a multi-parameter investigation of olcanic plume … · i also owe a thank you to mr. bob meyer...

A MULTI-PARAMETER INVESTIGATION OF VOLCANIC

PLUME BEHAVIOR AND RESULTANT ENVIRONMENTAL

IMPACT AT A PERSISTENTLY DEGASSING VOLCANO, MASAYA, NICARAGUA

by

Patricia Amanda Nadeau B.Sc. Hons., McGill University, 2004

THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF SCIENCE

In the Department

of Earth Sciences

© Patricia Amanda Nadeau 2006

SIMON FRASER UNIVERSITY

Fall 2006

All rights reserved. This work may not be reproduced in whole or in part, by photocopy

or other means, without permission of the author.

ii

APPROVAL

Name: Patricia Amanda Nadeau

Degree: Master of Science

Title of Thesis: A Multi-Parameter Investigation of Volcanic Plume Behavior and Resultant Environmental Impact at a Persistently Degassing Volcano, Masaya, Nicaragua

Examining Committee:

Chair: Dr. Andrew Calvert Professor, Department of Earth Sciences

___________________________________________

Dr. Glyn Williams-Jones Senior Supervisor Assistant Professor, Department of Earth Sciences

___________________________________________

Dr. Diana Allen Supervisor Associate Professor, Department of Earth Sciences

___________________________________________

Dr. Kirstie Simpson Supervisor Volcanologist, Geological Survey of Canada

___________________________________________

Dr. John Stix External Examiner Associate Professor, Department of Earth and Planetary Sciences McGill University

Date Defended/Approved: ___________________________________________

iii

ABSTRACT

Recent advances in volcanic gas sensing technology make possible detailed

investigations of the behavior of boundary layer gas plumes. An intensive survey of SO2 flux

during two month-long campaigns revealed an apparent decrease of 30-50% downwind of

Masaya volcano, Nicaragua. Dry deposition of S and aerosol conversion cannot account for

such losses. Local topography modifies regional trade winds, causing plume acceleration.

The resulting along-axis dilution of the plume leads to underestimates of total SO2

emissions. This apparent depletion can be accounted for by accurate determination of

plume speed at the location of each SO2 flux measurement. Interaction of acidic plumes

with elevated topography results in widespread vegetation damage downwind, which may be

characterized by a multi-parameter approach incorporating ground-based datasets and

Landsat NDVIs at Masaya and other volcanic systems. A thorough understanding of plume

behavior is essential for accurate evaluation of volcanic SO2 output and resultant

environmental impacts.

Keywords: Masaya volcano; volcanic SO2 emissions; FLYSPEC; remote sensing; environmental impact

iv

ACKNOWLEDGEMENTS

I would like to thank my senior supervisor, Glyn Williams-Jones for his guidance,

support, and enthusiasm over the course of this project – there were definitely times when I

needed that enthusiasm. I was his first grad student, he was my first graduate supervisor – I

think we each survived the breaking-in process pretty well! I certainly wouldn’t be going on

to a Ph.D. if I didn’t have a great experience with him here at SFU. I would also like to

thank my supervisory committee members, Diana Allen and Kirstie Simpson, for all of their

help and input along the way, especially considering that volcanic SO2 was a bit off the

beaten path for them. Thanks also to John Stix, my external examiner, for taking time out of

his exceedingly busy schedule to review this thesis.

Most of this research would have been impossible without the help of a great group

of field assistants – I’d like to thank Kirstie Simpson for enduring many long days with no

relief from her driving duties; Guillaume Mauri, who I’m sure would have much preferred to

be somewhere else doing SP; and Annie Bérubé, Marianne Gagnon, and Geneviève Pépin,

all of whom sacrificed a big chunk of their vacation time to help me out – I hope they got as

much out of having them around as I did. I also owe a thank you to Jessica Liggett for (very

patiently) teaching me how to drive standard, which kept the 2006 group from suffering the

same fate as Kirstie. I’m also grateful for the help of a number of people in Nicaragua: the

staff of Parque Nacional Volcán Masaya for allowing us access wherever we needed it;

Carlos Molina-Palma and his family for their incredible hospitality and help with just about

anything we needed; and Sergio for saving us with his impeccable painting skills. Many

thanks to Tamsin Mather, Lizzette Rodríguez, and Matt Watson for their help with aerosol

data processing, and to John Porter and Norm O’Neill for lending us Microtops

instruments. I’d also like to thank Keith Horton for making sure we always had working

FLYSPECs, even when that required emergency fed-ex-ing of things to Nicaragua. Thanks

also go to Doew Steyn for his help in figuring out wind issues.

Big thanks to my friends in the grad group at SFU – there were many talks, coffee

breaks, and pub trips that helped keep me sane over the course of this thesis! Thanks also to

v

my other friends scattered from Colonie to Montreal to wherever else – I wouldn’t have

made it without you guys either. I also owe a thank you to Mr. Bob Meyer from Lisha Kill

over ten years ago – I have to think I wouldn’t be here where I am if I hadn’t enjoyed my 8th

grade Earth Science class so much. Finally, thanks to my family, especially my parents, for

their unfailing support in whatever I do, even if they can’t quite figure out how exactly I

ended up doing this whole “volcano thing.” I wouldn’t be here without their help.

vi

TABLE OF CONTENTS

Approval __________________________________________________________ ii Abstract __________________________________________________________ iii Acknowledgements _____________________________________________________ iv Table of Contents ______________________________________________________ vi List of Figures _________________________________________________________ ix List of Tables _________________________________________________________ xiv Quotation __________________________________________________________xv Chapter 1: Introduction ________________________________________________1

1.1 Background ........................................................................................................................1 1.2 Geologic Setting ................................................................................................................2 1.3 Volcanic Activity ...............................................................................................................4 1.4 Previous Studies ................................................................................................................6 1.5 Objectives...........................................................................................................................8 1.6 References ........................................................................................................................11

Chapter 2: Beyond COSPEC: Recent Advances in SO2 Monitoring Technology _______________________________________________ 15

2.1 Abstract.............................................................................................................................15 2.2 Introduction .....................................................................................................................15 2.3 Instrumentation...............................................................................................................16

2.3.1 Mini-DOAS ..............................................................................................................16 2.3.2 FLYSPEC ................................................................................................................18

2.4 Field Methodology ..........................................................................................................21 2.4.1 Traditional methods .....................................................................................................21 2.4.2 Walking traverses ........................................................................................................21 2.4.3 Automated scanning ....................................................................................................21 2.4.4 Automated networks....................................................................................................23

2.4.4.1 Potential enhancements to data acquisition networks ..............................................26 2.4.5 Plume speed / velocity ..................................................................................................29

2.5 Other Gas Species...........................................................................................................35 2.6 Additional Remote Gas Sensing Techniques ..............................................................36 2.7 Aerosols ............................................................................................................................38 2.8 Conclusions......................................................................................................................38 2.9 References ........................................................................................................................39

vii

Chapter 3: SO2 Flux Measurements at Masaya Volcano, Nicaragua: Apparent Downwind Depletion as a Result of Topographic Modification of Winds_______________________________________43

3.1 Abstract.............................................................................................................................43 3.2 Introduction .....................................................................................................................43 3.3 Methodology ....................................................................................................................45

3.3.1 SO2 flux .....................................................................................................................45 3.3.2 Plume speed .................................................................................................................48 3.3.3 Aerosols ......................................................................................................................53

3.4 Uncertainties ....................................................................................................................54 3.4.1 Spectrometer bias .........................................................................................................54 3.4.2 Plume speed .................................................................................................................55 3.4.3 Aerosol measurements ..................................................................................................55

3.5 Results...............................................................................................................................56 3.5.1 SO2 flux .....................................................................................................................56 3.5.2 Aerosols ......................................................................................................................61

3.6 Discussion ........................................................................................................................61 3.6.1 Variability in Masaya’s SO2 emissions........................................................................61 3.6.2 Apparent loss of SO2 with increasing distance from vent................................................64 3.6.3 Possible sources of loss ..................................................................................................64

3.6.3.1 Deposition of sulfur .......................................................................................................64 3.6.3.2 Conversion of SO2 to sulfate aerosols.........................................................................65 3.6.3.3 Topographic modification of winds ............................................................................66

3.7 Conclusions......................................................................................................................75 3.8 Acknowledgements .........................................................................................................77 3.9 References ........................................................................................................................78

Chapter 4: A Multi-Parameter Evaluation of the Environmental Effects of Two Low-Lying, Persistently Degassing Volcanoes: Masaya, Nicaragua and Poás, Costa Rica ______________________________82

4.1 Abstract.............................................................................................................................82 4.2 Introduction .....................................................................................................................82 4.3 Background ......................................................................................................................84 4.4 Methodology ....................................................................................................................90 4.5 Results.............................................................................................................................101 4.6 Discussion ......................................................................................................................115 4.7 Conclusions....................................................................................................................130 4.8 References ......................................................................................................................131

Chapter 5: Conclusions ______________________________________________ 136 5.1 General Conclusions.....................................................................................................136 5.2 Recommendations for Future Work ..........................................................................137

Appendix A: Comprehensive Methodologies ______________________________ 139 A.1 SO2 Traverses.................................................................................................................139 A.2 Aerosol Traverses..........................................................................................................144 A.3 Plume Speed...................................................................................................................146 A.4 References ......................................................................................................................149

viii

Appendix B: Data ____________________________________________________ 150 B.1 2005 and 2006 SO2 traverses along the Ticuantepe and Llano Pacaya

roads. ...............................................................................................................................150 B.2 Previous SO2 fluxes (1972-2004) ................................................................................166 B.3 2006 comparison of FLYSPEC and FLYSPEC v2 on 2/26/2006 .......................171 B.4 2005 and 2006 spectrometer-based plume speed data as determined from

dual spectrometer method ...........................................................................................172 B.5 2005 and 2006 wind speed data as measured by hand-held anemometer ............176 B.6 Simultaneous aerosol/SO2 traverse data as measured on Ticuantepe and

Llano Pacaya roads........................................................................................................178 B.7 Profile data of NDVI, S dry deposition, and SO2 ground-level

concentration .................................................................................................................183 B.7.1 NDVI .....................................................................................................................183 B.7.2 Dry deposition of S ....................................................................................................188 B.7.3 SO2 ground-level concentration....................................................................................193

B.8 References ......................................................................................................................198

ix

LIST OF FIGURES

Figure 1-1: Geographic map of western Nicaragua showing the location of Masaya volcano with respect to other Nicaraguan volcanoes.............................................3

Figure 1-2: Relief image of Masaya Caldera and surrounding region. Volcanic cones within the caldera are labeled in red, while craters are labeled in yellow. UTM zone 16n. .............................................................................................5

Figure 1-3: Map of road network in the vicinity of Masaya volcano. Roads utilized for measuring SO2 flux, the Ticuantepe and Llano Pacaya roads, are shown. .........................................................................................................................10

Figure 2-1: Configuration of an Ocean Optics USB 2000 mini-DOAS spectrometer, coupled to a telescope via an optical fiber (figure not to scale). Dashed and non-dashed lines drawn within the spectrometer represent the paths of light at two distinct wavelengths, displaying how the spectrometer disperses radiation on to the CCD array. ................................17

Figure 2-2: Components of a FLYSPEC: an Ocean Optics USB 2000 spectrometer, sub-notebook computer, GPS with antenna, high and low calibration SO2 gas cells mounted above the spectrometer, and telescope. The “telescope” is a fiber-optic collimating lens mounted directly to the spectrometer input aperture. The lens, in combination with the UV band-pass filter window mounted on the case, provides a field of view of approximately 2.5°. Power for the spectrometer and GPS is supplied by the computer. ...........................................................................20

Figure 2-3: Example of a FLYSPEC field deployment. The spectrometer unit is mounted near the car’s side mirror with duct tape, while the GPS antenna is affixed to the roof. ..................................................................................22

Figure 2-4: A walking traverse with a backpack-mounted FLYSPEC at Vulcano, Italy ..............................................................................................................................24

Figure 2-5: An example of a scanning DOAS-based spectrometer. (modified from Edmonds et al., 2003) ...............................................................................................25

x

Figure 2-6: Calculating plume height and position from the SO2 concentration and angular data from two fixed scanning spectrometers. Angles α1 and α2 are determined from the scan angles coinciding with peaks in SO2 concentration during one revolution. Tan α1=h/d1 and tan α2=(h+z)/d2, where h is plume height and z is any elevation difference between the locations of the two spectrometers. The known horizontal distance between the spectrometers is equal to d1+d2 (modified from Edmonds et al., 2003) ...................................................................27

Figure 2-7: Schematic diagram of a degassing volcano, plume with cross-section, and a possible configuration for proposed network of stationary wide-angle miniature spectrometers, with their expanded fields of view delineated by dotted lines. ........................................................................................28

Figure 2-8: Example of a network of stationary network of spectrometers with conventional fields of view, represented by dashed lines. Data from collection points closely approximate the plume shape and average concentration pathlength of a traditional traverse. An example traverse is represented by a dark curve, and the plume approximated from the four spectrometers is marked by the gray-striped area. .......................................30

Figure 2-9: Field deployment for determination of plume speed at Masaya volcano, Nicaragua. The FLYSPECs were mounted downwind of the gas source on small, lightweight camera tripods separated by 20 m. Measurements were made for 30 minutes, with a GPS antenna in each configuration to ensure time-synchronization of the datasets. ...........................33

Figure 2-10: (a) The SO2 pathlength concentration for two FLYSPECs at Masaya volcano, Nicaragua on March 25, 2003. The instrument separation is 40.5 m, determined by tape measure. Inset is a 4-minute window showing an apparent time separation (16 s) between the 2 signals. (b) The SO2 signals for the entire 30-minute sampling period are compared to each other for timeshifts between -60 and 60 s at 0.1 s iterations. The maximum correlation coefficient (r2 = 0.959) for the signals occurs at a time difference of 13.1 s, which for a 40.5 m separation, results in a plume speed of 3.1 m/s....................................................34

Figure 3-1: Mean SO2 fluxes (normalized to a plume speed of 1 m/s) grouped by month. Error bars represent one standard deviation of repeat measurements. Note break in x-axis. See also Appendix B.1-2........................46

Figure 3-2: Map of road network in the vicinity of Masaya volcano. Roads utilized for measuring SO2 flux, the Ticuantepe and Llano Pacaya roads, are shown. Star indicates the location of the active vent. .........................................47

Figure 3-3: FLYSPEC components and field deployment. (a) FLYSPEC (b) FLYSPEC v2..............................................................................................................49

xi

Figure 3-4: Mean daily SO2 flux (normalized to a plume speed of 1 m/s) as measured on each of the two roads downwind of Masaya’s active vent. Error bars represent one standard deviation of repeat measurements. Note break in x-axis. See Tables 3-1 and 3-2. ......................................................59

Figure 3-5: Histograms displaying the variability of SO2 flux measurements on the Ticuantepe road for (a) a month (February 16, 2006 – March 12, 2006), and (b) a day (March 4, 2006) ..................................................................................63

Figure 3-6: Example of simultaneous particle and SO2 column measurements from March 8, 2006.............................................................................................................67

Figure 3-7: Digital elevation model of Masaya region highlighting the complex topography between the active vent and the two roads downwind (shown in red). Vertical exaggeration is 5x. ..........................................................69

Figure 3-8: Example of the effect of increasing plume speed on measured concentration path-lengths of SO2. A twofold increase in plume velocity corresponds to a twofold decrease in the amount of SO2 sensed by the spectrometer at a given time. Spectrometer fields of view are denoted by dashed gray lines. Continuous plume is portrayed as discrete puffs for illustrative purposes. ..............................................................70

Figure 3-9: A generalized pattern of airflow over a two-dimensional ridge. The point of maximum wind velocity is indicated by a dark gray star, and the points of minimum wind velocity are denoted by light gray stars. Variables for use in Equation 3-2 are also shown.................................................71

Figure 3-10: Photographs of Masaya volcano’s plume as seen from the space shuttle, with the active vent indicated by a star. The portion of the plume over the Llano Pacaya ridge is clearly less concentrated than other portions of the plume. Photos are from November 9, 1984 (top photo; NASA image STS 051a-32-066) and January 13, 1986 (bottom photo; NASA image STS061c-37-076). Images courtesy of the Image Science and Analysis Laboratory, NASA-Johnson Space Center.......................73

Figure 4-1: Geographic map of western Nicaragua showing the location of Masaya volcano with respect to other Nicaraguan volcanoes...........................................85

Figure 4-2: Relief image of Masaya Caldera. Volcanic cones within the caldera are labeled in red, while craters are labeled in yellow. UTM zone 16n. ..................86

Figure 4-3: The actively degassing vent in Santiago crater. Vent’s diameter is estimated to be approximately 30-40 m. ................................................................87

xii

Figure 4-4: (a) and (b) Views of the Llano Pacaya ridge looking west from the Ticuantepe road. (a) is a clear view in which the plume is passing unimpeded over the ridge. (b) depicts an instance of intense fumigation of the ridge by the plume such that the ridge is barely visible. (c) the landscape in the devastated region of the Llano Pacaya ridge, looking southeast. Note the lack of vegetation beyond dry grasses and a few small shrubs.................................................................................89

Figure 4-5: Geographic map of northwestern Costa Rica showing the location of Poás volcano with respect to other Costa Rican volcanoes. ...............................91

Figure 4-6: Map of the summit area of Poás volcano. UTM zone 16n. ..............................92

Figure 4-7: The active crater of Poás volcano, with Laguna Caliente and fumarole fields. Diameter of Laguna Caliente is approximately 300-350 m.....................93

Figure 4-8: Composite images of Landsat ETM+ red, green, and blue bands for each of the three scenes utilized, with important features labeled. ....................94

Figure 4-9: Digital elevation model of Masaya region. Masaya caldera is outlined in red for reference. ...................................................................................................99

Figure 4-10: Digital elevation model of Poás region. Laguna Caliente is outlined in light blue for reference............................................................................................100

Figure 4-11: Interpolated distributions of S dry deposition and SO2 ground level concentration derived from data in Delmelle et al. (2001, 2002). (a) S dry deposition. (b) SO2 ground level concentration. Shaded DEM and caldera (outlined in red) are shown for reference. ..............................................102

Figure 4-12: Average SO2 concentration path-lengths over Ticuantepe and Llano Pacaya roads in 2005 and 2006. Color scheme is relative; data from the two roads were averaged separately such that a color on one road is not indicative of the same concentration path-length on the other. Shaded DEM surface and red caldera outline are shown for reference. .........104

Figure 4-13: Location of profiles used for dataset sampling on Masaya data. Eight profiles originating from the active vent are shown in yellow. Also shown is a light blue ‘control’ profile, used to examine data relationships in an area relatively unaffected by the plume. Shaded DEM surface and red caldera outline are shown for reference. .......................105

Figure 4-14: Location of profiles used for dataset sampling on Poás data. Five parallel profiles originating in an unaffected area upwind of the active crater are shown in yellow. Shaded DEM surface and light blue Laguna Caliente outline are shown for reference. ..............................................106

xiii

Figure 4-15: Results of the NDVI transformations for each of the three scenes utilized, with Masaya caldera outlined in red for reference. (a) 2001 Masaya NDVI (b) 2004 Masaya NDVI (c) Poás NDVI. ................................107

Figure 4-16: Residual results of the subtraction of regional elevation-dependent trends from NDVI at Masaya, with caldera outlined in red for reference. (a) 2001 Masaya residuals (b) 2004 Masaya residuals.....................110

Figure 4-17: The area exhibiting the most severe degree of vegetation damage at Masaya, as determined from NDVI residual data from 2001 (background image shown here). The prominent zone of vegetation kill is delineated by the striped area, and the red caldera outline is included for reference. ............................................................................................114

Figure 4-18: Results of profiles of DEM and NDVI residual (or NDVI, as at Poás) datasets. Elevation profiles are shown by a black line, and NDVI data by a gray line. See Figures 4-13 and 4-14 for location of profiles. All profiles are shown with the eastern end at the origin. (a) 8 Masaya profiles (b) Masaya control profile (c) 5 Poás profiles .....................................116

Figure 4-19: Plot of the relationship between NDVI and elevation along Masaya profile 4. Small and large arrows highlight departures from the positive region trend and correspond to the El Panama and Llano Pacaya regions, respectively....................................................................................124

Figure 4-20: Perspective view of 2001 Masaya NDVI draped over DEM. Note excessively damaged areas on the highlands of El Panama and Llano Pacaya. Vertical exaggeration is 7.5x....................................................................125

Figure 4-21: Data along Masaya profile 4, showing elevated dry deposition velocities and ground-level concentrations of SO2 persisting beyond the area marked by vegetation damage as reflected by NDVI residual values. .............127

xiv

LIST OF TABLES

Table 3-1: Comparison of simultaneous measurements of ground-level wind speed by anemometer and plume speed measurements made according to the dual-spectrometer method. ...........................................................................52

Table 3-2: Summary of SO2 flux data as measured on the Ticuantepe road. ......................57

Table 3-3: Summary of SO2 flux data as measured on the Llano Pacaya road....................58

Table 3-4: Plume particle concentrations for five pairs of aerosol traverses. T and LP in traverse name indicate Ticauntepe and Llano Pacaya roads, and F and F2 represent FLYSPEC and FLYSPECv2, respectively. .........................62

xv

QUOTATION

"May every young scientist remember... and not fail to keep

his eyes open for the possibility that an irritating failure of

his apparatus to give consistent results may once or twice in a

lifetime conceal an important discovery."

~Patrick Blackett

Chapter 1 – Introduction

1

Chapter 1: INTRODUCTION

1.1 Background

Measured fluxes of volcanic SO2 emissions are important in the evaluation of the

state of activity of a particular volcano. Changes in SO2 flux from an established baseline

often indicate a coincident change in eruptive activity (e.g., Zapata et al., 1997). Quantities

of degassed SO2 can also facilitate estimations of degassed magma volumes within the

volcanic conduit as well as evaluation of volcanic contributions to global emissions of SO2 to

the atmosphere (e.g., Andres and Kasgnoc, 1998). Likewise, accurate representations of SO2

degassing aid in the assessment of the ramifications of the degassing on the local

environment (e.g., Delmelle et al., 2001, 2002). Because of the extent of the applicability of

SO2 flux data in geologic and environmental applications, the large errors often associated

with reported SO2 fluxes are of concern (e.g., Williams-Jones et al., 2006).

While SO2 is the easiest of volcanic gases to monitor, being the third most abundant

volcanic gas and present in the ambient atmosphere in only negligible amounts, direct

sampling of volcanic SO2 or other measurements very proximal to an active vent are not

always feasible. Active craters or fumaroles are often dangerous and inaccessible. As such,

SO2 is often measured through remote methods. Because SO2 absorbs specific wavelengths

of light in the ultraviolet range, measurements can easily be accomplished with the use of a

UV spectrometer, such as the FLYSPEC (Horton et al., 2006). By aiming the instrument’s

field of view at a parcel of air containing SO2, the amount of SO2 in the column of air in

question can be calculated. However, even when measuring SO2 remotely, field conditions

proximal to volcanoes are often less than ideal. Accordingly, SO2 data collection involves

techniques which are inherently limited and, unless accounted for, lead to erroneous

calculations of fluxes. Most importantly, lack of accessibility at most volcanoes requires data

to be collected many kilometers downwind of the degassing vent. Between the source vent

and the data acquisition site, a number of physical and chemical processes may take place,

thereby altering the amount of SO2 within the plume. Conversion of SO2 to sulfate aerosols

may play a role in measured fluxes being less than source vent fluxes, especially in conditions


2

with relative humidity near 100% (Eatough et al., 1994). Direct wet or dry deposition from

the plume also decreases the SO2 content of plumes (Delmelle et al., 2001). Dispersal and

dilution of the plume as it propagates away from the vent and mixes with the ambient

atmosphere are other factors adding to potential uncertainties in flux estimates. Topography

may also affect the extent to which each of these processes occurs. Further complicating

accurate representation of SO2 fluxes are natural variations in the volcano’s gas output on

timescales that may escape detection due to limitations of the temporal resolution of the data

collection (Branan et al., submitted). Furthermore, estimations of SO2 flux rely directly on

determination of the speed at which the measured plume propagates, often a difficult

parameter to accurately quantify.

Masaya volcano in western Nicaragua (Figure 1-1) presents a near-ideal locale to

investigate the changes which may affect a volcanic plume following emission from the vent.

The volcano has been emitting a steady, non-eruptive plume of gas for over 150 years, with

especially high levels of degassing since the onset of a degassing crisis in 1993 (Williams-

Jones et al., 2003). The local road network, with two roughly parallel roads at ~5 km and

~15 km downwind of the summit and perpendicular to the plume’s direction of

propagation, provides the infrastructure necessary to adequately assess the plume at two

different stages of its downwind evolution. Additionally, the more distal road lies on a ridge

with an elevation higher than that of Masaya’s summit. This offers a chance to investigate

the degree to which the plume interacts with the surrounding topography and the extent to

which that interaction affects traditional SO2 flux measurements.

1.2 Geologic Setting

Masaya volcano is just one of many volcanoes that dot the western coast of Central

America. All of the volcanoes were formed as a result of the oblique subduction of the

Cocos plate beneath the Caribbean Plate, and comprise what is known as the Central

American Volcanic Front (CAVF). The northern extent of the CAVF is Tacaná volcano in

Guatemala; the southern is Irazú volcano, Costa Rica. Between those two volcanoes exists a

string of volcanoes with a greater spatial density (average spacing of volcanic centers

~25 km) than other subduction zones around the world (Carr, 1984). Nicaragua is situated

at the center of a symmetrical pattern of physical and chemical properties of the rocks of the


3

Figure 1-1: Geographic map of western Nicaragua showing the location of Masaya volcano with respect to other Nicaraguan volcanoes.


4

CAVF: Nicaraguan volcanoes have the smallest edifice heights, lowest SiO2 contents, and

highest density lavas. To the north and south, the volcanoes of the CAVF generally are

taller and more silicic (Carr, 1984 and references therein). The Nicaraguan region represents

the segment of the CAVF with the thinnest underlying continental crust (Carr, 1984), and

the Masaya area comprises some of the only tholeiitic basalts on the CAVF (Ui, 1972; Carr,

1984). Structurally, the Masaya region is marked by the complex interplay of the Nicaraguan

Depression and the Managua Graben with the Masaya volcanic center itself (Girard and

van Wyk de Vries, 2005).

1.3 Volcanic Activity

Situated ~20 km southeast of the Nicaraguan capital of Managua, Masaya volcano is

part of a complex series of basaltic shields, cones, and calderas. The largest feature in the

complex is the Las Sierras shield, a series of dacitic to basaltic lava flows, ignimbrite deposits,

and lahar deposits (van Wyk de Vries, 1993). A large basaltic ignimbrite erupted at

approximately 30,000 BP, forming the Las Sierras (or Las Nubes) caldera (Bice, 1980; van

Wyk de Vries, 1993; Sebesta, 1997). It was within the Las Sierras caldera that a series of

basaltic lava flows formed the Masaya shield volcano. The oblong Masaya caldera (~11 by

~6 km) formed on the new shield as the result of basaltic ignimbrite eruptions between

~6500 and 2250 BP (Williams, 1983). Further edifice-building activity followed, creating the

Masaya and Nindiri cones (Rymer et al., 1998; Williams-Jones et al., 2003). Repeated vertical

oscillations of the near-surface magma column within these cones led to the formation of a

series of pit craters on both Nindiri and Masaya cones (Figure 1-2). The most recently

formed, and currently active, crater is Santiago, located on the eastern flank of Nindiri cone,

and is believed to have formed in 1853 (McBirney, 1956; Rymer et al., 1998).

Large historic explosive eruptions have not occurred at Masaya, and lava flows are

rare as of late, with the most recent being a fissure eruption on the flanks of Masaya cone in

1772 (Rymer et al., 1998). Since the last extrusive activity, lava lakes have appeared in the pit

craters at least twice (possibly in Nindiri crater in 1852, and in Santiago crater in 1948 and

1965), and smaller pools of lava, with associated small-scale lava fountaining (~40 m high),

have also appeared within the vents of the Santiago crater in 1989 and 1993 (Rymer et al.,

1998). Minor eruptions of ash and gas have recurred on time scales of months to years


5

Figure 1-2: Relief image of Masaya Caldera and surrounding region. Volcanic cones within the caldera are labeled in red, while craters are labeled in yellow. UTM zone 16n.


6

(e.g., Smithsonian Institution, 1989; 1997; 2001), with the most recent occurring in March,

2005 (Smithsonian Institution, 2006). Currently, activity is largely restricted to persistent

SO2 degassing from Santiago pit crater, where nocturnal incandescence of the two main

vents suggests the presence of a very shallow magma surface.

1.4 Previous Studies

Beginning with descriptions from the times of the Spanish Conquistadors, numerous

observations and studies of Masaya have been documented over the years. The physical

volcanology was first recorded by Sapper over the years from 1897 to 1927 (McBirney,

1956). McBirney (1956) gave detailed descriptions of Masaya caldera and the evolution of its

volcanic features. The early 1980s were marked by a number of investigations into the

petrology of the Masaya complex (e.g., Bice, 1980; Williams, 1983; Carr, 1984). In more

recent years, studies have focused on the interpretation of the structure and formation of the

Masaya caldera, and the behavior of the underlying magmatic system (e.g., Gregg and

Williams, 1996; Girard and van Wyk de Vries, 2005). These studies are often supplemented

with data from various geophysical survey techniques (e.g., Connor and Williams, 1990;

Métaxian et al., 1997; Rymer et al., 1998; Williams-Jones et al., 2003).

In addition to studies of the physical volcanology at Masaya, the persistent degassing

at and relative accessibility of Masaya have made it an ideal site for the study of volcanic

gases. SO2 monitoring by means of UV spectroscopy began with COSPEC measurements

by Stoiber et al. (1986) in 1972, with numerous studies of Masaya’s SO2 emissions having

been conducted since then (e.g., St-Amand, 1999; Delmelle et al., 1999; Williams-Jones et al.,

2003; McGonigle et al., 2004). Horrocks et al. (1999) evaluated the SO2/HCl and HCL/HF

ratios of the emitted plume through the use of open-path Fourier Transform Infrared

spectroscopy. Allen et al. (2002) utilized direct sampling methods (filters and diffusion

tubes) to quantify the initial amount of sulfate aerosol present in the plume upon emission.

Mather et al. (2003) expanded on research in the aerosol field by using a multi-stage

impactor to assess the concentrations of a number of soluble aerosol species in the plume.

Another similar study by Mather et al. (2006) measured sulfate aerosol flux as well as

SO2/HF and SO2/HCl ratios.


7

A number of studies at Masaya have also begun to investigate the effect of the plume

on the downwind environment. Dry deposition of sulfur and other acidic species from the

plume, as well as plume dispersion, were investigated by Delmelle et al. (2001; 2002).

Delmelle et al. (2003) conducted a survey of soils in the Masaya region, finding that

fumigation of the region by Masaya’s plume had led to significant degrees of soil

acidification by SO2, HCl, and HF. Delfosse et al. (2005a, b) examined the Masaya Andosols

as well, concluding that much of the volcanic sulfur impacting the more weathered soils of

the Llano Pacaya region is sequestered from the soil in the form of aluminum-hydroxy-

sulfate minerals, which prevent acidification to the extent found closer to the volcano in the

Ticuantepe region. Delmelle et al. (2001; 2002) and studies by Nicholson et al. (1996) and

Sandoval et al. (1996) at Poás volcano, Costa Rica, document the deleterious effects of direct

volcanic plume fumigation on local vegetation.

Aside from studies of the volcano and plume themselves, Masaya has also been the

site of a number of studies aimed at bettering the techniques for the measurement of SO2. A

key advancement in the improvement of constraints on plume speed was made in the study

by Williams-Jones et al. (2006). A dual-spectrometer method for measuring plume speed

was introduced, whereby similar signals appearing in the data of two time-synchronized

spectrometers aligned beneath the plume downwind of the degassing vent were correlated.

Based on the known distance separation between the spectrometers and the time lag

between the signals in the two datasets, plume speed was derived. Field campaigns at both

Kilauea and Masaya volcanoes were conducted, yielding more accurate plume speeds than

could be determined from ground-based anemometers. Similar experiments using three

spectrometers to determine plume velocity were also performed at Masaya (McGonigle et al.,

2005). Masaya was also the site of field trials implementing scanning and walking traverse

methods of measuring SO2 plumes (McGonigle et al., 2002).

Plume depletion rate and aerosol conversion were addressed by a number of studies,

at Masaya and other volcanoes: Oppenheimer et al. (1998) conducted a campaign at

Soufrière Hills volcano, Montserrat, concluding that loss of SO2 from the plume does cause,

to some degree, underestimation of true volcanic SO2 flux. The study, however, was based

on only two days of data, and relied largely on ground-based wind speed measurements in

the determination of flux and plume age. Plume geometry may also have been altered by the


8

variable winds associated with coastal environments, further skewing data. More recently,

McGonigle et al. (2004) performed a study showing that conversion of SO2 to sulfate

aerosols at Masaya was insignificant, as indicated by negligible variation in apparent SO2 flux

with varying plume age, temperature, and relative humidity. However, no actual

measurements of the plume’s aerosol content were conducted, and the basis for flux

calculations and plume age determination was ground-based wind speed. A better means by

which to assess aerosol content of plumes and associated fluxes and conversion rates was

demonstrated in Porter et al. (2002). Experiments at Kilauea volcano, Hawaii, integrated the

use of a sun photometer to measure the plume’s aerosol content with a UV spectrometer

measuring SO2 flux at various distances downwind of the degassing vent. However, flux

measurements were again based on only simple ground-based wind speed measurements as

proxies for plume velocities.

1.5 Objectives

The main objective of this study is to develop methods which more accurately

measure volcanic SO2 fluxes, using Masaya volcano as a natural laboratory. Individual goals

to be met are: to further build upon an extensive time series of flux data in order to account

for variations over small time-scales; to utilize new plume speed determination methods to

better constrain what is normally the greatest source of error in flux calculations; and to

complete simultaneous SO2 and aerosol flux measurements at each of two distances

downwind to account for obvious changes of the plume’s characteristics with distance.

Many attempts at improving monitoring techniques and constraints on SO2 fluxes

have been previously undertaken (Oppenheimer et al., 1998; Delmelle et al., 2001; Porter et

al., 2002; Gerlach, 2003; McGonigle et al., 2004; Williams-Jones et al., 2006), though often

independently of each other. Unique to this study is the integration of the monitoring

methods and the addition of an intensive campaign to monitor the plume’s aerosol content.

Measurements of both the plume’s aerosol and SO2 content on each of two roads at

different distances downwind (Ticuantepe and Llano Pacaya roads, Figure 1-3) and advanced

methods in plume speed measurements will give new insight into the behavior of Masaya’s

plume as it progresses through time and space.


9

A secondary objective of this project is to investigate environmental effects of

Masaya’s low-lying volcanic plume. Remotely-sensed data, such as Landsat-derived

vegetation indices (NDVI in particular) and digital elevation models (DEMs), are integrated

with previously existing ground-based sulfur deposition (Delmelle et al., 2001), ground level

SO2 concentration data (Delmelle et al., 2002), and SO2 flux data (Chapter 3; Appendix B.1-

3). Such datasets help in identifying the location and extent of the plume’s devastating effect

on vegetation. A second volcano, Poás, is also examined through the use of a DEM and

NDVI. As Poás has different geography, climate, vegetation, and volcanic structure from

Masaya, comparison of the datasets from the two systems offers insight into the applicability

of remote sensing methods to the detection of environmental effects at different types of

volcanoes.

The nature of the two separate objectives of this thesis lends itself to the separation

of the two ideas into independent chapters. This separation also facilitates the later

submission of each chapter to scientific journals. As such, chapter 3 of this thesis focuses

on the issue of the SO2 plume at Masaya volcano and the associated hurdles that come with

attempts at its quantification. Chapter 4 deals with the ability of remote sensing methods to

detect interactions of volcanic plumes with topography and incipient effects on the

surrounding vegetation. A preceding chapter is a literature review describing the various

manners in which volcanic SO2 is remotely measured. Each of these three main chapters is

already in article format for journal submission. Accordingly, in-depth methodologies and

comprehensive datasets have been omitted from the chapters themselves and have instead

been placed in Appendices A and B, respectively. Furthermore, because of the individual

articles, some repetition between chapters is unavoidable, especially in instances of

background information and figures.


10

Figure 1-3: Map of road network in the vicinity of Masaya volcano. Roads utilized for measuring SO2 flux, the Ticuantepe and Llano Pacaya roads, are shown.


11

1.6 References

Allen, A.G., Oppenheimer, C., Ferm, M., Baxter, P.J., Horrocks, L.A., Galle, B., McGonigle, A.J.S., and Duffell, H.J., 2002. Primary sulfate aerosol and associated emissions from Masaya Volcano, Nicaragua. Journal of Geophysical Research, D, Atmospheres, 107(D23), 4682, doi:10.1029/2002JD002120.

Andres, R.J., and Kasgnoc, A.D., 1998. A time-averaged inventory of subaerial volcanic sulfur emissions. Journal of Geophysical Research, 103: 25251-25261.

Bice, D.C., 1980. Tephra strigraphy and physical aspects of recent volcanism near Managua, Nicaragua. PhD Thesis, University of California, Berkeley, CA, USA.

Branan, Y.K., Harris, A., Watson, I.M., Phillips, J.C., Horton, K., Williams-Jones, G., and Garbeil, H. Investigation of at-vent dynamics and dilution of gas puffing using thermal infrared radiometers at Masaya volcano, Nicaragua. Journal of Volcanology and Geothermal Research, submitted.

Carr, M.J., 1984. Symmetrical and segmented variation of physical and geochemical characteristics of the Central American Volcanic Front. Journal of Volcanology and Geothermal Research, 20: 231-252.

Connor, C.B., and Williams, S.N., 1990. Interpretation of gravity anomalies, Masaya caldera complex, Nicaragua. In: Larue, D.K. and Draper, G. (Editors), Transactions of the 12th Caribbean Geological Conference, United States Virgin Islands: 495-502.

Delfosse, T., Delmelle, P., Givron, C., and Delvaux, B., 2005a. Inorganic sulphate extraction from SO2-impacted Andosols. European Journal of Soil Science, 56: 127-133.

Delfosse, T., Delmelle, P., Iserentant, A., and Delvaux, B., 2005b. Contribution of SO3 to the acid neutralizing capacity of Andosols exposed to strong volcanogenic acid and SO2 deposition. European Journal of Soil Science, 56: 113-125.

Delmelle, P., Baxter, P., Beaulieu, A., Burton, M., Francis, P., Garcia-Alvarez, J., Horrocks, L., Navarro, M., Oppenheimer, C., Rothery, D., Rymer, H., St. Amand, K., Stix, J., Strauch, W., Williams-Jones, G., 1999. Origin, effects of Masaya Volcano’s continued unrest probed in Nicaragua. EOS, Transactions of the American Geophysical Union, 80, 575-581.

Delmelle, P., Stix, J., Bourque, C.P.A., Baxter, P.J., Garcia-Alvarez, J., and Barquero, J., 2001. Dry Deposition and Heavy Acid Loading in the Vicinity of Masaya Volcano, a Major Sulfur and Chlorine Source in Nicaragua. Environmental Science and Technology, 35(7): 1289-1293.

Delmelle, P., Stix, J., Baxter, P.J., Garcia-Alvarez, J., and Barquero, J., 2002. Atmospheric dispersion, environmental effects and potential health hazard associated with the low-altitude gas plume of Masaya Volcano, Nicaragua. Bulletin of Volcanology, 64(6): 423-434.


12

Delmelle, P., Delfosse, T., and Delvaux, B., 2003. Sulfate, chloride and fluoride retention in Andosols exposed to volcanic acid emissions. Environmental Pollution, 126: 445-457.

Eatough, D.J., Caka, F.M., and Farber, R.J., 1994. The conversion of SO2 to sulfate in the atmosphere. Israel Journal of Chemistry, 34: 301-314.

Girard, G., and van Wyk de Vries, B., 2005. The Managua Graben and Las Sierras-Masaya volcanic complex (Nicaragua); pull-apart localization by an intrusive complex: results from analogue modeling. Journal of Volcanology and Geothermal Research, 144: 37-57.

Gregg, T.K.P., and Williams, S.N., 1996. Explosive mafic volcanoes on Mars and Earth: Deep magma sources and rapid rise rate. Icarus, 122: 397-405.

Horrocks, L., Burton, M., Francis, P., and Oppenheimer, C., 1999. Stable gas plume composition measured by OP-FTIR spectroscopy at Masaya Volcano, Nicaragua, 1998-1999. Geophysical Research Letters, 26(23): 3497-3500.

Horton, K.A., Williams-Jones, G., Garbeil, H., Elias, T., Sutton, A.J., Mouginis-Mark, P, Porter, J.N., and Clegg, S., 2006. Real-time measurement of volcanic SO2 emissions: Validation of a new UV correlation spectrometer (FLYSPEC). Bulletin of Volcanology, 68: 323-327.

Mather, T.A., Allen, A.G., Oppenheimer, C., Pyle, D.M., and McGonigle, A.J.S., 2003. Size-Resolved Characterisation of Soluble Ions in the Particles in the Tropospheric Plume of Masaya Volcano, Nicaragua: Origins and Plume Processing. Journal of Atmospheric Chemistry, 46: 207-237.

Mather, T.A., Pyle, D.M., Tsanev, V.I., McGonigle, A.J.S., Oppenheimer, C., and Allen, A.G., 2006. A reassessment of current volcanic emissions from the Central American arc with specific examples from Nicaragua. Journal of Volcanology and Geothermal Research, 149: 297-311.

McBirney, A.R., 1956. The Nicaragua volcano Masaya and its caldera. EOS, Transactions, American Geophysical Union, 37: 83-96.

McGonigle, A.J.S., Oppenheimer, C., Galle, B., Mather, T.A., and Pyle, D.M., 2002. Walking traverse and scanning DOAS measurements of volcanic gas emission rates. Geophysical Research Letters, 29(20), 1985, doi:10.10292002GL015827.

McGonigle, A.J.S., Delmelle, P., Oppenheimer, C., Tsanev, V.I., Delfosse, T., Williams-Jones, G., Horton, K., and Mather, T.A., 2004. SO2 depletion in tropospheric volcanic plumes. Geophysical Research Letters, 31, L13201, doi:10.1029/2004GL019990.

McGonigle, A.J.S., Hilton, D.R., Fischer, T.P., and Oppenheimer, C., 2005. Plume velocity determination for volcanic SO2 flux measurements. Geophysical Research Letters, 32, L11302, doi:10.1029/2005GL022470.


13

Métaxian, J.-P., Lesage, P., and Dorel, J., 1997. Permanent tremor of Masaya Volcano, Nicaragua: Wave field analysis and source location. Journal of Geophysical Research, 102: 22529-22545.

Nicholson, R.A., Roberts, P.D., and Baxter, P.J., 1996. Preliminary studies of acid and gas contamination at Poás volcano, Costa Rica. In: Appleton, J.D., Fuge, R., and McCall, G.J.H. (Editors), Environmental Geochemistry and Health, Geological Society, Special publications, 113: 239-244.

Oppenheimer, C., Francis, P., and Stix, J., 1998. Depletion rates of sulfur dioxide in troposheric volcanic plumes. Geophysical Research Letters, 25(14): 2671-2674.

Porter, J.N., Horton, K.A., Mouginis-Mark, P.J., Lienert, B., Sharma, S.K., Lau, E., Sutton, A.J., Elias, T., and Oppenheimer, C., 2002. Sun photometer and lidar measurements of the plume from the Hawaii Kilauea Volcano Pu'u O'o Vent; aerosol flux and SO2 lifetime. Geophysical Research Letters, 29(16), doi:10.1029/2002GL014744.

Rymer, H., van Wyk de Vries, B., Stix, J., and Williams-Jones, G., 1998. Pit crater structure and processes governing persistent activity at Masaya Volcano, Nicaragua. Bulletin of Volcanology, 59: 345-355.

Sandoval, L., Valdés, J., Martínez, M., Barquero, J., and Fernández, E., 1996. Efecto de las emisiones del Volcán Poás sobre la vegetación, Costa Rica. Ingenieria y Ciencia Quimica, 16: 62-64.

Sebesta, J., 1997. Dynamic development of the relief in the Managua area, Nicaragua. Acta Universitatis Carolinae/Geographica, 2: 93-109.

Smithsonian Institution, 1989. Masaya. Bulletin of the Global Volcanism Network, 14(6).




St-Amand, K., 1999. The distribution and origin of Radon, CO2, and SO2 gases and multifractal behaviour of SO2 at Masaya Volcano, Nicaragua. MSc Thesis, University of Montreal, Montreal, Canada.

Stoiber, R.E., Williams, S.N., and Huebert, B.J., 1986. Sulfur and halogen gases at Masaya caldera complex, Nicaragua: Total flux and variations with time. Journal of Geophysical Research, 91: 12215-12231.

Ui, T., 1972. Recent volcanism in the Masaya-Granada area, Nicaragua. Bulletin of Volcanology, 36: 174-190.

van Wyk de Vries, B., 1993. Tectonic and magma evolution of Nicaraguan volcanic systems. PhD Thesis, Open University, Milton Keynes, UK.

Williams, S.N., 1983. Plinian airfall deposits of basaltic composition. Geology, 11: 211-214.


14

Williams-Jones, G., Rymer, H., and Rothery, D.A., 2003. Gravity changes and passive SO2 degassing at the Masaya caldera complex, Nicaragua. Journal of Volcanology and Geothermal Research, 123: 137-160.

Williams-Jones, G., Horton, K.A., Elias, T., Garbeil, H., Mouginis-Mark, P.J., Sutton, A.J., and Harris, A.J.L., 2006. Accurately measuring volcanic plume velocity with multiple UV spectrometers. Bulletin of Volcanology, 68: 328-332.

Zapata G., J.A., Calvache V., M.L., Cortés J., G.P., Fischer, T.P., Garzon V., G., Gómez M., D., Narváez M., L., Ordóñez V., M., Ortega E., A., Stix, J., Torres C., R., and Williams, S.N., 1997. SO2 fluxes from Galeras Volcano, Colombia, 1989-1995: Progressive degassing and conduit obstruction of a Decade Volcano. Journal of Volcanology and Geothermal Research, 77: 195-208.

Chapter 2 – Beyond COSPEC

15

Chapter 2: BEYOND COSPEC: RECENT ADVANCES IN SO2

MONITORING TECHNOLOGY

*Accepted for publication as a chapter in The COSPEC Cookbook: Making SO2 Gas Measurements at Active Volcanoes, Proceedings of Volcanology, IAVCEI

2.1 Abstract

Recent years have seen the advent of a new generation of small, ultraviolet

spectrometers based on differential optical absorption technology (DOAS) that are widely

applicable in the field of volcanic SO2 monitoring. The new spectrometers, such as the

FLYSPEC and the mini-DOAS, are cheap, small, light, and designed for deployment in

harsh volcanic environments. Computer and GPS integration of the small spectrometers

allow for improvements on previously existing field methods, and introduce the possibility

of new methods, such as automated scanning, accurate determination of plume speed, and

networks of spectrometers. These spectrometers, along with new satellites and other

innovative remote sensing technologies, have greatly improved the ability of volcanologists

to measure SO2 and other volcanic gases. Such advances in volcanic gas monitoring

technology have implications in other fields as well; SO2 fluxes are used in modeling its

effects on climate change, as well as in petrologic studies.

2.2 Introduction

In the years since its 1971 introduction to the volcanological community, the

correlation spectrometer (COSPEC) long maintained its position as the primary tool for the

remote sensing of volcanic SO2 emissions (Stoiber et al., 1983). The technology was first

developed by Barringer Research as a means for monitoring industrial output of SO2 and

NOX to the environment. Volcanologists quickly realized the applicability of such an

instrument to volcanic gas monitoring, and began field campaigns around the world (e.g.,

Stoiber and Jepsen, 1973; Stoiber et al., 1980; Stoiber et al., 1986; Elias et al., 1998).

However innovative the initial application of COSPEC to volcano monitoring, little

was done in the years that followed to improve on the instrumentation itself. It was heavy


16

and cumbersome (~20 kg, 102 cm x 53 cm x 28 cm), and not intended for field work in

rough and variable volcanic terrains. Aside from the instrument itself, a power source was

necessary, as was a paper chart data recorder and some sort of platform for the instrument.

Price was also prohibitive, with a single instrument costing tens of thousands of US dollars

(Galle et al., 2002). In fact, the only significant advance of COSPEC technology over the

years was the optional integration of a digital data interface, making replacement of the paper

chart recorder possible. The size and weight remained essentially the same, while costs

climbed as a result of limited demand, and for years, COSPEC remained the sole

spectrometer utilized to remotely measure volcanic SO2 flux.

Beginning in the late 1990s, miniature, low-cost spectrometers, as well as increasingly

powerful and compact computer technology led to the advent of alternatives to COSPEC.

Based largely on differential optical absorption spectroscopy (DOAS; Platt, 1994), new

spectrometers, such as mini-DOAS (Galle et al., 2002), RMDI (Wardell et al., 2003), MUSE

(Rodríguez et al., 2004), and FLYSPEC (Horton et al., 2006) offered a smaller, cheaper,

more robust alternative to the COSPEC, while also improving several factors in the data

collection methodology and offering the opportunity for a wider range of field applications.

2.3 Instrumentation

2.3.1 Mini-DOAS

First described by Galle et al. (2002), the mini-DOAS utilizes an Ocean Optics

USB2000 spectrometer, comprising a telescope, circular-to-linear optical fiber, a 50-μm slit,

collimating mirror, plane grating, curved mirror, and a 2048-element linear charge-coupled-

device (CCD) array (Figure 2-1). The spectral resolution of the instrument is ~0.6 nm in the

245-380 nm wavelength range, and spectra are output from the instrument to a computer,

which also serves as the power source, via a USB connection. Also connected to the laptop

is a GPS antenna, allowing for position- and time-stamping of data as they are collected.

The entire package is considerably more portable than the COSPEC, weighing less

than 1 kilogram (without laptop computer). Likewise, the dimensions of the mini-DOAS

(~90 mm x ~65 mm x ~35 mm) are considerably smaller than the COSPEC dimensions.

Compared to COSPEC, less power is required to deploy a mini-DOAS: 1W via a laptop


17

Figure 2-1: Configuration of an Ocean Optics USB 2000 mini-DOAS spectrometer, coupled to a telescope via an optical fiber (figure not to scale). Dashed and non-dashed lines drawn within the spectrometer represent the paths of light at two distinct wavelengths, displaying how the spectrometer disperses radiation on to the CCD array.

Reprinted from the Journal of Volcanology and Geothermal Research, Vol. 119, B. Galle, C.M. Oppenheimer, A. Geyer, A. McGonigle, and M. Edmonds, “A mini-DOAS spectrometer applied in

remote sensing of volcanic SO2 emissions,” Pages 241-254, Copyright (2003), with permission from Elsevier.


18

computer connection versus up to 23W (Galle et al., 2002). Costs are also significantly lower

for a mini-DOAS, with the entire configuration, including a laptop computer, often

attainable for approximately 10% the cost of a COSPEC (McGonigle et al., 2003). Beyond

physical specifications, the mini-DOAS’ spectroscopic technique differs from that of the

COSPEC. Instead of utilizing a correlation mask, whereby levels of radiation at wavelengths

specific to the atmospheric absorption windows of SO2 are calibrated by means of

comparison to internal cells of known SO2 concentration, the mini-DOAS records full

radiation spectra covering a wide range of wavelengths. Data collected in the wavelength

ranges characteristic of SO2 are then utilized to determine the amount of SO2 present within

the plume.

Solar radiation spectra are captured by the instrument’s telescope at user-defined

intervals, depending on ambient atmospheric conditions and the data collection method.

Dark spectra are routinely measured during data collection by blocking any radiation from

entering the telescope. Subtraction of dark spectra from sky spectra is required to account

for instrument noise within the data set. Reference spectra of SO2-free sky are also collected

for utilization in the data retrievals.

Spectra are post-processed by means of a number of filtering techniques. Final

spectra are compared and scaled to a laboratory reference spectrum for SO2 in order to

determine the column amount of SO2 present at the time of the measurements. Field trials

utilizing the COSPEC and mini-DOAS in tandem have confirmed comparable results for

the two instruments (Galle et al., 2002; Elias et al., 2006).

2.3.2 FLYSPEC

The FLYSPEC, described in Horton et al. (2006) and Elias et al. (2006), is similar to

the mini-DOAS in the respect that the spectrometer itself is an Ocean Optics USB2000.

For the FLYSPEC, a 25–μm slit results in a spectral resolution of 0.25 nm over a range of

wavelengths from 177-330 nm. Both spectrometers utilize a laptop computer for data

processing and as a power source, and are also significantly smaller, lighter, and cheaper than

the COSPEC. The FLYSPEC itself weighs less than two kilograms, including a small, sub-

notebook computer.


19

Aside from differing spectral specifications, there are a number of physical variations

between the two instruments. The fiber optic cable is omitted from the FLYSPEC; the

telescope is attached directly to the spectrometer in order to reduce light loss. The

FLYSPEC has a larger field of view than the mini-DOAS, at 44 mrad rather than 20 mrad.

Additionally, the FLYSPEC has a UV filter to reduce stray light, and includes SO2 calibration

cells within its configuration. Further differentiating the FLYSPEC from its counterparts is

the combination of all components within a robust, weather-proof case, well-suited for field

deployment (Figure 2-2).

Rather than the mini-DOAS method of collecting spectra and comparing to a

laboratory reference spectrum, the FLYSPEC relies on calibration cells, similar to COSPEC.

Dark spectra and reference spectra are still utilized, and a process similar to the post-

processing of mini-DOAS data takes place, though instead of comparing final spectra to a

laboratory spectrum to determine SO2 concentration, the final spectra from the FLYSPEC

are calibrated to the reference spectra produced by the insertion of calibration cells of

known concentration into the instrument’s field of view. Calibrations are performed in the

field, in the same atmospheric conditions as the collection of the larger, primary dataset. By

using calibration-cell reference spectra obtained in the field, atmospheric effects such as

Fraunhofer lines or Ring-effect Raman spectra (which would not be present in model

laboratory spectra) will be present in all spectra, negating the need for data corrections

(Horton et al., 2006). Like the mini-DOAS, the FLYSPEC also records and stores full

spectra, allowing for re-processing and comparison to other spectra following the DOAS

methodology. FLYSPEC results in field trials are also comparable to those of COSPEC and

mini-DOAS (Elias et al., 2006; Horton et al., 2006).

Despite differences between the two newest tools in the field of remote SO2

monitoring at volcanoes, both provide a viable alternative to the COSPEC. The compact

size, relatively low cost, DOAS techniques, and computer/GPS integration of both the

FLYSPEC and the mini-DOAS have allowed for an increased range of possible applications

of the instruments in the field of monitoring volcanic SO2.


20

Figure 2-2: Components of a FLYSPEC: an Ocean Optics USB 2000 spectrometer, sub-notebook computer, GPS with antenna, high and low calibration SO2 gas cells mounted above the spectrometer, and telescope. The “telescope” is a fiber-optic collimating lens mounted directly to the spectrometer input aperture. The lens, in combination with the UV band-pass filter window mounted on the case, provides a field of view of approximately 2.5°. Power for the spectrometer and GPS is supplied by the computer.

Reprinted from the Bulleting of Volcanology, Vol. 68, K.A. Horton, G. Williams-Jones, H. Garbeil, T. Elias, A.J. Sutton P. Mouginis-Mark, J.N. Porter, and S. Clegg, “Real-time

measurement of volcanic SO2 emissions: Validation of a new UV correlation spectrometer (FLYSPEC),” Figure 1, page 324, Copyright (2006), with kind permission of Springer Science

and Business Media.


21

2.4 Field Methodology

2.4.1 Traditional methods

As with traditional COSPEC methodology (Stoiber et al., 1983), the FLYSPEC and

the mini-DOAS are well-suited for both stationary scans and vehicle-based traverses beneath

plumes. With the mini-DOAS, the optical fiber linking the spectrometer and the telescope

also provides increased ease of use in vehicular traverses, as only the telescope requires

external mounting, allowing the spectrometer itself to be manipulated as needed within the

car, plane, helicopter, or boat. The FLYSPEC, while lacking the optical fiber and separate

telescope, is still very convenient in traverses, as the entire instrument is small enough to be

easily mounted on the exterior of the vehicle (Figure 2-3), whereas COSPEC required

significant space within the vehicle.

2.4.2 Walking traverses

At many volcanoes worldwide, automobile traverses using COSPEC, FLYSPEC, or

mini-DOAS are impossible, as a consequence of remote locations, insufficient infrastructure,

or a combination of both. COSPEC had been employed sporadically in walking traverses in

situations where plume geometries prohibited scans or vehicular traverses (Stoiber et al.,

1983); however, the weight and size of the instrument, as well as the requisite constant

traverse speed prior to GPS integration kept the walking traverse methodology from

becoming commonplace. The newer, smaller, GPS-integrated FLYSPEC and mini-DOAS

facilitate such walking traverses, as demonstrated at both Kilauea (Horton et al., 2006) and

Masaya volcanoes (McGonigle et al., 2002). The small spectrometers are easily mounted in a

carrier on the operator’s back (Figure 2-4), or, in the case of mini-DOAS, the telescope may

be affixed to a helmet and the body of the spectrometer simply carried by the operator

(McGonigle et al., 2002).

2.4.3 Automated scanning

While spectrometers like mini-DOAS and FLYSPEC lend themselves to manual

scanning of plumes as with COSPEC methodology, their small size and high degree of

computer integration make possible automated scans of plumes. The addition of a rotating

motor to an individual spectrometer’s configuration is relatively simple; a small mirror or


22

Figure 2-3: Example of a FLYSPEC field deployment. The spectrometer unit is mounted near the car’s side mirror with duct tape, while the GPS antenna is affixed to the roof.


23

prism attached to the motor can be programmed to scan across the breadth of a plume at

any desired speed or resolution (Edmonds et al., 2003).

Such automated scans, with scan duration and angular range more easily constrained

than with manual scans, may be especially useful at volcanoes with unusual crater geometry

or awkward plume configurations, which do not allow for traverses beneath the plume.

However, as with COSPEC scans, problems and uncertainties in determining the height of

and distance to the plume persist.

2.4.4 Automated networks

At volcanoes that are highly active, constant or near-constant monitoring of the

system is desirable. While instruments such as tiltmeters and seismometers produce high-

resolution, real- or near-real-time data and are easily deployed remotely, the COSPEC

requires the presence of an operator. COSPEC data are collected only in discrete scans or

traverses, and these data also require post-processing (Edmonds et al., 2003). Newer, small

spectrometers, with their computer integration, real-time data output, and ability to make

measurements through telemetry, provide an opportunity to produce high-resolution,

continuous SO2 flux data.

One means by which near-constant measurements can be collected is to modify

spectrometers, such that they perform regular scans of a volcano’s plume. Automated

scanning mini-DOAS networks have been employed already at Mt. Etna, Italy (McGonigle

et al., 2003; Salerno et al., 2004), Stromboli, Italy (Salerno et al., 2004), and Soufrière Hills

volcano, Montserrat (Edmonds et al., 2003; Young et al., 2003). At Soufrière Hills, two

mini-DOAS spectrometers (Scanspecs) were placed on the downwind slopes of the volcano.

Each of the instruments was outfitted with a stepper motor to rotate a prism, which reflects

light into the spectrometer’s telescope (Figure 2-5), allowing for the spectrometers to

repeatedly scan the SO2 plume eight hours per day. As a consequence of the angular scans,

data are in the form of SO2 slant columns rather than the vertical columns obtained by

automobile or walking traverses. Accordingly, SO2 flux calculations for the Scanspecs’ slant

columns require knowledge of plume height, in addition to other variables such as plume

speed. The use of multiple, time-synchronized spectrometers allows for determination of

plume height; each instrument scans through the plume and, based on the geometry of the


24

Figure 2-4: A walking traverse with a backpack-mounted FLYSPEC at Vulcano, Italy


25

Figure 2-5: An example of a scanning DOAS-based spectrometer. (modified from Edmonds et al., 2003)

Schematic portion of figure reprinted from the Bulletin of Volcanology, Vol. 65, M. Edmonds, R.A. Herd, B. Galle, and C.M. Oppenheimer, “Automated, high time-resolution measurements of

SO2 flux at Soufrière Hills Volcano, Montserrat,” Figure 2, page 580, Copyright (2003), with kind permission of Springer Science and Business Media.


26

spectrometers’ location and angular data at the time of peak SO2 concentration, the height of

the plume’s maximum concentration is derived (Figure 2-6). Distances from the

spectrometers to the plume are then calculated. These distances, in combination with the

angular range over which SO2 was detected in the scans, are subsequently used to determine

the plume’s true width. Once the plume’s width is known, total flux of SO2 can be

calculated as with traditional traverses (Edmonds et al., 2003).

While this type of automated scanning had proven to be successful at Mt. Etna,

Stromboli, and Soufrière Hills, it may not be suitable in other types of volcanic terrain. At

Kilauea or Masaya, for example, the generally low topography results in plumes that are quite

often very close to ground level. In such cases, the plume’s geometry may not permit full

scans of the plume; some segments of the plume may appear to be below the horizon,

rendering the data acquisition impossible.

2.4.4.1 Potential enhancements to data acquisition networks

At volcanoes where plume and topography geometries are not conducive to

scanning, but automated, real-time, semi-permanent SO2 monitoring is desired, other

possibilities may exist. The FLYSPEC and mini-DOAS instruments, at present, possess

fields of view of 44 and 20 mrads, respectively, which limit the instruments to ‘seeing’ only

small segments of plumes at any given time. If the optical configurations of the instruments

could be altered, through the use of a ‘fish-eye’ lens, for example, so as to broaden the field

of view significantly, a small number of stationary instruments aligned perpendicular to the

plume’s propagation could ‘see’ the entire width of the plume (Figure 2-7). An ideal network

would fully encircle the volcano, though practicality would likely allow for a smaller number

of spectrometers arranged downwind of the volcano according to the predominant wind

direction. Each spectrometer would record an average path length concentration for the

whole of its respective sector of the plume’s cross-section. The multiple spectrometers

would yield continuous data across the width of the plume, allowing for real-time flux

measurements with a temporal resolution even greater than that of automated scanning

systems.

Scan-based data are limited by the finite period of time necessary to complete a full

scan of the plume (80-400 s; Edmonds et al., 2003); stationary wide-angle spectrometer


27

Figure 2-6: Calculating plume height and position from the SO2 concentration and angular data from two fixed scanning spectrometers. Angles α1 and α2 are determined from the scan angles coinciding with peaks in SO2 concentration during one revolution. Tan α1=h/d1 and tan α2=(h+z)/d2, where h is plume height and z is any elevation difference between the locations of the two spectrometers. The known horizontal distance between the spectrometers is equal to d1+d2 (modified from Edmonds et al., 2003)

Reprinted from the Bulletin of Volcanology, Vol. 65, M. Edmonds, R.A. Herd, B. Galle, and C.M. Oppenheimer, “Automated, high time-resolution measurements of SO2 flux at Soufrière Hills Volcano, Montserrat,” Figure 6, page 583, Copyright (2003), with kind permission of

Springer Science and Business Media.


28

Figure 2-7: Schematic diagram of a degassing volcano, plume with cross-section, and a possible configuration for proposed network of stationary wide-angle miniature spectrometers, with their expanded fields of view delineated by dotted lines.


29

networks could produce flux data at the same temporal resolution of spectral acquisitions,

usually 300-1000 ms (Horton et al., 2006). Further, data from the stationary network would

represent integrated instantaneous cross-sections and, therefore, plume fluxes as well,

whereas data from scans and traditional traverses, during which the plume is continuously

propagating overhead, correspond to only average fluxes over the finite time period required

to complete each measurement.

Alternatively, several stationary spectrometers with a traditional small field of view

may be sufficient to continuously monitor volcanic plumes. Arranged in a similar fashion to

the proposed ‘fish-eye lens’ network, the instruments would not measure the entire breadth

of the plume, but may provide enough data to closely approximate the plume’s true cross-

section (Figure 2-8). As such, the limited spatial resolution of such a network would be

compensated by the high temporal resolution. In each case, the optimum spatial density of

spectrometers across a given plume’s transect would need to be determined for each volcano

for typical weather conditions. Extensive field testing would also be required, for both the

wide-angle and the standard field of view stationary methods, in order to determine the

accuracy of each technique relative to the established standards of scanning and traversing

the plume. If found to systematically over- or under-estimate SO2 fluxes with respect to

other measurement techniques, initial field comparisons could determine a scaling factor for

use in normalizing future stationary measurements.

2.4.5 Plume speed / velocity

One of the crucial uncertainties associated with the determination of SO2 flux,

whether through the use of COSPEC, FLYSPEC, or mini-DOAS, is the velocity at which

the measured plume propagates past the spectrometer’s field of view. The formula for

calculation of SO2 flux is as follows:

F = [SO2] · cosθ · d · v · c [2-1]

where [SO2] is the concentration path length [ppm·m] of SO2 in the gas column, θ [°] is the

deviation from perpendicular of the road to the plume’s path of propagation, d [m] is the

apparent width of the plume as it appears during the traverse, v [m/s] is the plume velocity,

and c is a unit conversion factor (final SO2 flux values are often reported in metric

tonnes/day, t/d). Clearly, the net flux is directly proportional to plume velocity; therefore,


30

Figure 2-8: Example of a network of stationary network of spectrometers with conventional fields of view, represented by dashed lines. Data from collection points closely approximate the plume shape and average concentration pathlength of a traditional traverse. An example traverse is represented by a dark curve, and the plume approximated from the four spectrometers is marked by the gray-striped area.


31

an accurate appraisal of plume velocity is needed in order to best determine the true SO2

flux. Generally, researchers have had to rely on wind speed measurements as a proxy for

this plume velocity. Ground-based anemometer data (e.g., Elias et al., 1998; Williams-Jones

et al., 2003), radiosonde information from local airports, and visual assessments of speed

from video recordings of plume have all been utilized in place of the plume’s velocity in such

calculations (Doukas, 2002; Williams-Jones et al., 2006). In-plume wind speed

measurements have also been conducted, and yield relatively accurate results (Doukas, 2002).

However, logistical difficulties, such as volcanoes with low-lying topography and plumes,

and the prohibitive costs associated with aircraft-based measurements, limit the widespread

application of the method.

With the introduction of these small spectrometers, new techniques have been

developed to determine a more accurate, cost-effective representation of volcanic plume

velocity. The low cost of FLYSPEC and mini-DOAS allow for the procurement of multiple

spectrometers (Edmonds et al., 2003; Horton et al., 2006), and with only two instruments it

is possible to derive the apparent speed of a plume. To do so, the location of the most

concentrated part of the plume is located by completing a preliminary traverse beneath the

plume. If local geography and infrastructure beneath the concentrated segment permits,

each of the spectrometers is placed on a tripod and set to collect SO2 path-length

concentrations. By placing the time-synchronized instruments along a line parallel to the

plume’s propagation (Figure 2-9), similar signals are recorded on each of the instruments.

The combination of the known distance between the spectrometers (from the integrated

GPS units or a tape measure) and the time lag between corresponding points on the data

plots allows for the calculation of plume speed by comparing the data signals recorded by

each spectrometer, and noting the time lag between them (Figure 2-10; Williams-Jones et al.,

2006).

An ideal spatial separation for the two spectrometers in this technique is typically 20

– 50 m (Williams-Jones et al., 2006), with some exceptions. In instances when a plume is

moving very quickly, the time separation between the datasets of two spectrometers placed

close together may approach the temporal resolution of the data. Variations in the plume’s

concentration cannot be adequately resolved for proper correlation of the two

spectrometers’ data signals under such circumstances, and the two instruments must be


32

moved farther away from each other. However, the possibility of greater error as a result of

dispersion over the greater distance between the two spectrometers must be considered.

Williams-Jones et al. (2006) report an uncertainty of 0.28-2.73% due to inline dispersion over

10-50 m for a plume 500 m high as measured 5 km from the degassing vent; uncertainty

would increase with greater spectrometer separation.

Problems also arise in situations when the plume is moving very slowly. Too large a

spatial spectrometer separation will allow for mixing and other alterations to the plume’s

physical structure such that the signal recorded by the downwind spectrometer is dissimilar

to the initial signal recorded further upwind. However, compensation for this is not possible

by simply moving the spectrometers closer together, as the spectrometers’ fields of view will

overlap, especially for high plumes. Therefore, the dual-spectrometer plume speed method

will not give accurate results for very slow plumes (~<2 m/s) and should not be used for

such circumstances.

There may be errors associated with plume speed measurements as a result of

variations in plume direction over the course of a plume speed measurement, and

accordingly, the spectrometers not being aligned parallel to the center of the plume. To

compensate, the full half-hour dataset can be subdivided into smaller time windows, and

only those with the best correlations (i.e., when the spectrometers were aligned with the

plume, and, therefore, measuring the same parcels of gas) used in assessing the plume speed.

Time windows are also useful in accounting for variations in the plume speed over the half-

hour. As well, as an alternative to the determination of plume speed, additional

spectrometers may be added to the two-spectrometer configuration in order to determine

plume velocity (McGonigle et al., 2005a; Williams-Jones et al., 2006) and changes in the

plume’s vector over the course of a measurement.

Beyond a simple multi-spectrometer approach to determining plume speed, a

rotating mirror may be integrated into a single spectrometer’s field housing, allowing for

scanning of two sectors of the sky during a single traverse. Doing such allows for

determination of plume height and, subsequently, plume transport speed (McGonigle et al.,

2005b). Such a method does become complicated, however, in locales lacking a flat road

network on which to traverse, as is the case at many volcanoes. In deriving plume height,

the pitch of the instrument at the time of each spectral acquisition must be known.


33

Figure 2-9: Field deployment for determination of plume speed at Masaya volcano, Nicaragua. The FLYSPECs were mounted downwind of the gas source on small, lightweight camera tripods separated by 20 m. Measurements were made for 30 minutes, with a GPS antenna in each configuration to ensure time-synchronization of the datasets.


34

Figure 2-10: (a) The SO2 pathlength concentration for two FLYSPECs at Masaya volcano, Nicaragua on March 25, 2003. The instrument separation is 40.5 m, determined by tape measure. Inset is a 4-minute window showing an apparent time separation (16 s) between the 2 signals. (b) The SO2 signals for the entire 30-minute sampling period are compared to each other for timeshifts between -60 and 60 s at 0.1 s iterations. The maximum correlation coefficient (r2 = 0.959) for the signals occurs at a time difference of 13.1 s, which for a 40.5 m separation, results in a plume speed of 3.1 m/s.

Reprinted from the Bulletin of Volcanology, Vol. 68, G. Williams-Jones, K.A. Horton, T. Elias, H.

Garbeil, P.J. Mouginis-Mark, A.J. Sutton, and A.J.L. Harris, “Accurately measuring volcanic plume velocity with multiple UV spectrometers,” Figure 2, page 330, Copyright (2006), with kind

permission of Springer Science and Business Media.


35

Corrections for pitch compensate for undulating topography and resultant deviations of the

instrument’s field of view from the intended azimuth and inclination, but such continuous

measurements of pitch may not always be practical or feasible.

Plume speed would ideally be measured continuously for the duration of a given

day’s SO2 flux measurement collection. Constant recording by two spectrometers would

identify changes in the speed of the plume over the course of the day, allowing for accurate

plume speeds to be determined for each measurement of SO2 flux. However, such

continuous measurements are generally neither practical, as in situations where the plume

shifts laterally many times over the course of a day, nor feasible, based on the number of

spectrometers necessary for flux measurements and simultaneous plume speed

measurements. Short of continuous monitoring of plume speed, the number of plume

speed measurements made should be as large as possible without compromising the quality

of other data sets being collected.

2.5 Other Gas Species

The original COSPEC was developed as a means to monitor industrial emissions of

both SO2 and NOX. This proved to be of limited use in the volcanic sector, as simultaneous

measurements of multiple gases with the COSPEC is impossible as a result of the correlation

mask spectroscopy methodology. Data concerning nitrogen species within volcanic plumes

are of minimal value relative to information on SO2 fluxes and, therefore, SO2 data collection

took priority during COSPEC field campaigns. Further, in instances when NOX data may

have been called for, high costs would have prohibited the use of two spectrometers in

tandem during traverses or scans, and limited time would probably have restricted scientists

from performing separate scans or traverses to focus on NOX.

The advent of DOAS as a means to study volcanic plumes has introduced the

possibility of obtaining, with relative ease, data on gas species other than SO2. DOAS

technology differs from COSPEC in its ability to record full spectra, thereby obtaining

information for wavelengths of ultraviolet radiation in the atmospheric absorption windows

of gases other than SO2. Given calibration cells or laboratory spectra for other species

within the range of wavelengths of the instrument, fluxes of a wide variety of volcanic gases

can theoretically be derived.


36

Attempts have recently been made to quantify BrO at Soufrière Hills volcano

(Bobrowski et al., 2003), NO2 at a power plant in the U.K. (McGonigle et al., 2004), and H2S

at Solfatara and Vulcano volcanoes in Italy (O’Dwyer et al., 2003) through the use of mini-

DOAS spectrometers and laboratory reference spectra. Similar studies could also

theoretically be conducted using calibration cells of different gas species.

2.6 Additional Remote Gas Sensing Techniques

DOAS technology is most commonly used in the form of ground-based or airborne

spectrometers, as with FLYSPEC and mini-DOAS instruments, but recent satellite

technology has expanded the applicability of DOAS to volcanic SO2 sensing. The Global

Ozone Monitoring Experiment (GOME) sensor aboard the second European Remote

Sensing satellite, while primarily targeting ozone, possesses the capability to detect

atmospheric absorption by other trace gases. The same UV absorption signal used to

determine quantities of SO2 from ground-based spectrometers can also be used to derive

SO2 concentrations from GOME data (Bortoli et al., 2000). Other UV satellite sensors, such

as the Total Ozone Mapping Spectrometer (TOMS), also detect volcanic SO2. However,

while TOMS has provided a reliable record of SO2 output from large eruptions over the past

two decades, the technology is limited. TOMS is generally capable of detecting only

stratospheric SO2; smaller eruptions or quiescent degassing that reach only into the

troposphere go largely undetected (Carn et al., 2003). A newer tool, the Ozone Monitoring

Instrument (OMI), part of the payload on NASA’s EOS/Aura satellite, is based on

technology similar to that of TOMS. Compared to TOMS, OMI offers a greater degree of

spatial and spectral resolution, as well as lower detection limits for SO2, which translates to

the ability to detect smaller eruption clouds than TOMS, and even passive degassing (Carn,

2006).

More recent developments in satellite sensing of volcanic SO2 have been in the

infrared portion of the electromagnetic spectrum. AVHRR, MODIS, ASTER, SEVIRI, and

VISSR are all instruments with bands capable of detecting, through manipulation of infrared

data, elevated levels of SO2 into the upper troposphere (Tupper et al., 2004). Most recently,

Carn et al. (2005) have demonstrated the capacity of AIRS to sense volcanic SO2 from

moderate eruptions at altitudes as low as ~6 km. The high spectral resolution of AIRS also


37

may permit detection of volcanic CO2, but detection of any passively degassed plumes

remains problematic as a result of the low spatial resolution (13.5 km/pixel) of the

instrument (Carn et al., 2005).

LIDAR, a laser-based technology, and its derivative Differential Absorption LIDAR

(DIAL) provide another ground-based option in the remote observation of volcanic gases.

Laser pulses are directed at the plume, and based on the return time for back-scattered

radiation of specific wavelengths, concentrations of various gases can be determined

(Weibring et al., 2002; McGonigle and Oppenheimer, 2003). Plume speed may also be

integrated with LIDAR data to determine fluxes of SO2.

Fourier Transform Infrared spectroscopy (FTIR) has also been employed as an

alternative method for monitoring volcanic gases. FTIR yields data in the form of a wide

variety of molar ratios, such as SO2/HCl and HCl/HF, and, in combination with SO2 flux

data from a COSPEC, FLYSPEC, or mini-DOAS, the data may be used to determine fluxes

of trace gases from volcanoes, as at Masaya volcano (Horrocks et al., 2003). While FTIR

equipment is quite bulky and difficult to deploy in most field conditions, and may eventually

be surpassed by instruments utilizing DOAS technology in the sensing of gases other than

SO2, it is still a key means by which volcanic H2O and CO2 may be detected. Because of the

high levels of H2O and CO2 in the ambient atmosphere, their volcanic components are often

quite difficult to detect; the ability of FTIR technology to do so makes possible a fuller

quantification of the extent of volcanic degassing.

As with FTIR, detection of H2O and CO2 is now also possible with a new

development in volcanic gas monitoring: a multi-sensor system (MSS) comprising a

temperature/humidity sensor, electrochemical sensors, and an infrared analyzer. The

instrument measures H2O, CO2, and SO2, and the addition of specialized sensors and

alkaline filter techniques would expand the range of measurable compounds to H2S, H2, Cl,

and F (Shinohara, 2005). Despite the number of components, the system is portable and

weighs only 5 kg. This portability is necessary, as the MSS verges on being a type of direct

sampling rather than remote sensing in the manner of instruments like COSPEC or LIDAR.

Measurements with the MSS must be taken either very near, or in, the degassing crater, or

within the plume itself, as from a helicopter or airplane.


38

2.7 Aerosols

Regardless of the method utilized in the monitoring of volcanic SO2, one caveat is

the possibility of conversion of SO2 to sulfate aerosols:

OH + SO2 (+O2, H2O) H2SO4 + HO2 [2-2]

Should such conversion occur downwind of the degassing volcano, fluxes of SO2 measured

by spectrometers may underestimate the true, original flux from the volcano’s vent. To

mitigate such underestimates, some SO2 monitoring campaigns have begun to integrate the

use of a sun photometer to measure aerosol quantity along with SO2 flux (Porter et al., 2002;

Mather et al., 2004; Chapter 3; Appendix A). Satellites such as TOMS and AIRS also

possess the capability to sense atmospheric aerosols in tandem with volcanic SO2

measurements (Carn et al., 2005).

2.8 Conclusions

The COSPEC proved invaluable to volcanic SO2 research for nearly three decades.

In recent years, the technology has been surpassed by smaller, cheaper, more versatile

instruments based on differential optical absorption spectroscopy (DOAS) techniques. Both

mini-DOAS and FLYSPEC are capable of SO2 sensing in the manner of traditional

COSPEC methods, and also have the potential for a wide range of other applications in the

field of volcano monitoring. With this new generation of spectrometer, gas monitoring will

be able to keep pace with the likes of other real-time monitoring methods, as have been

employed in deformation and seismology studies. Both single instrument scanning and

stationary networks of these small spectrometers can now be implemented alongside

seismometers and GPS stations on volcanoes worldwide, finally presenting an opportunity

for continuous geochemical monitoring. Further, gas monitoring is no longer limited to only

SO2; the capability of DOAS methodology to detect a variety of other volcanic gas species is

a significant improvement over COSPEC, and will likely play an increasingly important role

in the future as the range of detectable gases broadens.


39

2.9 References

Bobrowski, N., Hönninger, G., Galle, B., and Platt, U., 2003. Detection of bromine monoxide in a volcanic plume. Nature, 423: 273-276.

Bortoli, D., Costa, M.J., Ravegnani, F., Kostadinov, I., Giovanelli, G., and Petritoli, A., 2000. Atmospheric SO2 vertical column retrieval from GOME data analysis in the Mediterranean area. 4th ERS-ENVISAT Symposium, Gothenburg, Sweden: October 6-20, 2000.

Carn, S.A., Krueger, A.J., Bluth, G.J.S., Schaefer, S.J., Krotkov, N.A., Watson, I.M., and Datta, S., 2003. Volcanic eruption detection by the Total Ozone Mapping Spectrometer (TOMS) instruments; a 22-year record of sulphur dioxide and ash emissions. In: Oppenheimer, C., Pyle, D.M., and Barclay, J. (Editors), Volcanic Degassing, Geological Society, London, Special Publications, 213:177-202.

Carn, S.A., Strow, L.L., de Souza-Machado, S., Edmonds, Y., and Hannon, S., 2005. Quantifying tropospheric volcanic emissions with AIRS; the 2002 eruption of Mt. Etna (Italy). Geophysical Research Letters, 32(2), L02301, doi:10.1029/2004GL021034.

Carn, S.A., 2006. TOMS Volcanic Emissions Group website. Ozone Monitoring Instrument (OMI) - Total Ozone Mapping Spectrometer (TOMS). Retrieved July 6, 2006 from http://toms.umbc.edu/

Doukas, M.P., 2002. A new method from GPS-based wind speed determinations during airborne volcanic plume measurements. U.S. Geological Survey Open-File Report 02-395, pp 1-13.

Edmonds, M., Herd, R.A., Galle, B., and Oppenheimer, C.M., 2003. Automated, high time-resolution measurements of SO2 flux at Soufrière Hills Volcano, Montserrat. Bulletin of Volcanology, 65(8): 578-586.

Elias, T., Sutton, A.J., Stokes, J.B., Casadevall, T.J., 1998. Sulfur dioxide emission rates of Kilauea Volcano, Hawaii, 1979-1997. U.S. Geological Survey Open-File Report 98-462.

Elias, T., Sutton, A.J., Oppenheimer, C., Horton, K.A., Garbeil, H., Tsanev, V., McGonigle, A.J.S., and Williams-Jones, G., 2006. Intercomparison of COSPEC and two miniature ultraviolet spectrometer systems for SO2 measurements using scattered sunlight. Bulletin of Volcanology, 68: 313-322.

Galle, B., Oppenheimer, C., Geyer, A., McGonigle, A.J.S., Edmonds, M., and Horrocks, L., 2002. A miniaturised ultraviolet spectrometer for remote sensing of SO2 fluxes; a new tool for volcano surveillance. Journal of Volcanology and Geothermal Research, 119(1-4): 241-254.


40

Horrocks, L.A., Oppenheimer, C., Burton, M.R., and Duffel, H.J., 2003. Compositional variation in tropospheric volcanic gas plumes; evidence from ground-based remote sensing. In: Oppenheimer, C., Pyle, D.M., and Barclay, J. (Editors), Volcanic Degassing, Geological Society, London, Special Publications, 213: 349-369.


Mather, T.A., Tsanev, V.I., Pyle, D.M., McGonigle, A.J.S., Oppenehimer, C., and Allen, A.G., 2004. Characterization and evolution of tropospheric plumes from Lascar and Villarrica volcanoes, Chile. Journal of Geophysical Research, 109, D21303, doi:10.1029/2004JD004934.


McGonigle, A.J.S., and Oppenheimer, C., 2003. Optical sensing of volcanic gas and aerosol emissions. In: Oppenheimer, C., Pyle, D.M., and Barclay, J. (Editors), Volcanic Degassing, Geological Society, London, Special Publications, 213: 149-168.

McGonigle, A.J.S., Oppenheimer, C., Hayes, A.R., Galle, B., Edmonds, M., Caltabiano, T., Salerno, G., Burton, M., and Mather, T.A., 2003. Sulphur dioxide fluxes from Mount Etna, Vulcano, and Stromboli measured with an automated scanning ultraviolet spectrometer. Journal of Geophysical Research, 108(B9), 2455, doi:10.1029/2002JB002261.

McGonigle, A.J.S., Thomson, C.L., Tsanev, V.I., and Oppenheimer, C., 2004. A simple technique for measuring power station SO2 and NO2 emissions. Atmospheric Environment, 38(1): 21-25.

McGonigle, A.J.S., Hilton, D.R., Fischer, T.P., and Oppenheimer, C., 2005a. Plume velocity determination for volcanic SO2 flux measurements. Geophysical Research Letters, 32, L11302, doi:10.1029/2005GL022470.

McGonigle, A.J.S., Inguaggiato, S., Aiuppa, A., Hayes, A.R., and Oppenheimer, C., 2005b. Accurate measurement of volcanic SO2 flux: Determination of plume transport speed and integrated SO2 concentration with a single device. Geochemistry Geophysics Geosystems, 6(1), Q02003, doi:10.1029/2004GC000845.

O'Dwyer, M., Padgett, M.J., McGonigle, A.J.S., Oppenheimer, C., and Inguaggiato, S., 2003. Real-time measurement of volcanic H2S and SO2 concentrations by UV spectroscopy. Geophysical Research Letters, 30(12), 1652, doi:10.1029/2003GL017246.

Platt, U., 1994. Differential Optical Absorption Spectroscopy (DOAS). In: Sigrist, M.W. (Editor), Air Monitoring by Spectroscopic Techniques, Chemical Analysis Series, 127. Wiley, New York: 27-84.


41


Rodríguez, L.A., Watson, I.M., Rose, W.I., Branan, Y.K., Bluth, G.J.S., Chigna, G., Matías, O., Escobar, D., Carn, S.A., and Fischer, T.P., 2004. SO2 emissions to the atmosphere from active volcanoes in Guatemala and El Salvador, 1999-2002. Journal of Volcanology and Geothermal Research, 138: 325-344.

Salerno, G., Burton, M., Caltabiano, T., Mure, F., Bruno, N., Longo, V., and Condarelli, D., 2004. First results from automated SO2 flux monitoring networks on Mt. Etna and Stromboli volcano, Italy. IAVCEI 2004 General Assembly, Pucón, Chile: November 14-19, 2004.

Shinohara, H., 2005. A new technique to estimate volcanic gas composition: plume measurements with a portable multi-sensor system. Journal of Volcanology and Geothermal Research, 143(4): 319-333.

Stoiber, R.E., and Jepsen, A., 1973. Sulfur Dioxide Contributions to the Atmosphere by Volcanoes. Science, 182: 577-578.

Stoiber, R.E., Williams, S.N., and Malinconico, L.L., 1980. Mount St. Helens, Washington, 1980 volcanic eruption; magmatic gas component during the first 16 days. Science, 208(4449): 1258-1259.

Stoiber, R.E., Malinconico, L.L., and Williams, S.N., 1983. Use of the correlation spectrometer at volcanoes. In: Tazieff, H. and Sabroux, J.C. (Editors), Forecasting Volcanic Events. Elsevier, New York: 424-444.


Tupper, A., Carn, S., Davey, J., Kamada, Y., Potts, R., Prata, F., and Tokuno, M., 2004. An evaluation of volcanic cloud detection techniques during recent significant eruptions in the western 'Ring of Fire'. Remote Sensing of the Environment, 91(1): 27-46.

Wardell, L.J., Morrow, B., Stix, J., 2003. Initial field trials for development of RMDI, a miniaturized UV/visible spectrometer designed for multi-gas remote sensing of volcanic plumes. EOS Transactions of the American Geophysical Union, 84(46), Fall Meeting Supplement, Abstract V11C-0513.

Weibring, P., Swartling, J., Edner, H., Svanberg, S., Caltabiano, T., Condarelli, D., Cecchi, G., and Pantani, L., 2002. Optical monitoring of volcanic sulphur dioxide emissions--comparison between four different remote-sensing spectroscopic techniques. Optics and Lasers in Engineering, 37(2-3): 267-284.

Williams-Jones, G., Rymer, H., and Rothery, D.A., 2003. Gravity changes and passive SO2 degassing at the Masaya caldera complex, Nicaragua. Journal of Volcanology and Geothermal Research, 123(1-2): 137-160.


42


Young, S.R., Voight, B., and Duffell, H.J., 2003. Magma extrusion dynamics revealed by high-frequency gas monitoring at Soufriere Hills Volcano, Montserrat. In: Oppenheimer, C., Pyle, D.M., and Barclay, J. (Editors), Volcanic Degassing, Geological Society, London, Special Publications, 213: 219-230.

Chapter 3 – SO2 Flux and Apparent Downwind Depletion

43

Chapter 3: SO2 FLUX MEASUREMENTS AT MASAYA VOLCANO, NICARAGUA: APPARENT DOWNWIND DEPLETION

AS A RESULT OF TOPOGRAPHIC MODIFICATION OF

WINDS

*To be submitted to the Bulletin of Volcanology

3.1 Abstract

A series of 701 SO2 measurements at Masaya in 2005 and 2006 reveals that SO2

fluxes 15 km downwind of the active vent are 30-50% less than those measured only 5 km

from the vent. Measurements from this and previous studies indicate that dry deposition of

sulfur from the plume and conversion of SO2 to sulfate aerosols within the plume each may

amount to a maximum of 10% loss, and are not sufficient to account for the larger apparent

loss measured. However, the SO2 flux measurement site 15 km downwind is located on a

ridge over which local trade winds, and the entrained plume, accelerate. It is interpreted that

the greater wind speeds cause localized dilution of the plume along the axis of propagation.

Lower concentrations of SO2 measured on the ridge lead to calculations of lower fluxes, as

calculated at the same plume speed as measurements from only 5 km downwind, and is

responsible for the apparent loss of SO2. Future campaigns to measure SO2 flux at Masaya

will require individual plume speed measurements to be taken at each flux measurement site

to compensate for dilution and subsequent calculation of lower fluxes.

3.2 Introduction

Volcanic gas emissions play an important role in the understanding of volcanic

processes and the forecasting of volcanic eruptions (e.g., Stoiber et al., 1983; Zapata et al.,

1997; Williams-Jones et al., 2003). Sulfur dioxide is the gas most commonly monitored, due

to (1) the relative abundance of SO2 in emissions in comparison to background

concentrations, and (2) the fact that SO2 also absorbs radiation in the ultraviolet range,

making it easily detectable through solar remote sensing techniques (Stoiber et al., 1983).


44

Campaigns to measure SO2 emissions using ground-based UV-spectrometers have been

undertaken at volcanoes worldwide since the 1970s (e.g., Stoiber and Jepsen, 1973; Stoiber et

al., 1983; Elias et al., 1998; Rodríguez et al., 2004). Recent advances in spectroscopic

technology have led to the development of small, low cost spectrometers that allow SO2

monitoring campaigns to be conducted with greater ease at an increasing number of

volcanoes worldwide (Galle et al., 2002; Horton et al., 2006; Chapter 2).

In making SO2 measurements, it is important to note the reactivity of the species and

the potential for loss of SO2 from the plume that could lead to underestimates of the actual

flux. SO2 may convert to sulfate aerosols in the atmosphere, be adsorbed onto ash particles

within an ash-laden plume (Witham et al., 2005, and references therein), and scavenged from

the airborne plume by means of wet or dry deposition (Delmelle et al., 2001; 2002). Recent

studies have reported a range of SO2-sulfate conversion rates from those that are negligible

for the typical timescales of SO2 measurements (McGonigle et al., 2004) to near 100%

depletion of a volcanic plume’s SO2 content (Oppenheimer et al., 1998). These studies

comprised small data sets, measured only SO2 rather than SO2 and aerosols, and were based

on calculations of plume age derived from wind speed measurements that may not have

been representative of the true plume speed. In light of the prevalence of inaccurate

estimations of plume speed, it was suggested that all reported fluxes of SO2 be normalized to

a common plume speed of 1 m/s (Zapata et al., 1997). More recently, methods to better

constrain plume speed have been developed, increasing the accuracy and precision of SO2

flux measurements (McGonigle et al., 2005a, b; Williams-Jones et al., 2006a, b). Further, the

use of sun photometers as a means to measure volcanic aerosols has grown more

commonplace (e.g., Watson and Oppenheimer, 2000, 2001; Porter et al., 2002; Mather et al.,

2004). In light of these new developments, it was felt that the issue of SO2 loss in volcanic

plumes should be revisited.

Masaya volcano, ~20 km southeast of Managua in western Nicaragua, provides an

ideal natural laboratory to study potential SO2 loss. Masaya, a basaltic complex consisting of

nested calderas, cones, and pit craters (McBirney, 1956; Rymer et al., 1998; Girard and van

Wyk de Vries, 2005), though persistently active, rarely experiences eruptive activity. Rather,

the volcano has undergone a series of episodes of quiescent degassing for over 150 years,

with intermittent gas crises associated with periods of elevated SO2 output (Stoiber et al.,


45

1986 and references therein). 1431 individual measurements of SO2 flux reveal that over the

period from 1972-2006, Masaya emitted an average of ~1100 metric tonnes/day (t/d) of

SO2 (Appendix B.1-2, and references therein). When the 1411 measurements for which

plume speeds were reported are normalized to a standard plume speed of 1 m/s, SO2 flux at

Masaya has a mean daily flux of ~160 t/d (Figure 3-1; Appendix B.1-2, and references

therein). Current activity is marked by the persistence of a gas crisis that began in 1993

(Rymer et al., 1998; Delmelle et al., 1999; Williams-Jones et al., 2003; Smithsonian

Institution, 2006). During the dry season in Nicaragua (December to April), easterly trade

winds typically blow the volcanic SO2 plume from Masaya towards the Pacific Ocean.

Though the terrain downwind of the volcano is irregular and marked by steep ridges and

valleys, a well-developed road network exists in the region west of the volcano (Figure 3-2).

The Ticuantepe and Llano Pacaya roads are roughly perpendicular to the plume’s normal

propagation direction at distances of ~5 and ~15 km downwind of the vent, respectively

(Figure 3-2). The combination of consistent degassing, well-developed infrastructure at two

distances downwind of the active vent, and general accessibility of the volcano make it a

prime location for the study of the plume’s evolution through time and space. This paper

reports the results of two intensive, month-long field campaigns at Masaya volcano aimed at

utilizing the latest field methods to better evaluate the presence or absence of significant

depletion of SO2 in volcanic plumes.

3.3 Methodology

All measurements were made between February 25 and March 16, 2005, and

February 16 and March 12, 2006. SO2 measurements were made only on days when the

consistent easterly trade winds prevailed; days dominated by strong on-shore breezes (as

determined by low wind speeds, wide plumes, and eastward motion of clouds) were not

utilized for data collection. Aerosol measurements were made only on days that were

predominantly cloud-free.

3.3.1 SO2 flux

In order to make simultaneous measurements on the two downwind roads, SO2 data

were collected by conducting multiple traverses beneath the volcanic plume using one of two

46

Figure 3-1: Mean SO2 fluxes (normalized to a plume speed of 1 m/s) grouped by month. Error bars represent one standard deviation of repeat measurements. Note break in x-axis. See also Appendix B.1-2.

Chapter 3 – SO

2 Flux and Apparent Dow

nwind D

epletion


47

Figure 3-2: Map of road network in the vicinity of Masaya volcano. Roads utilized for measuring SO2 flux, the Ticuantepe and Llano Pacaya roads, are shown. Star indicates the location of the active vent.


48

small UV spectrometers known as the FLYSPEC (Horton et al., 2006) and the FLYSPEC

v2. Each instrument consists of a collimating lens mounted on an Ocean Optics USB2000

spectrometer. The spectrometers were each housed within sturdy plastic cases outfitted with

UV filter windows and were powered through a USB connection to sub-notebook

computers. Additionally, each instrument had a GPS antenna and calibration cells integrated

into the set-up: the FLYSPEC had an external GPS antenna and manually-operated

calibration cells, while the FLYSPEC v2 was fully integrated and had both a GPS antenna

and an automated calibration cell array within its plastic case. Depending on the

configuration of the spectrometer and the UV window in the plastic case, the instruments

were attached either to the passenger-side side mirror, or to the roof, of the field car (Figure

3-3). By measuring diffuse, zenith sky UV radiation, the FLYSPEC measures an integrated

vertical column amount of SO2 within the plume (Horton et al., 2006).

In 2005, traverses were made using one FLYSPEC; in 2006 data were often collected

simultaneously by two spectrometers, one FLYSPEC and one FLYSPEC v2, each traversing

at different distances downwind of Masaya’s degassing vent. The total number of traverses

was 701, or approximately 20 measurements per day. Spectra were captured at intervals of 1,

2, or 3 scans per second, and then averaged to one measurement per second in order to

coincide with the frequency of the GPS data collection. The spectra captured along a

traverse beneath the plume were representative of an integrated cross-section of the plume,

which, after corrections for deviations from perpendicular of the road to the plume, was

multiplied by a plume speed in order to determine a flux of SO2 (e.g., Casadevall et al., 1981;

Stoiber et al., 1983; Williams-Jones et al., 2006b).

3.3.2 Plume speed

Plume speeds were measured according to the multiple-spectrometer method

described by Williams-Jones et al. (2006a). Two FLYSPECs, placed a known distance apart

(~30-50 m) along the plume’s axis of propagation, recorded variations of concentration

within the plume as it passed over. Iterative correlation of the ~30 minute datasets yielded a

time-separation for the instruments, which, when combined with the known spatial

separation, allowed for calculation of the speed of the plume at plume height. Hand-held

anemometer measurements were also made each day, both between traverses and during the


49

Figure 3-3: FLYSPEC components and field deployment. (a) FLYSPEC (b) FLYSPEC v2

a.)


50

b.)


51

collection of FLYSPEC-based plume speed data. Anemometer measurements were made

~2 m above ground level and were 5-minute averages.

The accurate determination of the speeds of volcanic plumes has often proved

problematic. Despite recent advances in the ability to measure the ‘true’ plume speed

(McGonigle et al., 2005a, b; Williams-Jones et al., 2006a), which have greatly reduced the

uncertainties associated with SO2 fluxes as with this study, datasets from previous years did

not have the advantage of such methods. Accordingly, much archival SO2 flux data from

Masaya may be skewed as a result of inaccurate estimations of plume speed (Appendix B.2).

Simple scaling of hand-held anemometer estimations of plume speeds to plume speeds

determined by the dual-spectrometer method is not possible; ground-based anemometer

measurements are not consistent relative to measured plume speeds (Table 3-1). Likewise,

radiosonde data have shown that the speed of winds aloft varies greatly from day to day

relative to wind speeds at ground level (Williams-Jones et al., 2006). Rather than make

attempts to scale plume speeds measured by different methods, Zapata et al. (1997)

suggested the normalization of all reported fluxes to a plume speed of 1 m/s. This omits the

largest source of uncertainty in the measurements and facilitates the comparison of datasets

from year to year, as well as from volcano to volcano. Thus, in addition to accounting for

measured plume and wind speeds, SO2 fluxes were also always normalized to a standard

plume speed of 1 m/s.

In order to find true, rather than normalized, fluxes, the dual-spectrometer plume

speed determination method was attempted as often as possible. Hardware problems at the

start of the 2006 field campaign prevented the implementation of the method until February

26. Similarly, on March 14, 2005, and March 9 and 10, 2006, the plume was located over

regions that did not allow for an adequate spatial separation between the two spectrometers.

Plume speed data files for February 27, 2005 and February 24, 2006 became corrupted and

were not useable for calculation of plume speeds. Measurements on March 2, 4, 6, and 8,

2006 initially yielded improbably high values for plume speeds. Upon inspection of the data,

it became apparent that the plume was moving too quickly to be adequately sampled given

the temporal sampling resolution of the FLYSPECs and the distance between the two

spectrometers, and data were therefore not representative of the true plume speeds. For all

problematic dates, in cases when hand-held anemometer data was available, it was


52

Table 3-1: Comparison of simultaneous measurements of ground-level wind speed by anemometer and plume speed measurements made according to the dual-spectrometer method.

Date Road Anemometer Wind Speed

(m/s)

Anemometer Maximum Gust (m/s)

Dual-Spectrometer Plume Speed

(m/s) 2/25/2005 Llano Pacaya 3.7 7 10.3 2/26/2005 Llano Pacaya 2.8 5.9 9 3/5/2005 Ticuantepe 1.4 4.4 3.4 3/6/2005 Ticuantepe 1.7 4.1 2.1 3/7/2005 Ticuantepe 2.2 5.2 12 3/11/2005 Ticuantepe 3.6 8.5 2.1 3/12/2005 Ticuantepe 2.6 7.5 3.5 3/15/2005 Ticuantepe 1.1 2.1 12.4 3/16/2005 Ticuantepe 1.1 2.9 10.5 2/23/2006 Llano Pacaya 7.6 12.5 12.1 2/26/2006 Llano Pacaya 6.5 9.9 22.5 2/27/2006 Llano Pacaya 5.6 12.4 12.4 2/28/2006 Llano Pacaya 9.3 13.2 19 3/1/2006 Llano Pacaya 7.2 10.2 20.6 3/3/2006 Llano Pacaya 4.9 8.3 7.3 3/7/2006 Llano Pacaya 6.1 10.9 7 3/12/2006 Llano Pacaya 6.1 10.2 16.8


53

substituted for a plume speed derived from spectrometer data. However, as ground-level

winds are often substantially less than those aloft, it is important to note that estimates of

flux relying on such ground-level wind data are likely significantly misrepresenting the true

flux. Some dates lacked both adequate spectrometer-based and hand-held anemometer wind

data; on these dates, fluxes were calculated only as values normalized to a wind speed of

1 m/s.

3.3.3 Aerosols

Of the 701 SO2 traverses, approximately 60 incorporated the use of Microtops II sun

photometers to measure aerosol optical depths. The Microtops II is a 5-channel Volz-type

sun photometer (Morys et al., 1996; Porter et al., 2001), with data capture taking place in all

channels simultaneously. Two of the instruments used had channels at wavelengths of

380 nm, 440 nm, 500 nm, 675 nm, and 870 nm, while the third had channels at 380 nm,

500 nm, 675 nm, 936 nm, and 1020 nm. The instruments used in this study were calibrated

at Mauna Loa observatory, both by the Solar Light Company and independently.

Measurements are made by aligning the instrument’s field of view with the sun for the

duration of the measurement. Scan length may be adjusted by the Microtops II operator in

order to best suit the field conditions; scan length in this study was set to 10 scans,

equivalent to a sampling time of ~3 seconds. Of the 10 scans for each measurement, the

Microtops II was programmed to select the 4 strongest signals to be averaged for further

processing. The short scan length was chosen for agreement with sampling times used at

other volcanoes (J. Porter, pers. comm., 2005), and also as a compromise between the need

for rapid measurements (to reduce the effect of the plume’s temporal variations on

measurements) and ensuring good data quality and a representative dataset.

As SO2 data were being captured over the course of a single traverse, the field

vehicle stopped at intervals of ~300 m to allow for a Microtops II measurement. Five

photometer measurements were made at each stop in order to allow for averaging, and to

constrain variability (coefficients of variation for sets of 5 measurements were generally less

than 1.5%). Erroneous scans due to traffic blocking the instrument’s field of view were

removed from the dataset. Each measurement site was recorded by a hand-held GPS to

facilitate post-processing of the raw data. Though the Microtops II calculates aerosol optical


54

depths in real-time via a built-in microprocessor system, raw voltages in the 500 nm channel

were re-processed using software after Watson and Oppenheimer (2000). The post-

processing allowed for re-calculation of solar zenith angles based on the GPS data.

Additionally, because each plume traverse dataset included background measurements that

were made in plume-free sky (as determined by FLYSPEC measurements) before and after

encountering the plume, the software allowed true plume optical depths to be determined by

subtracting background optical thicknesses from that of the plume. Subsequently, standard

methodology for the calculation of particle size distributions was used (for more detailed

discussion of methodology see Watson and Oppenheimer, 2000; Mather et al., 2004;

Appendix A). From the calculated size distributions, a particle column density for each

measurement site on each traverse was calculated. These values were subsequently averaged

for each individual traverse, and multiplied by the length of the traverse and normalized to a

plume speed of 1 m/s in order to determine a particle flux for each traverse.

3.4 Uncertainties

3.4.1 Spectrometer bias

Elias et al. (2006) performed a rigorous comparison of the performances of three

unique UV-spectrometers (COSPEC, FLYSPEC, and mini-DOAS). The study showed that

the three instruments reported comparable results when utilized side-by-side for traverses

beneath the SO2 plume at Kilauea, Hawaii. Despite the proven interchangeability of these

three types of spectrometers, further precautions, as outlined below, were taken in order to

avoid any spectrometer bias within the datasets of the two FLYSPECs.

All measurements in 2005, and some in 2006, were completed with only a single

FLYSPEC making measurements on a given day. On days with only a single spectrometer,

measurements consisted of half a day on the proximal Ticuantepe road and half the day on

the distal Llano Pacaya road. This routine was varied daily so that the morning traverses

alternated between the proximal and distal roads, effectively eliminating the possibility of any

diurnal variations in SO2 output leading to discrepancies within the dataset. The two

FLYSPECs utilized in the 2006 field campaign were tested side-by-side for 6 traverses

beneath the plume to ensure agreement of the two datasets (Appendix B). To further limit

any instrumental bias, the two spectrometers were deployed for daily measurements such


55

that one spectrometer was used proximally, and the other distally. At mid-day, the

FLYSPECs were switched. As with days consisting of a single FLYSPEC making

measurements, the dual-spectrometer routine was varied: if one spectrometer was utilized in

the morning on the Llano Pacaya road, and the other on the Ticuantepe road, the starting

locations would be reversed for the next morning.

3.4.2 Plume speed

Accurate determination of a volcanic plume’s speed has long been recognized to be

the greatest source of uncertainty in the measurement of gas fluxes from volcanoes (e.g.,

Stoiber et al., 1983; Doukas, 2002; McGonigle et al, 2005a, b; Williams-Jones et al., 2006a, b).

Ground-based anemometer measurements and readings from nearby airports have been

used frequently as proxies for plume speed. However, wind profiles can vary by orders of

magnitude both vertically and laterally; wind speed at an airport several kilometers away, or

ground wind speeds beneath the plume are not necessarily representative of the speed of the

plume itself, which can result in uncertainties of at least 10–40% (e.g., Stoiber et al., 1983;

Doukas, 2002) to more than 100% (Williams-Jones et al., 2006a) for the derived fluxes.

While the uncertainties associated with determining the speed of a volcanic plume have not

been eliminated, by utilizing a dual-spectrometer method as a means to track variations in

the plume’s concentration over a known distance, the uncertainties have likely been reduced

to below 15% (Williams-Jones et al., 2006a). On days when this method was not possible in

practice, normalization to a wind speed of 1 m/s was used in tandem with ground-based

anemometer measurements for the calculation of SO2 fluxes.

3.4.3 Aerosol measurements

To ensure measurement of plume aerosols only, use of the Microtops II requires

cloud-free background sky. While this rarely posed a problem on the Llano Pacaya road, the

Ticuantepe road was often quite cloudy, especially early in the day. The morning clouds

over Ticuantepe are likely the result of vertical air movement associated with the dissipation

of a nocturnal temperature inversion, while clouds that persisted into the afternoon hours

may have been due to orographic lifting as the easterly trade winds blew in the direction of

the Llano Pacaya ridge. Clouds often precluded aerosol measurements on the Ticuantepe

road. As such, a disproportionate amount of aerosol data is from the Llano Pacaya road,


56

and few days resulted in the desired quasi-simultaneous aerosol traverses on the two roads.

Because the dataset of aerosol measurements on the Ticuantepe road is not as

comprehensive as initially intended, the number of measurements made may not be

sufficient to fully identify and remove noise due to natural variability in the plume’s aerosol

content. Further, despite efforts to avoid cloudy sky measurements, cirrus clouds were

present on a number of occasions. If very thin portions of the clouds were not detected by

the operator of the Microtops II, measurements may have inadvertently been taken with

small amounts of cirrus cloud in the field of view of the instrument.

Further complicating measurements are the differing operational geometries of the

FLYSPEC and the Microtops II. The FLYSPEC makes measurements based on scattered

sunlight in the vertical field of view of the instrument. In contrast, the Microtops II requires

the sun to be directly in the instrument’s field of view, thereby forcing a slant column

measurement. Depending on the time of day and the plume’s position over the road, the

differing fields of view may have caused some measurements, which, while ‘out of the

plume’ with respect to the FLYSPEC measurements, to still be ‘in plume’ for Microtops II

measurements. For example, if the field vehicle had already passed beneath the plume, the

FLYSPEC would indicate only background levels of SO2. However, if the sun was behind

the car for that particular traverse, the Microtops II measurement coinciding with

background SO2 levels overhead may have actually been measuring back towards the plume,

perhaps with a significant portion of it in the field of view. Post-processing of aerosol data

indicated that this was likely the case for some traverses; pre-traverse and post-traverse

optical thicknesses for some traverses differed substantially despite both being considered

‘out of the plume.’ Therefore, aerosol measurements were not always consistent with the

corresponding SO2 measurements.

3.5 Results

3.5.1 SO2 flux

Results of this study’s 701 measurements of SO2 flux at Masaya are reported in Table

3-2 and Table 3-3. Figure 3-4 gives a graphical representation of those fluxes normalized to

a standard plume speed of 1 m/s, separated by road. For the two month-long periods over

which measurements were made, the average daily flux, normalized to a plume speed of


57

Table 3-2: Summary of SO2 flux data as measured on the Ticuantepe road.

Date Number

of Traverses

Plume Speed (m/s)

Mean Daily Flux @ Standard Plume Speed of 1 m/s

(tonnes/day)

S.D. Mean Daily Flux

@ Measured Plume Speed (tonnes/day)

S.D.

2005 2/25/05 6 10.3 LP 164 47 1684 485 2/26/05 8 9.0 LP 350 94 3153 846 2/27/05 5 423 44 3/3/05 7 2.4 T 141 41 339 100 3/5/05 10 3.4 T 54 13 182 44 3/6/05 15 2.1 T 65 25 136 53 3/7/05 9 12.0 T 68 23 811 282 3/11/05 11 2.1 T 65 25 137 53 3/12/05 9 3.5 T 71 18 249 65 3/14/05 12 3.0 T* 85 35 254 104 3/15/05 7 12.4 T 78 24 965 294 3/16/05 10 10.5 T 116 49 1217 518

Average# 9 6.4 140 37 830 258 2006

2/16/06 3 6.5 LP* 99 34 646 221 2/17/06 14 3.0 T* 113 34 340 102 2/18/06 11 2.2 T* 86 26 189 56 2/19/06 12 1.4 T* 82 30 115 43 2/20/06 10 2.5 T* 84 32 211 80 2/21/06 14 1.1 T* 105 28 115 31 2/23/06 7 12.1 LP 164 38 1983 458 2/24/06 11 2.4 T* 130 55 311 132 2/26/06 23 22.5 LP 158 88 3552 1976 2/27/06 20 12.4 LP 98 48 1212 595 2/28/06 13 19.0 LP 151 39 2876 734 3/1/06 23 20.7 LP 156 54 3224 1115 3/2/06 12 8.0 LP* 165 47 1321 377 3/3/06 25 7.3 LP 154 47 1127 340 3/4/06 29 7.9 LP* 119 42 943 330 3/6/06 8 9.6 LP* 103 32 989 303 3/7/06 6 7.0 LP 120 35 837 242 3/8/06 24 4.5 LP* 117 39 528 174 3/9/06 12 161 32 3/10/06 9 176 50 3/12/06 7 16.8 LP 179 58 3014 972

Average# 14 8.8 130 42 1239 436

Ticuantepe Average# 12 8.0 133 40 1089 371

Superscripts T and LP are indicative of the road on which the plume speed (derived from 5-minute windows of full plume speed scan) utilized for flux calculation was measured. Dashed line denotes a landslide in Santiago crater, which blocked the vent, lowering SO2 fluxes.

# all averages were determined using individual measurements from the entire 701-measurement dataset. * plume speed data in which hand-held anemometer data have been substituted for spectrometer-based

measurements.


58

Table 3-3: Summary of SO2 flux data as measured on the Llano Pacaya road.

Date Number

of Traverses

Plume Speed (m/s)

Average Daily Flux @ Standard Plume Speed of 1 m/s (tonnes/day)

S.D.

Average Daily Flux @ Measured

Plume Speed (tonnes/day)

S.D.

2005 2/25/05 6 10.3 LP 204 38 2098 396 2/26/05 7 9.0 LP 299 96 2688 863 2/27/05 6 252 53 3/7/05 2 12.0 T 53 27 636 322

3/12/05 6 3.5 T 46 11 161 38 3/14/05 6 3.0 T* 43 13 128 38 3/15/05 4 12.4 T 54 15 663 181

Average# 5 8.4 136 36 1062 306 2006 2/16/06 2 6.5 LP* 67 30 436 193 2/17/06 10 7.9 LP* 56 23 445 185 2/18/06 12 7.8 LP* 44 9 346 71 2/19/06 11 9.3 LP* 48 13 446 117 2/20/06 8 6.3 LP* 54 13 342 85 2/21/06 12 5.2 LP* 51 13 264 69 2/23/06 6 12.1 LP 86 40 1035 486 2/24/06 12 9.4 LP* 72 21 676 199 2/26/06 8 22.5 LP 99 42 2228 946 2/27/06 11 12.4 LP 63 19 778 233 2/28/06 10 19.0 LP 60 17 1144 317 3/1/06 8 20.7 LP 74 17 1524 362 3/2/06 23 8.0 LP* 94 33 749 262 3/3/06 25 7.3 LP 100 22 727 159 3/4/06 24 7.9 LP* 79 30 622 234 3/6/06 31 9.6 LP* 54 23 523 220 3/7/06 9 7.0 LP 67 14 466 95 3/8/06 20 4.5 LP* 64 18 287 80 3/9/06 5 85 12

3/10/06 6 134 26 3/12/06 9 16.8 LP 82 16 1381 271

Average# 13 10.5 73 21 759 241

Llano Pacaya

Average# 11 10.0 89 25 832 257

Superscripts T and LP are indicative of the road on which the plume speed (derived from 5-minute windows of full plume speed scan) utilized for flux calculation was measured. Dashed line denotes a landslide in Santiago crater, which blocked the vent, lowering SO2 fluxes.

# all averages were determined using individual measurements from the entire 701-measurement dataset. * plume speed data in which hand-held anemometer data have been substituted for spectrometer-based

measurements.

59

Figure 3-4: Mean daily SO2 flux (normalized to a plume speed of 1 m/s) as measured on each of the two roads downwind of Masaya’s active vent. Error bars represent one standard deviation of repeat measurements. Note break in x-axis. See Tables 3-1 and 3-2.

Chapter 3 – SO

2 Flux and Apparent Dow

nwind D

epletion


60

1 m/s, was approximately 140 ± 37 t/d in 2005 and 130 ± 42 t/d in 2006 as measured on the

Ticuantepe road. Fluxes on the Llano Pacaya road were 136 ± 36 t/d and 73 ± 21 t/d in

2005 and 2006, respectively. With this normalization, the maximum recorded average daily

flux on Ticuantepe was on February 27, 2005 at 423 ± 44 t/d, while the minimum was

54 ± 13 t/d, measured on March 5, 2005. On the Llano Pacaya road, the maximum

recorded average flux was 299 ± 96 t/d on February 26, 2005. The minimum recorded

average flux on Llano Pacaya was 43 ± 13 t/d on March 14, 2005. However, there were also

a number of dates on which SO2 was below detection levels on the Llano Pacaya road.

Using the dual spectrometer plume speed method, it was determined that plume

speeds ranged from 2.1 to 22.5 m/s on days in which measurements were made. Overall,

winds were weaker in 2005 than in 2006, with a 2005 average measured plume speed

(calculated using a 5-minute moving window) of 6.8 ± 4.4 m/s, as compared to an average of

12.2 ± 6.4 m/s in 2006. On some days during the 2005 field season, the easterly trade winds

which normally dominate the region’s weather were weakened such that an on-shore breeze

from the Pacific coast prevailed. Under such conditions, Masaya’s plume spreads out

laterally along the Llano Pacaya ridge, causing any traverse measurements to significantly

overestimate true SO2 fluxes. For the purposes of this study, data were not collected on days

when the plume was deemed to be ‘smeared’ (as determined by uncharacteristically long

profiles and low wind speeds). Taking into account only dates with accurately measured,

spectrometer-based plume speeds, average daily fluxes ranged from 115 ± 31 to

3552 ± 1976 t/d, with average values of 1483 ± 510 t/d on Ticuantepe and 1194 ± 359 t/d

on Llano Pacaya.

Between February 27, 2005 and March 3, 2005 a significant drop in SO2 flux was

noted. On the morning of March 3, 2005 Masaya Volcano National Park workers reported

that a landslide had occurred the previous night within the active Santiago crater. A visibly

reduced plume from the crater’s active vent suggested that the partial collapse of the crater

wall may have caused a blockage that was inhibiting the escape of SO2 from the volcano.

The visual observations were confirmed by the measured drop in SO2 emissions, when

normalized to a plume speed of 1 m/s, from an average of ~400 t/d prior to the landslide to

an average of <100 t/d following the landslide. At measured plume speeds, average fluxes


61

of SO2 immediately after the slide were nearly ten times less than those pre-slide

(3153 ± 846 t/d on February 26, 2005 versus 339 ± 100 t/d on March 3, 2005). Flux

measurements made in the days immediately following the landslide were conducted only on

the Ticuantepe road. Though significantly lower than prior to the vent blockage, SO2

column thicknesses on the Ticuantepe road remained above background levels and were

easily detected, whereas SO2 was below detection levels on the Llano Pacaya road for several

days following the vent blockage. By 2006, SO2 was again detected on the Llano Pacaya

road, though the visible plume at the crater was still significantly diminished relative to the

start of the 2005 field campaign.

3.5.2 Aerosols

Due to the wide range of atmospheric variables that control the size of, amount of,

and rate of conversion to sulfate aerosols, all of which are necessary to calculate fluxes of

sulfate aerosols (e.g., Mather et al., 2004), aerosol data are presented as conservative

estimates of particle fluxes (Table 3-4). Of the approximately 60 traverses that incorporated

aerosol measurements, only 5 pairs were sufficiently close in time to be utilized in the

assessment of sulfate conversion between the two roads. For the five pairs of traverses, the

minimum particle concentration on Ticuantepe was 1.1 x 1019 particles/s, while the

maximum was 1.5 x 1019 particles/s. The corresponding range for Llano Pacaya was

8.0 x 1018 particles/s to 2.4 x 1019 particles/s.

3.6 Discussion

3.6.1 Variability in Masaya’s SO2 emissions

The long-term degassing cycles at Masaya volcano have been attributed to

convective overturn within the magma plumbing system and alterations to the geometry of

Masaya’s crater and shallow substructure (Rymer et al., 1998; Williams-Jones et al., 2003).

These large-scale variations in SO2 flux (e.g., 10 t/d in 1992 to 6118 t/d in 1998; Appendix

B.2) occur over years or decades and lead to a coefficient of variation close to 100% of the

mean when each individual measurement from 1972 to 2006 is taken into account.

However, on shorter time scales, variability persists (Figure 3-5) that cannot be accounted

for by such large-scale magmatic processes. Data from entire month-long field campaigns


62

Table 3-4: Plume particle concentrations for five pairs of aerosol traverses. T and LP in traverse name indicate Ticauntepe and Llano Pacaya roads, and F and F2 represent FLYSPEC and FLYSPECv2, respectively.

Date and Traverses Ticuantepe Particle Flux

(in particles/s @ Standard Plume Speed of 1 m/s)

Llano Pacaya Particle Flux (in particles/s @ Standard

Plume Speed of 1 m/s)

3/1/2006 – T20(F) and LP6(F2) 1.5 x 1019 2.4 x 1019 3/6/2006 – T2(F2) and LP5(F) 1.3 x 1019 1.7 x 1019 3/6/2006 – T6(F2) and LP9(F) 1.3 x 1019 1.3 x 1019 3/8/2006 – T4(F2) and LP4(F) 1.1 x 1019 1.2 x 1019 3/8/2006 – T6(F2) and LP8(F) 1.3 x 1019 8.0 x 1018


63

Figure 3-5: Histograms displaying the variability of SO2 flux measurements on the Ticuantepe road for (a) a month (February 16, 2006 – March 12, 2006), and (b) a day (March 4, 2006)

a.)

b.)


64

as well as single days of measurements resulted in values of SO2 flux with coefficients of

variation near 30%. Similar variability is found upon investigation of the historical dataset,

regardless of the disparity in the number of measurements between the current study and

previous field campaigns. This variability, apparently inherent to SO2 flux measurements at

Masaya, may be due to a number of factors, from the non-instantaneous nature of a single

measurement, to natural heterogeneities in the plume as a result of puffing at the vent

(Branan et al., submitted), to non-uniform interaction of the plume with the environment.

3.6.2 Apparent loss of SO2 with increasing distance from vent

In previous SO2 monitoring campaigns at Masaya (e.g., Stoiber et al., 1986; Williams-

Jones et al., 2003; CCVG-IAVCEI, 2003; McGonigle et al., 2004), it was assumed that, while

dispersion would inevitably occur as the plume propagated further downwind, no

measurable mass loss would occur: any vertical dispersion is accounted for by the fact that

the spectrometers (COSPEC, FLYSPEC, or mini-DOAS) measure concentrations over an

‘infinite’ vertical column; lateral dispersion in the cross-wind direction was assumed to affect

measurements only in that profiles downwind would be wider than those made at more

proximal locations; and dispersion along the axis of the plume’s propagation is not an issue

due to the relatively continuous nature of the plume. However, inspection of the SO2 data

from both 2005 and 2006 indicated that there is a significant discrepancy between the

magnitudes of the SO2 fluxes calculated for the Ticuantepe and Llano Pacaya roads.

Regardless of the overall magnitude of SO2 output or degassing regime (freely degassing or

partial blockage of the vent by the landslide), fluxes calculated for the traverses on the more

distal Llano Pacaya road are generally ~33% to ~50% less than fluxes calculated for the

proximal Ticuantepe road (Figure 3-4, Table 3-2, Table 3-3). This suggests that there is

some downwind loss of SO2.

3.6.3 Possible sources of loss

3.6.3.1 Deposition of sulfur

Loss by deposition of S on the earth’s surface was initially thought to be one

possibility in accounting for the discrepancy. Delmelle et al. (2001) estimated that

approximately 10% of the total SO2 flux from Masaya is lost to soils and vegetation through


65

dry deposition of S. However, this estimate is based on SO2 flux measurements conducted

solely on the Llano Pacaya road. Given our observations that fluxes measured on Llano

Pacaya are up to 50% less than those measured on the more proximal Ticuantepe road, the

volume of dryly deposited sulfur measured by Delmelle et al. (2001) may amount to only 5%

of the total SO2 emissions at Masaya. In reality, the total loss of S to dry deposition is likely

to be even less than 5%; Ticuantepe itself is 5 km downwind of the active vent and fluxes

measured there are likely to be underestimates of actual SO2 emissions as well. Other

estimates of S dry deposition from an SO2 plume with a single point source are on the same

order: approximately 8% (Bourque and Arp, 1996). Accordingly, dry deposition of S from

the SO2 plume is not sufficient to fully explain the difference in measured fluxes.

3.6.3.2 Conversion of SO2 to sulfate aerosols

During daytime conditions, the principal oxidant in the conversion of SO2 to sulfate

aerosols is recognized as the hydroxyl radical, such that the net atmospheric reaction for

sulfate generation (and subsequent regeneration of hydroxyl) is as follows (Eatough et al.,

1994, and references therein):

OH + SO2 (+O2, H2O) H2SO4 + HO2 [3-1]

Should such conversion occur downwind of the degassing volcano, fluxes of SO2 measured

by spectrometers may underestimate the actual flux from the volcano’s vent.

Ideally, measurement of sulfate aerosol fluxes would be as simple as that of SO2

fluxes. However, manipulation of Microtops II data in order to derive accurate sulfate

fluxes requires precise knowledge of atmospheric conditions such as relative humidity and

sun angle, as well as conditions within the plume itself (i.e., temperature, ash content, water

content, proportion of the plume comprising non-sulfate species) and plume age. Each of

these variables can affect the number of sulfate particles in the plume, the size distribution of

sulfate particles in the plume, and the rate of conversion of SO2 to sulfate. As the in-plume

variables are nearly-impossible to constrain, especially on the time scale of the multiple

Microtops II aerosol measurements, even the conservative estimates of particle flux

presented here are not sufficient to accurately quantify the rates of sulfate conversion and

SO2 loss from the plume. The particle fluxes calculated for the 5 pairs of traverses are all of

the same order of magnitude, with no more than a twofold difference in the magnitudes


66

seen on the Ticuantepe versus Llano Pacaya roads. Some traverses show particle column

densities following the same general trend as SO2 concentration path-length within the

plume (Figure 3-6, Appendix B.6), possibly indicating that variations in aerosol content

within the plume merely correspond to equivalent variations in the overall concentration of

the plume as a whole. Still other profiles exhibited little to no apparent correlation between

SO2 concentration path-lengths and particle column densities, with highly variable values of

column density over the course of a given traverse. The low resolution of the aerosol data in

this study, the numerous assumptions associated with determining a plume’s aerosol content,

and the inconsistencies in the collected dataset indicate that the Microtops measurements

made in this study may not be a wholly reliable estimate of aerosol conversion in the plume

at Masaya. For this reason, it was determined that other means would be better suited to

discounting aerosol conversion as the cause of the apparent loss of SO2 downwind.

Previous studies of aerosol conversion in volcanic plumes have quoted conversion

rates ranging from negligible (e.g., McGonigle et al., 2004) to close to 100%

(e.g., Oppenheimer et al., 1998). However, conversion of SO2 to sulfate in dry, summer

conditions like those at Masaya have been found to be approximately 5-10% per hour

(Eatough et al., 1994). Based on plume speeds measured with the dual-spectrometer

method, plume travel times over the ~10 km separation between the Ticuantepe and Llano

Pacaya roads range from 7 to 79 minutes. Even 79 minutes is not sufficient to account for

the 33-50% SO2 loss measured. Further, with the Ticuantepe road lying only 5 km from the

active vent, no significant conversion to aerosols is likely to take place prior to the plume

reaching the road. Based on the slow conversion rates that are likely for the Masaya region,

it is probable that most, if not all, aerosols in the plume are due to high-temperature, near-

vent conversion (Allen et al., 2002; Mather et al., 2003) and maintain a relatively constant

concentration over the spatial scope of this study.

3.6.3.3 Topographic modification of winds

Another possible reason for the discrepancy in fluxes between the proximal and

distal roads is the effect of topography. Despite a number of previous studies of volcanic

gas at Masaya (e.g., Stoiber et al., 1986; Horrocks et al., 1999; CCVG-IAVCEI, 2003;

Williams-Jones et al., 2003), all have failed to address the geometry of the roads with respect


67

Figure 3-6: Example of simultaneous particle and SO2 column measurements from March 8, 2006.


68

to elevation as well as location. The Llano Pacaya ridge, ~15 km downwind, at its highest

point, is more than 400 m higher than the degassing summit of Masaya. A series of

secondary ridges and valleys branch off the main ridge in the direction of the volcano

(Figure 3-7). The regional trade winds generally blow northeast to southwest, almost directly

over the highest points on the Llano Pacaya ridge. The potential for interaction of the trade

winds with the complex topography cannot be ignored. Wind field modifications by

topography at other volcanoes have been previously noted (e.g., Graziani et al., 1997; Favalli

et al., 2004), however no investigation of wind speed variations has been conducted at

Masaya as yet.

The ~400 m elevation difference over a horizontal distance of ~15 km from

Masaya’s summit to the Llano Pacaya road leads to orographic acceleration of winds (Oke,

1987). As such, a given parcel of the plume not only will be diluted across the width of the

plume as it propagates away from the vent, but also along the axis of the plume (Figure 3-8),

most noticeably at the highest elevation, where acceleration is greatest (Oke, 1987). Should

winds be significantly faster over the Llano Pacaya road than over the Ticuantepe road, SO2

column abundances and, therefore, derived fluxes, would vary greatly between the two

measurement sites. Oke (1987) gives the following equation as a simple, first-order means to

calculate the likely acceleration factor of winds over a 2-dimensional ridge:

Vmax / Vup = 1 + b(H/X) [3-2]

where Vmax / Vup is the actual acceleration factor, or the ratio of the maximum wind velocity

relative to the velocity of the wind upwind of the ridge; b is the theoretical maximum

acceleration factor for the topography in question and is equal to 2 for a 2-dimensional ridge,

1.6 for a 3-dimensional hill, and 0.8 for a step-up (Taylor and Lee, 1984; Oke, 1987); H is the

height of the ridge; and X is the distance from the peak of the ridge to the point where the

elevation is equal to H/2 (Figure 3-9). In general, actual maximum acceleration factors,

Vmax / Vup, for this topographic geometry range from 1.6 to 1.8 (Oke, 1987). Winds are

perturbed to a lesser degree within an envelope ranging from just upwind of the ridge to just

downwind of the ridge, and up to an altitude approximately equivalent to X. In the case of

Masaya, there is likely the further complicating factor of wind flow in valleys and the

associated acceleration. When the wind flow is parallel to the valley bottom, valleys produce

jetting of winds, with maximums at the narrowest points (Oke, 1987). The valleys on the


69

Figure 3-7: Digital elevation model of Masaya region highlighting the complex topography between the active vent and the two roads downwind (shown in red). Vertical exaggeration is 5x.


70

Figure 3-8: Example of the effect of increasing plume speed on measured concentration path-lengths of SO2. A twofold increase in plume velocity corresponds to a twofold decrease in the amount of SO2 sensed by the spectrometer at a given time. Spectrometer fields of view are denoted by dashed gray lines. Continuous plume is portrayed as discrete puffs for illustrative purposes.


71

Figure 3-9: A generalized pattern of airflow over a two-dimensional ridge. The point of maximum wind velocity is indicated by a dark gray star, and the points of minimum wind velocity are denoted by light gray stars. Variables for use in Equation 3-2 are also shown.


72

eastern side of the Llano Pacaya ridge are generally narrowest towards the top of the ridge,

which would exacerbate the effects of the ridge-induced acceleration. This is especially

important given the likely confinement of the plume to the steep valleys near the Llano

Pacaya ridge (Delmelle et al., 2002). Taking into account acceleration over the Llano Pacaya

ridge and possible exacerbation of the effect by acceleration in the secondary valleys

branching from the ridge, an expected range of acceleration factors at Masaya is at least 1.2

to 1.4. However, equation 3-2 is only a simplistic approach to determining theoretical

acceleration factors over complex topographic features, such as the Llano Pacaya ridge.

Taylor and Teunissen (1985) conducted an extensive wind speed monitoring

campaign in the region surrounding Askervein hill in Scotland in an effort to study boundary

layer flow over low hills. Askervein is a small, roughly elliptical hill with a major axis of

~2 km, a minor axis of ~1 km, and a maximum elevation of ~125 m above the surrounding

flat land. This configuration, though on a smaller scale, is not dissimilar to the Llano

Pacaya’s approximately 15 km x 8 km dimensions and elevation of ~400 m above the

surrounding topography. In the Askervein study, winds perpendicular to the long axis of the

hill were found to be 1.6 to 1.9 times faster on top of the hill than on flat land upwind of the

hill. While the relief of the Llano Pacaya ridge is not quite as extreme as Askervein relative

to its surroundings, it also has the complicating factor of secondary ridges and valleys, which

may further enhance acceleration of winds over the ridge (Oke, 1987). Accordingly,

reasonable acceleration factors at Llano Pacaya are expected to be similar to those

documented at Askervein, with wind speeds atop the ridge reaching up to twice those in the

vicinity of the Ticuantepe road.

Photographs of Masaya’s plume taken from the space shuttles support the assertion

that topographically-induced wind acceleration causes dilution of the plume over the Llano

Pacaya ridge (Figure 3-10). In each of the images, there appears to be a visible dilution of

the plume beginning immediately adjacent to the windward side of the ridge, and a

subsequent semi-transparent section of the plume directly over the ridge. Leeward of the

ridge, the plume appears to re-gain characteristics similar to those prior to encountering the

ridge, becoming more concentrated and opaque. However, the degree of the topographic

acceleration’s effect on the plume and its propagation speed will depend largely on the

height of the plume and ambient atmospheric conditions. The region surrounding Masaya is


73

Figure 3-10: Photographs of Masaya volcano’s plume as seen from the space shuttle, with the active vent indicated by a star. The portion of the plume over the Llano Pacaya ridge is clearly less concentrated than other portions of the plume. Photos are from November 9, 1984 (top photo; NASA image STS 051a-32-066) and January 13, 1986 (bottom photo; NASA image STS061c-37-076). Images courtesy of the Image Science and Analysis Laboratory, NASA-Johnson Space Center.


74

prone to the development of a nocturnal temperature inversion (Delmelle et al., 2002),

which can interfere with the plume’s ability to propagate over the Llano Pacaya ridge,

sometimes forcing it to spread laterally along the windward side of the ridge. Heating of the

surface during daylight hours generally dissipates the capping inversion, which allows the

plume to clear the ridge. Depending on the height to which the plume rises, it will either be

near-ground-level and, therefore, be affected by orographic acceleration, or it will rise above

the envelope influenced by the acceleration and continue over the ridge unaffected. Smaller

topographic features may complicate the issue further, causing localized interaction with and

acceleration of the plume.

In general, acceleration factors up to near 2 are likely to be the cause of the apparent

loss of SO2 over the 10 km from the Ticuantepe road to the Llano Pacaya road. If this is the

case, the discrepancy will persist as long as one single plume speed is applied to

measurements made at different localities. A single value for plume speed when calculating

flux, whether it be a hand-held anemometer measurement of ground wind speed, a multi-

spectrometer measurement of plume speed or velocity, or a normalization to a plume speed

of 1 m/s (Zapata et al.,1997; Williams-Jones et al., 2006a) is not adequate at Masaya. While

the normalization approach remains valid for inter-volcano comparisons of fluxes at

volcanoes without topographic issues as at Masaya, and for comparing year-to-year data

from Masaya’s individual roads, it should not be used as a means to compare fluxes

measured on Ticuantepe to those measured on Llano Pacaya. Resulting normalized fluxes

will be skewed, with fluxes appearing misleadingly low over the Llano Pacaya road.

At volcanoes with complex topography and the possibility of topography-wind

interactions, plume speed measurements need to be conducted in the same locality as the

SO2 measurements themselves. Future studies at Masaya should include plume speed

measurements on both the Ticuantepe road and the Llano Pacaya road, with a total of at

least four spectrometers, in order to assess the difference in magnitude of plume speeds at

these locations. This will confirm the degree to which SO2 loss between the two roads is

apparent (due to the assumption of a uniform plume speed in calculations of SO2 flux), and

how much may be true loss of SO2 from the plume. In this study, only two spectrometers

were available for use. Further, it was also only in post-processing of the data that the

discrepancy in flux from one road to the next was recognized. Prior to this recognition of


75

topographic effects on apparent fluxes, it had been assumed that a single value for plume

speed was sufficient for any measurements regardless of location. In the interim, preliminary

estimates of the acceleration factors for the geography of the Masaya-Llano Pacaya region

indicate that the expected wind acceleration factors affecting Masaya’s plume are sufficiently

large to account for all of the 33-50% losses measured in this study.

The findings in this study have implications beyond field protocol for making SO2

flux measurements. SO2 fluxes from Masaya have been used in calculating worldwide

volcanic SO2 input to the atmosphere (Andres and Kasgnoc, 1998), the bulk of which were

measurements made on the Llano Pacaya road. True SO2 fluxes are conceivably twice those

used in such calculations, and as Masaya was already one of the largest non-eruptive emitters

of SO2, a twofold increase in SO2 from Masaya could significantly alter the estimated budget

of volcanic SO2. Similarly, remotely sensed SO2 fluxes are quoted with respect to the ‘excess

sulfur problem’ (e.g., Andres et al., 1991; Wallace et al., 2003). Volumes of SO2 released by

volcanoes are often more than can be accounted for by petrologic estimates of magmas’ pre-

eruptive sulfur contents. If erroneously low SO2 fluxes, as with this study’s Llano Pacaya

measurements, are referenced with respect to the ‘excess sulfur,’ true SO2 fluxes will

exacerbate the ‘excess sulfur problem.’

3.7 Conclusions

Three decades of SO2 flux measurements (1431 individual traverses) at Masaya

volcano, Nicaragua have consistently resulted in coefficients of variation of approximately

30% over the short time periods associated with individual field campaigns (Appendix B.2).

The 701 measurements from this study, while more extensive in number than previous

studies, exhibit the same natural variability. However, what became evident upon the

implementation of this more rigorous approach to SO2 monitoring was a significant

discrepancy between fluxes measured at two different distances downwind of Masaya’s

active vent. Fluxes measured on the Llano Pacaya road, ~15 km downwind of the vent,

were ~30-50% lower than those recorded on the Ticuantepe road, only ~5 km from Masaya.

In order to account for the apparent loss over 10 km, dry deposition of sulfur from

the plume, conversion of sulfur to sulfate aerosols, and topographic interactions were

considered. Previous estimates of losses to dry deposition were approximately 10%


76

(Bourque and Arp, 1996; Delmelle et al., 2001). The 10% estimate at Masaya is likely to be

even lower, as it was based on SO2 measurements made on the Llano Pacaya road, now

shown to exhibit fluxes less than those measured closer to the degassing vent. Accordingly,

dry deposition cannot be solely responsible for the measured SO2 loss seen between the two

roads. Likewise, the estimates of the plume’s aerosol content in this study do not indicate

sufficient variation to account for a 30-50% loss of SO2. Further, published rates of

conversion of SO2 to sulfate would lead to degrees of conversion much smaller than those

needed to explain the losses seen at Masaya. Most sulfate aerosols present in Masaya’s

plume are likely to be of near-source, high temperature origin (Allen et al., 2002; Mather et

al., 2003).

The complex topography surrounding the low-lying Masaya volcano is interpreted to

be at the root of the apparent losses of SO2 with distance downwind. Simple first-order

calculations and parallels drawn to meteorological field studies indicate that the Llano Pacaya

ridge is likely significantly modifying the persistent trade winds that have made Masaya an

ideal site for SO2 measurements. Up to 400 m higher than the summit of Masaya, the Llano

Pacaya ridge forces winds and, therefore, the entrained plume, to accelerate. Consequently,

the plume is diluted along its axis of propagation over the Llano Pacaya road relative to the

Ticuantepe road. This dilution results in smaller SO2 column thicknesses over Llano Pacaya,

and when these integrated column thicknesses are multiplied by either a standard,

normalizing plume speed of 1 m/s (e.g., Zapata et al., 1997) or a single measured plume

speed, it results in a misleading discrepancy between fluxes recorded on the Ticuantepe and

Llano Pacaya roads. As a consequence of the acceleration of local winds, Llano Pacaya

fluxes routinely appear 30-50% lower than those measured on Ticuantepe. Further study

involving simultaneous plume speed measurements on both roads will confirm that

acceleration of winds over the Llano Pacaya ridge is the cause of the apparent loss of SO2.

Distinct plume speed measurements for each distance downwind will remedy the problem,

with faster winds over Llano Pacaya compensating for the lower column thicknesses.

It is important to note that orographic acceleration of winds may be the case for all

volcanoes with low, tropospheric plumes and complex surrounding topography like that at

Masaya. In those instances, one must be wary of using one blanket plume speed value for all

data collected, as it can result in misleading variations within the SO2 flux dataset, either by


77

creating non-existent ‘losses’ or exacerbating the true discrepancies between values from

different measurement locations. Single plume speeds or normalizing factors may still be

useful in comparing year-to-year datasets for individual measurement locations, but should

not be used in comparing measurements from different distances downwind, as with the

Ticuantepe and Llano Pacaya roads. Use of erroneous fluxes of SO2 may result in flawed

estimations of both global volcanic SO2 emissions and volumes of degassed magma at

individual volcanoes.

3.8 Acknowledgements

This research was supported by an NSERC Discovery grant to Glyn Williams-Jones

and would not have been possible without the help of many people. We would like to thank

Keith Horton for help with the FLYSPECs; Norm O’Neill and John Porter for the use of

their Microtops II instruments; Lizzette Rodríguez and Matt Watson for aerosol processing

software; Annie Bérubé, Marianne Gagnon, Guillaume Mauri, Geneviève Pépin, and Kirstie

Simpson for help with field work; Doew Steyn for insight into the acceleration of wind over

topography; and John Stix, Diana Allen, and Kirstie Simpson for their input when reviewing

this chapter.


78

3.9 References

Allen, A.G., Oppenheimer, C., Ferm, M., Baxter, P.J., Horrocks, L.A., Galle, B., McGonigle, A.J.S., and Duffell, H.J., 2002. Primary sulfate aerosol and associated emissions from Masaya Volcano, Nicaragua. Journal of Geophysical Research, D, Atmospheres, 107(D23), 4682, doi:10.1029/2002JD002120

Andres, R.J., Rose, W.I., Kyle, P.R., deSilva, S., Francis, P., Gardeweg, M., and Moreno Roa, H., 1991. Excessive sulfur dioxide emissions from Chilean volcanoes. Journal of Volcanology and Geothermal Research, 46, 323-329.


Bourque, C. P.-A., and Arp, P.A., 1996. Simulating sulfur dioxide plume dispersion and subsequent deposition downwind from a stationary point source: a model. Environmental Pollution, 91(3): 363-380.

Branan, Y.K., Harris, A., Watson, I.M., Phillips, J.C., Horton, K., Williams-Jones, G., and Garbeil, H. Investigation of at-vent dynamics and dilution of gas puffing using thermal infrared radiometers at Masaya volcano, Nicaragua. Journal of Volcanology and Geothermal Research, submitted.

Casadevall, T.J., Johnston, D.A., Harris, D.M., Rose Jr., W.I., Malinconico, L.L., Stoiber, R.E., Bornhorst, T.J., Williams, S.N., Woodruff, L., and Thompson, J.M., 1981. SO2 emission rates at Mount St. Helens from March 29 through December, 1980. In: Lipman, P.W., and Mullineaux, D.R. (Editors), The 1980 Eruptions of Mount St. Helen’s. U.S. Geological Survey Professional Paper 1250: 193-200.

CCVG-IAVCEI, 2003. Eighth Field Workshop on Volcanic Gases, March 26 to April 1, 2003. In: The Commission on the Chemistry of Volcanic Gases & International Association of Volcanology and Chemistry of the Earth’s Interior, Nicaragua and Costa Rica, <http:// http://www.iavcei.org/>

Delmelle, P., Stix, J., Bourque, C.P.A., Baxter, P.J., Garcia-Alvarez, J., and Barquero, J., 2001. Dry deposition and heavy acid loading in the vicinity of Masaya Volcano, a major sulfur and chlorine source in Nicaragua. Environmental Science and Technology, 35: 1289-1293.

Delmelle, P., Stix, J., Baxter, J., Garcia-Alvarez, J., and Barquero, J., 2002. Atmospheric dispersion, environmental effects and potential health hazard associated with the low-altitude gas plume of Masaya volcano, Nicaragua. Bulletin of Volcanology, 64: 423-434.

Doukas, M.P., 2002. A new method from GPS-based wind speed determinations during airborne volcanic plume measurements. U.S. Geological Survey Open-File Report 02-395, pp 1-13.

Eatough, D.J., Caka, F.M., and Farber, R.J., 1994. The conversion of SO2 to sulfate in the atmosphere. Israel Journal of Chemistry, 34: 301-314.


79

Elias, T., Sutton, A.J., Stokes, J.B., and Casadevall, T.J., 1998. Sulfur dioxide emission rates of Kilauea Volcano, Hawaii, 1979-1997. U.S. Geological Survey Open-File Report 98-462.

Elias, T., Sutton, A.J., Oppenheimer, C., Horton, K.A., Garbeil, H., Tsanev, V., McGonigle, A.J.S., and Williams-Jones, G., 2006. Intercomparison of COSPEC and two miniature ultraviolet spectrometer systems for SO2 measurements using scattered sunlight. Bulletin of Volcanology, 68: 313-322.

Favalli, M., Mazzarini, F., Pareschi, M.T., and Boschi, E., 2004. Role of local wind circulation in plume monitoring at Mt. Etna volcano (Sicily): Insights from a mesoscale numerical model. Geophysical Research Letters, 31, L09105, doi:10.1029/2003GL019281.

Girard, G., and van Wyk de Vries, B., 2005. The Managua Graben and Las Sierras-Masaya volcanic complex (Nicaragua); pull-apart localization by an intrusive complex: results from analogue modeling. Journal of Volcanology and Geothermal Research, 144: 37-57.

Graziani, G., Martilli, A., Pareschi, M.T., and Valenza, M., 1997. Atmospheric dispersion of natural gases at Vulcano island. Journal of Volcanology and Geothermal Research, 75: 283-308.

Horrocks, L., Burton, M., Francis, P., and Oppenheimer, C., 1999. Stable gas plume composition measured by OP-FTIR spectroscopy at Masaya Volcano, Nicaragua, 1998-1999. Geophysical Research Letters, 26(23): 3497-3500.


Mather, T.A., Allen, A.G., Oppenheimer, C., Pyle, D.M., and McGonigle, A.J.S., 2003. Size-Resolved Characterisation of Soluble Ions in the Particles in the Tropospheric Plume of Masaya Volcano, Nicaragua: Origins and Plume Processing. Journal of Atmospheric Chemistry, 46: 207-237.

Mather, T.A., Tsanev, V.I., Pyle, D.M., McGonigle, A.J.S., Oppenehimer, C., and Allen, A.G., 2004. Characterization and evolution of tropospheric plumes from Lascar and Villarrica volcanoes, Chile. Journal of Geophysical Research, 109, D21303, doi:10.1029/2004JD004934.


McGonigle, A.J.S., Delmelle, P., Oppenheimer, C., Tsanev, V.I., Delfosse, T., Williams-Jones, G., Horton, K., and Mather, T.A., 2004. SO2 depletion in tropospheric volcanic plumes. Geophysical Research Letters, 31, L13201, doi:10.1029/2004GL019990.


80

McGonigle, A.J.S., Hilton, D.R., Fischer, T.P., and Oppenheimer, C., 2005a. Plume velocity determination for volcanic SO2 flux measurements. Geophysical Research Letters, 32, L11302, doi:10.1029/2005GL022470.

McGonigle, A.J.S., Inguaggiato, S., Aiuppa, A., Hayes, A.R., and Oppenheimer, C., 2005b. Accurate measurement of volcanic SO2 flux: Determination of plume transport speed and integrated SO2 concentration with a single device. Geochemistry Geophysics Geosystems, 6(1), Q02003, doi:10.1029/2004GC000845.

Morys, M. F., Mims III, F.M., and Anderson, S. E., 1996. Design, calibration, and performance of MICROTOPS II hand-held ozonometer. 12th International Symposium on Photobiology, Vienna, Austria, International Congress on Photobiology. <http://www.solar.com/ftp/papers/mtops.pdf>

Oke, T. R., 1987. Boundary Layer Climates, Routledge, 2nd edn., Methuen, New York, 435 pp.

Oppenheimer, C., Francis, P., and Stix, J., 1998. Depletion rates of sulfur dioxide in tropospheric volcanic plumes. Geophysical Research Letters, 25(14): 2671-2674.

Porter, J.N., Miller, M., Pietras, C., and Motell, C., 2001. Ship-based Sun Photometer Measurements Using Microtops Sun Photometers. Journal of Atmospheric and Oceanic Technology, 18: 765-774.


Rodríguez, L.A., Watson, I.M., Rose, W.I., Branan, Y.K., Bluth, G.J.S., Chigna, G., Matías, O., Escobar, D., Carn, S.A., and Fischer, T.P., 2004. SO2 emissions to the atmosphere from active volcanoes in Guatemala and El Salvador, 1999-2002. Journal of Volcanology and Geothermal Research, 138: 325-344.




Stoiber, R.E., Malinconico, L.L., and Williams, S.N., 1983. Use of the correlation spectrometer at volcanoes. In: Tazieff, H. and Sabroux, J.C. (Editors), Forecasting Volcanic Events. Elsevier, New York: 424-444.



81

Taylor, P.A., and Lee, R.J., 1984. Simple guidelines for estimating wind speed variations due to small scale topographic features. Climatological Bulletin, 18: 3-32.

Taylor, P.A., and Teunissen, H.W., 1985. The Askervein Hill Project: Report on the Sept./Oct. 1983 Main Field Experiement. Environment Canada Atmospheric Environment Service, Research Report MSRB-84-6, 314 pp.

Wallace, P.J., Carn, S.A., Rose, W.I., Bluth, G.J.S., and Gerlach, T., 2003. Integrating Petrologic and Remote Sensing Perspectives on Magmatic Volatiles and Volcanic Degassing. EOS, Transactions of the American Geophysical Union, 84(42), 441, 446-447.

Watson, I.M., and Oppenheimer, C., 2000. Particle size distributions of Mount Etna’s aerosol plume constrained by Sun photometry. Journal of Geophysical Research, 105: 9823-2829.

Watson, I.M., and Oppenheimer, C., 2001. Photometric observations of Mt. Etna’s different aerosol plumes. Atmospheric Environment, 35: 3561-3572.


Williams-Jones, G., Horton, K.A., Elias, T., Garbeil, H., Mouginis-Mark, P.J., Sutton, A.J., and Harris, A.J.L., 2006a. Accurately measuring volcanic plume velocity with multiple UV spectrometers. Bulletin of Volcanology, 68: 328-332.

Williams-Jones, G., Stix, J., and Nadeau, P.A., 2006b. Using the COSPEC in the field. In: Stix, J., Williams-Jones, G., and Hickson, C. (Editors), The COSPEC Cookbook: Making SO2 Gas Measurements at Active Volcanoes, Proceedings of Volcanology, IAVCEI. In press.

Witham, C.S., Oppenheimer, C., and Horwell, C.J., 2005. Volcanic ash-leachates: a review and recommendations for sampling methods. Journal of Volcanology and Geothermal Research, 141: 299-326.

Zapata G., J.A., Calvache V., M.L., Cortés J., G.P., Fischer, T.P., Garzon V., G., Gómez M., D., Narváez M., L., Ordóñez V., M., Ortega E., A., Stix, J., Torres C., R., and Williams, S.N., 1997. SO2 fluxes from Galeras Volcano, Colombia, 1989-1995: Progressive degassing and conduit obstruction of a Decade Volcano. Journal of Volcanology and Geothermal Research, 77: 195-208.

Chapter 4 – Environmental Effects of Persistent Volcanic Degassing

82

Chapter 4: A MULTI-PARAMETER EVALUATION OF THE

ENVIRONMENTAL EFFECTS OF TWO LOW-LYING, PERSISTENTLY DEGASSING VOLCANOES: MASAYA, NICARAGUA AND POÁS, COSTA RICA

*To be submitted to Geochemistry, Geophysics, Geosystems (G3)

4.1 Abstract

Investigation of the relationship between elevation and the health of vegetation, as

determined by Landsat NDVI data at Masaya volcano, Nicaragua, confirms previous

observations that severe vegetation devastation in the vicinity of Masaya is most evident at

high elevations on the Llano Pacaya ridge downwind of the volcano. The acidic plume

released from the low-altitude volcano routinely fumigates the highlands, resulting in the

documented vegetation kill. At Poás volcano, Costa Rica, a diffuse plume of SO2 mantles

the topography adjacent to the vent, also causing significant destruction of the native cloud

forest. While the most severe vegetation kill at both volcanoes appears restricted to the

highest elevations, previous studies have shown that effects of the plume at Masaya are

found beyond the area identified from NDVI data, suggesting that NDVI is sufficient only

in estimating a minimum affected zone. More subtle effects, likely related to the complex

interplay of variables such as humidity, stomatal conductance, and plume composition, exist

and are not detected by NDVI. Further study is necessary to fully ascertain the extent of the

environmental effects of low-lying volcanic plumes like those at Masaya and Poás.

4.2 Introduction

Sulfur dioxide emissions have widely been recognized as a significant contributor to

the detrimental effects of industrial pollution. Numerous studies (e.g., Likens and Bormann,

1974; Williams et al., 1977; Kozlowski, 1980) have been conducted to assess the types and

extents of the environmental effects of the release of SO2 to the atmosphere. Active

volcanoes may also act as point-sources for sizeable emissions of SO2. Many attempts have


83

been made to quantify SO2 emissions from volcanoes (e.g., Stoiber et al., 1986; Casadevall et

al., 1994; Elias et al., 1998; Williams-Jones et al., 2003). Large-scale eruptions that inject

significant quantities of SO2, ash, and aerosols into the stratosphere have also been studied

in order to evaluate their impacts on global climate change (e.g., Pollock et al., 1976; Luhr,

1991; Hansen et al., 1992; Gerstell et al., 1993). Industrial emission of SO2 into the

troposphere and large-scale volcanic eruptions to the stratosphere represent the endpoints of

a continuum of sorts; intermediate to the two aforementioned cases are low-lying,

persistently degassing volcanoes that, in the absence of significant eruptive activity,

consistently emit large quantities of SO2 to the troposphere. Recent assessments of global

volcanic SO2 emissions have shown that passive emissions exceed the amount of SO2

released in acute eruptive phases (Andres and Kasgnoc, 1998). Yet, until recently, a paucity

of data existed on effects of tropospheric volcanic emissions, especially in cases of volcanoes

with complex topography that lends itself to persistent fumigation of the surrounding areas.

A persistently degassing basaltic shield, Masaya volcano, Nicaragua, undergoes cycles

of intense degassing in the absence of significant eruptive activity (Stoiber et al, 1986; Rymer

et al., 1998; Williams-Jones et al., 2003; Smithsonian Institution, 1981-2006; Chapter 3).

Substantial evidence has suggested that the volcano’s plume consistently fumigated the

highlands downwind of the active vent, leading to widespread devastation of vegetation (e.g.,

McBirney, 1956; Johnson and Parnell, 1986; Delmelle et al., 2002). Delmelle et al. (2001;

2002) made measurements of ground-level SO2 concentrations as well as the dry deposition

velocities of S, Cl, and F in the region downwind of Masaya, noting that the most severe

effects on vegetation were restricted to areas with the highest elevations. Further studies at

Masaya have investigated the effects of the fumigation on the soils downwind (Delmelle et

al., 2003; Delfosse et al., 2005a, b). These studies identified the chemical consequences of

the plume’s fumigation, and Delmelle et al. (2001; 2002) specifically highlighted the

detrimental effects of the degassing on vegetation, especially at high elevations, largely

through qualitative methods. This study evaluates the fumigation and incipient

environmental effects at Masaya, focusing on the effects of elevation, through the use of

remote sensing, specifically Landsat Enhanced Thematic Mapper Plus (ETM+) derived

Normalized Vegetation Indices (NDVIs) and a digital elevation model (DEM) generated

from digital photogrammetry. Subsequent application of similar remote sensing

methodology to Poás volcano in Costa Rica, allows for comparison of the effects of


84

persistent volcanic degassing at a humid volcano with dispersed fumarole fields (Poás) and a

low-lying, dry volcano with a single degassing vent (Masaya).

4.3 Background

The Masaya volcanic complex, situated approximately 20 km southeast of Managua,

Nicaragua (Figure 4-1), lies on the Central American Volcanic Front (CAVF), the result of

the subduction of the Cocos Plate beneath the Caribbean Plate. Unlike the many

stratovolcanoes that comprise the CAVF (e.g., Arenal, Costa Rica; Concepcion, Nicaragua),

Masaya consists of a 6 km by 11 km caldera in which sits a series of basaltic cones and pit

craters. The two largest cones are Masaya and Nindiri. Nindiri comprises three craters: the

main Nindiri crater, the site of an old lava lake, and two secondary pit craters, San Pedro and

Santiago (Figure 4-2). Santiago crater is currently the site of Masaya’s active vent. Extrusive

activity at Masaya is rare; the last lava flows were produced in 1772 as a result of fissure

vents on the flanks of Masaya cone (Rymer et al., 1998). Small vent-clearing explosions,

with small amounts of juvenile material, have occurred more recently (March, 2005;

Smithsonian Institution, 2006a). In the absence of significant explosive and extrusive events,

the predominant activity at Masaya has been varying degrees of passive degassing, usually

from a single vent in Santiago crater (Figure 4-3). Over the course of the most recent gas

crisis, which began in 1993 and has persisted to the time of this writing, average daily SO2

fluxes (normalized to a plume speed of 1 m/s) have ranged from 45 t/d to 312 t/d

(Appendix B). The most recent estimates, as of March 2006, are average SO2 emissions of

approximately 1240 ± 440 metric tonnes/day (at true plume speeds) as measured ~5 km

from the active vent (Chapter 3; Appendix B).

While many volcanoes emit SO2 at high altitude above the atmospheric boundary

layer, thus limiting the effects the plume may have on the local environment, certain

volcanoes, such as Masaya, degas into the boundary layer, where the emissions are far more

likely to have a significant impact on the vegetation of the surrounding region. Further,

Masaya caldera and its currently active vent are located within a larger-scale caldera, the Las

Sierras caldera (Bice, 1980; van Wyk de Vries, 1993; Sebesta, 1997). The western rim of the

Las Sierras caldera is approximately 15 km to the west of Masaya’s active vent and rises to

elevations (maximum elevation ~930 m a.s.l) exceeding those of Masaya’s summit


85

Figure 4-1: Geographic map of western Nicaragua showing the location of Masaya volcano with respect to other Nicaraguan volcanoes.


86

Figure 4-2: Relief image of Masaya Caldera. Volcanic cones within the caldera are labeled in red, while craters are labeled in yellow. UTM zone 16n.


87

Figure 4-3: The actively degassing vent in Santiago crater. Vent’s diameter is estimated to be approximately 30-40 m.


88

(~600 m a.s.l.). Unlike volcanoes that emit plumes unimpeded into the upper atmosphere

(e.g., Popocatépetl volcano, Mexico), Masaya’s gas plume is routinely blown by easterly trade

winds towards the rim of the Las Sierras caldera, the Llano Pacaya ridge. As a result, the

plume interacts directly with the land surface at Llano Pacaya rather than just the

atmosphere. Field observations have identified a marked ‘kill-off’ zone along the ridge,

where vegetation is reduced to little more than burnt, dry grasses (Figure 4-4).

Garrec et al. (1984) conducted one of the first studies of the effects of Masaya’s

persistent degassing on vegetation in the region, noting necrotic leaf damage on a variety of

crops and native plants. Analyses of various plants sampled at 5 km and 15 km downwind

of the active vent revealed fluoride concentrations in plant leaves an order of magnitude

greater than expected natural levels. Sulfur concentrations, however, despite their greater

abundance in the volcanic plume, did not differ notably from natural sulfur contents of

plants. More recent studies (Delmelle et al., 2001; 2002) measured ground level

concentrations of SO2 and dry deposition velocities of SO2 and HCl to constrain the lateral

dispersion of the plume. SO2 concentrations were determined through the use of a network

of 55 diffusion samplers distributed over a region up to 44 km downwind of Masaya. Dry

deposition was measured utilizing a similar spatial network, substituting lead dioxide-treated

sulfation plates. Results indicated that 1.5 x 108 g SO2 (<10% of total daily volcanic SO2

emissions) and 5.7 x 107 g HCl were deposited daily over a 1250 km2 area to the southwest

of the degassing vent. SO2 concentrations above background levels were detected over a

similar area. The greatest degree of vegetation devastation was found to be within a 22 km2

zone, which corresponded to a region that coupled high altitude with high SO2

concentrations. Sulfur deposition affects not only vegetation, but the soils of the area as

well; Delmelle et al. (2003) conducted a survey of the Andosol soils in the Masaya region,

finding that fumigation of the region by Masaya’s plume causes significant soil acidification

by SO2, HCl, and HF. Delfosse et al. (2005a, b) concluded that much of the volcanic sulfur

impacting the more weathered soils of the Llano Pacaya region is sequestered from the soil

in the form of aluminum-hydroxy-sulfate minerals, which serve to prevent acidification to

the extent found closer to the volcano in the Ticuantepe region.

For comparison to Masaya volcano, this study also evaluates the relationship

between degassing, topography, and vegetation kill at Poás volcano, Costa Rica (Figure 4-5).


89

Figure 4-4: (a) and (b) Views of the Llano Pacaya ridge looking west from the Ticuantepe road. (a) is a clear view in which the plume is passing unimpeded over the ridge. (b) depicts an instance of intense fumigation of the ridge by the plume such that the ridge is barely visible. (c) the landscape in the devastated region of the Llano Pacaya ridge, looking southeast. Note the lack of vegetation beyond dry grasses and a few small shrubs.


90

Located in the Cordillera Central of Costa Rica at an elevation of 2708 m a.s.l., Poás is a

stratovolcano of basaltic to dacitic composition. Activity at Poás is generally characterized

by sporadic phreatic or phreatomagmatic eruptions from the active crater (Figure 4-6).

Recent activity had been restricted to passive degassing from dispersed fumaroles and the

acidic crater lake, Laguna Caliente (Smithsonian Institution, 2003; Figure 4-7) until the first

phreatic activity in 14 years in March, 2006 (Smithsonian Institution, 2006b). While Poás

and Masaya share the common characteristic of persistent degassing leading to vegetation kill

(Nicholson et al., 1996; Sandoval et al., 1996), Poás is quite different from Masaya. In

contrast to Masaya, Poás is significantly higher in elevation, located in a humid climate, and

vegetated by a thick cloud forest. Further, whereas Masaya emits a large plume from one

central crater, Poás’ plume is dispersed, and SO2 fluxes are much lower than those at Masaya

(52 ± 34 t/d in 2001; G. Williams-Jones, pers. comm., 2006), as the Laguna Caliente acts as a

buffer, scavenging most of the acidic emissions before they are released to the atmosphere

(Fournier et al., submitted). At Masaya, the plume is relatively unimpeded until it encounters

rises in the topography at locations approximately 15 km from the vent. In contrast, the

plume at Poás mantles the ground surface immediately adjacent to the crater and only lofts

freely above the ground when the topography drops away significantly. The different

climates, vegetation, degassing styles, and topographic geometry at Poás and Masaya will aid

in the characterization of the environmental effects of persistently degassing volcanoes in a

ranges of physical settings.

4.4 Methodology

To examine the vegetation kill at Masaya and Poás, a variety of datasets were

employed. In total, three Landsat ETM+ scenes were utilized over the course of this

particular study. Two scenes of Masaya volcano (path 017, row 052) from March 25, 2001

and April 18, 2004 were obtained from the Global Land Cover Facility (GCLF) and

Landsat.org, respectively (Figure 4-8a, b). One scene of Poás volcano (path 015, row 053)

from December 1, 2003 was purchased from the United States Geological Survey (Figure

4-8c). Each image was radiometrically and geometrically corrected prior to our acquiring it.

All scenes were chosen for minimal cloud cover as well as their acquisition dates. Images

from March and April coincide with the dry season at Masaya, which is also the time of year


91

Figure 4-5: Geographic map of northwestern Costa Rica showing the location of Poás volcano with respect to other Costa Rican volcanoes.


92

Figure 4-6: Map of the summit area of Poás volcano. UTM zone 16n.

Reprinted from the Journal of Volcanology and Geothermal Research, Vol. 97, H. Rymer, J.

Cassidy, C.A. Locke, M.V. Barboza, J. Barquero, J. Brenes, and R. Van der Laat, “Geophysical studies of the recent 15-year eruptive cycle at Poás Volcano, Costa Rica,” Pages 425-442,

Copyright (2000), with permission from Elsevier.


93

Figure 4-7: The active crater of Poás volcano, with Laguna Caliente and fumarole fields. Diameter of Laguna Caliente is approximately 300-350 m.

94

Figure 4-8: Composite images of Landsat ETM+ red, green, and blue bands for each of the three scenes utilized, with important features labeled. (a) 2001 scene of Masaya; (b) 2004 scene of Masaya; and (c) Poás scene. Star in (a) and (b) indicates Masaya’s active vent.

a.)

Chapter 4 – Environm

ental Effects of Persistent Volcanic Degassing

95

b.)



96

c.)




97

when other ground-based geochemical datasets were collected. The 2001 scene of Masaya

was the primary data source evaluated; the second Masaya scene was used for a temporal

comparison at Masaya, and the Poás scene was used as a means to evaluate whether or not

the relationships deduced from Masaya’s NDVI data could be applied to other volcanoes.

The 2001 Landsat image of Masaya was downloaded from the GCLF as GeoTIFF

data, and was subsequently imported into the raster GIS program, IDRISI Kilimanjaro

(Clark Labs, 2004). As the scene consisted primarily of the Pacific Ocean, the image was

cropped to include only Masaya Caldera, the Llano Pacaya Ridge, and the area further

downwind, up to the Pacific Ocean, ultimately a total area of approximately 55 km x 40 km.

A Normalized Differential Vegetation Index (NDVI) was constructed from the red (Landsat

ETM+ band 3; 0.63-0.69 µm) and near-infrared, or NIR, (Landsat ETM+ band 4; 0.75-

0.90 µm) bands of the data (Rouse et al., 1973). NDVI is sensitive to chlorophyll content

and, therefore, is representative of ‘greenness,’ or the relative health of vegetation. Of the

numerous vegetation indices that exist, NDVI was chosen because of its wide use relative to

other ratio-based indices and its exploitation of vegetation’s high reflectance in the NIR

portion of the electromagnetic spectrum rather than the green portion, as is the case with

some other indices (Tucker, 1979). Further, NDVI is the vegetation index best suited for

conditions of high substrate and vegetation heterogeneities as is the case at both Masaya and

Poás (Lawrence and Ripple, 1998), and is applicable even in semi-arid regions like Masaya

(Anderson et al., 1993). NDVI is calculated as follows:

NDVI = NIR - RED / NIR + RED [4-1]

Because unhealthy, dead, or damaged vegetation is less ‘green’ than its healthy counterparts

and lacks chlorophyll, which is highly reflective in the NIR portion of the electromagnetic

spectrum, the damaged vegetation will have NDVI values much lower than the values for

healthy vegetation. Values range from -1 for dead vegetation or other barren surface

features such as water, ice, or clouds, to 1 for healthy, lush, green vegetative cover.

The IDRISI program contains a simple NDVI command, into which the file names

for bands 3 and 4 are entered. The output is the NDVI image itself. This process was used

for each of the three Landsat scenes, with the second Masaya scene having been cropped to

match the dimensions of the 2001 image, and the Poás data cropped to an area sufficient to


98

examine the crater and surrounding vegetation kill area. Because vegetation is generally

greener, healthier, and more abundant at higher elevations near Masaya, owing to cooler

temperatures and greater precipitation, linear regressions of the dependence of NDVI on

elevation were determined using IDRISI:

2001 NDVI = -0.162380 + 0.000448 * Elevation [4-2]

2004 NDVI = -0.167488 + 0.000413 * Elevation [4-3]

In order to isolate the effects of SO2 fumigation on NDVI, each regional trend was then

subtracted from the Landsat-derived NDVI. A regression was not performed using the data

from Poás, as the size of the area of interest is significantly smaller than that at Masaya. Poás

is also located within a cloud forest, which displays largely uniform vegetative health for the

area concerned.

A digital elevation model (DEM) with an initial resolution of 80 m, generated from

digital photogrammetry surveys of the Masaya region of Nicaragua, was obtained from the

Instituto Nicaragüense de Estudios Territoriales (INETER). To compensate for gaps in the

original dataset, the DEM was re-sampled to a resolution of 96 m in Golden Software’s

Surfer program using a radial basis function gridding method. Following re-sampling, the

DEM was exported to IDRISI for incorporation into the NDVI dataset. The DEM was

cropped to the same dimensions as those of the cropped Landsat data (Figure 4-9).

A DEM of Costa Rica was generated from a Shuttle Radar Topography Mission

(SRTM) image obtained from the GLCF. The DEM was originally downloaded in .TIFF

format with a resolution of 90 m. After importation into IDRISI, the data were re-sampled

to a resolution of 96 m to compensate for gaps in the DEM and to coincide with the

resolution of the Masaya dataset. The DEM was also cropped to include only the region of

interest surrounding the degassing crater at Poás (Figure 4-10).

Both S dry deposition and SO2 ground level concentration data for Masaya were

used from previous studies (Delmelle et al., 2001; 2002). Data points for each data set were

imported into Golden Software’s Surfer program and converted into a contoured,

continuous image by means of a radial basis function gridding method. Given the irregular

nature of the topography at Masaya, smooth contours that resulted for the two variables are

not likely in reality; rather, the contoured surfaces presented herein are meant to be general

99

Figure 4-9: Digital elevation model of Masaya region. Masaya caldera is outlined in red for reference.



100

Figure 4-10: Digital elevation model of Poás region. Laguna Caliente is outlined in light blue for reference.




101

representations of the plume’s aerial coverage. Once the full extent of each dataset had been

approximated in Surfer, the files were imported into IDRISI using the SRFIDRIS function

and manipulated by means of cropping and re-sampling in order to correspond to the

dimensions and resolution of the NDVI and DEM images (Figure 4-11). SO2 UV-

spectrometer traverse data (Chapter 3; Appendix B) were also digitized to provide additional

data in determining the spatial distribution of the areas subjected to the most concentrated

portions of the plume (Figure 4-12).

Population density data for Nicaragua were obtained from the Center for

International Earth Science Information Network (CIESIN), Columbia University, and

Centro Internacional de Agricultura Tropical (CIAT). Data comprised values for 2000 in

persons per km2, with a scaling factor such that the associated total population was

representative of United Nations totals for Nicaragua in 2000. The data were initially in

ASCII grid format, with latitude/longitude coordinates and a grid resolution of 2.5 arc-

minutes. The file was imported into IDRISI and, subsequently, projected into UTM

coordinates with a 96 m resolution in order to coincide with other datasets.

To evaluate the datasets, linear sampling profiles were digitized for both the Masaya

and Poás regions. A total of 8 profiles were created for Masaya, each beginning at the

degassing crater and radiating across the area over which the gas plume most often disperses.

Profiles ranged from 29 km to 36 km in length (Figure 4-13). Five profiles were created to

parallel the longitudinal axis of the kill-off zone at Poás; profiles radiating from a central

point were not utilized because much of the degassing at Poás is generated from scattered

fumaroles rather than one single vent. The five Poás profiles were each approximately

10 km long (Figure 4-14). Values were sampled along each profile for the NDVI, NDVI

residuals, DEM, dry deposition, and ground-level SO2 concentration data layers.

4.5 Results

Results from the NDVI transformations on each of the 3 Landsat data sets are

displayed in Figure 4-15. Residual results from the subsequent subtraction of regional trends

from Masaya’s NDVI datasets are shown in Figure 4-16. NDVI values ranged from -0.74 to

0.69 in 2001, and from -0.95 to 0.88 in 2004, at Masaya, and from -0.68 to 0.60 at Poás.

102

Figure 4-11: Interpolated distributions of S dry deposition and SO2 ground level concentration derived from data in Delmelle et al. (2001, 2002). (a) S dry deposition. (b) SO2 ground level concentration. Shaded DEM and caldera (outlined in red) are shown for reference.

a.)



103

b.)



104

Figure 4-12: Average SO2 concentration path-lengths over Ticuantepe and Llano Pacaya roads in 2005 and 2006. Color scheme is relative; data from the two roads were averaged separately such that a color on one road is not indicative of the same concentration path-length on the other. Shaded DEM surface and red caldera outline are shown for reference.



105

Figure 4-13: Location of profiles used for dataset sampling on Masaya data. Eight profiles originating from the active vent are shown in yellow. Also shown is a light blue ‘control’ profile, used to examine data relationships in an area relatively unaffected by the plume. Shaded DEM surface and red caldera outline are shown for reference.



106

Figure 4-14: Location of profiles used for dataset sampling on Poás data. Five parallel profiles originating in an unaffected area upwind of the active crater are shown in yellow. Shaded DEM surface and light blue Laguna Caliente outline are shown for reference.



107

Figure 4-15: Results of the NDVI transformations for each of the three scenes utilized, with Masaya caldera outlined in red for reference. (a) 2001 Masaya NDVI (b) 2004 Masaya NDVI (c) Poás NDVI.

a.)



108

b.)



109

c.)



110

Figure 4-16: Residual results of the subtraction of regional elevation-dependent trends from NDVI at Masaya, with caldera outlined in red for reference. (a) 2001 Masaya residuals (b) 2004 Masaya residuals

a.)



111

b.)




112

Both Masaya images display low values of NDVI within the caldera, as well as throughout

the regions of lowest elevation in the area. Values are higher along the Llano Pacaya ridge

except for the portion of the ridge in the path of Masaya’s plume. At Poás, NDVI values are

relatively low in the area downwind of the crater, but are high in all other regions.

Comparison of the NDVI images for Masaya from 2001 and 2004 fails to identify

any significant temporal changes. The extent and degree of plume-damaged vegetation is

similar for both years, though the 2004 image indicates the widespread persistence of

relatively unhealthy vegetation to higher elevations on the scale of the entire Llano Pacaya

ridge. This may be due to the date of the image’s acquisition by the satellite, which is a

month farther into Nicaragua’s dry season than the 2001 scene. Subtraction of the 2004

residual image from the 2001 residual image yields a difference of 0.005, possibly indicating

slightly healthier vegetation in 2001. However, the 0.005 difference is well within the

standard deviation of the mean residual value for each year (mean NDVI residual of -0.008

with a standard deviation of 0.126 for 2001; mean NDVI residual of -0.013 with a standard

deviation of 0.135 for 2004). There may, in fact, be real change over the three-year period;

however, there is an absence of data concerning precipitation, temperature, and other

variables which may have affected the vegetation. The NDVI data would need to be

adjusted in order to account for the variations in the conditions external to the plume’s

fumigation in order to fully assess a change in vegetation health as a result of the plume.

Further, a dataset consisting of only two dates is not an adequate time-series for determining

temporal changes. Cloud-free Landsat scenes of Masaya on which to perform NDVI

transformations are rare, and the accumulation of scenes for a larger time-series was not

possible. The assessment of change between the 2001 and 2004 NDVI residuals, therefore,

was based on visual assessment and the negligible difference between the mean values of the

two years’ datasets.

Given the apparent lack of significant variation between 2001 and 2004, only the

2001 scene from Masaya was analyzed further. Supporting this is the examination of Figure

4-12, which illustrates the relatively small spatial sector over which the plume generally

varies. The trade winds that prevail at both volcanoes are very consistent, confining each

plume to the same area year after year. This is evident, especially at Poás, where the

landscape varies from lush cloud forest to complete devastation by SO2 over a scale of only


113

tens of meters; this can also be deduced from the NDVI data, where adjacent pixels can vary

by as much as 0.7. As such, despite the more diffuse boundaries of affected areas at Masaya,

the constant fumigation in approximately the same location over a number of years would

not allow for any vegetation to recover from previous damage. While the reappearance of

vegetation, even within the active crater, after the degassing crisis of the 1980’s was

documented within approximately 18 months of the preceding drop in SO2 flux

(Smithsonian Institution, 1989-1991), the consistently high levels of degassing since 1993

(Chapter 3, Appendix B) would have either left vegetation at a relatively constant state of

devastation between 2001 and 2004, or perhaps caused the vegetation to be further adversely

affected. The lack of change between the 2001 and 2004 NDVI residual images suggests

that the former situation is the case, though this could be evaluated more closely with a

larger time-series of Landsat data.

Based on the 2001 NDVI residuals at Masaya, an area of 35 km2 was determined to

be the area of most severe vegetation kill (Figure 4-17), slightly larger than the 22 km2

identified by Delmelle et al. (2002) through air photo analysis. However, the areas over

which the effects of the plume exist, based on dry deposition and ground-level SO2

concentration data, are far more extensive, with corresponding areas of 1284 km2 and

1466 km2, respectively (Figure 4-11). These values are also somewhat larger than the

~1250 km2 for both variables suggested by Delmelle et al. (2001, 2002). At Poás, the zone

of most extensive damage to vegetation was found to be only 6 km2 (Figure 4-15c). In the

absence of extensive ground-based chemical data at Poás, it is impossible to discern the area

over which the plume’s presence may persist. Preliminary studies of the gas plume at Poás

indicate a reach of the plume much greater than that indicated simply by vegetation kill

(Nicholson et al., 1996; Sandoval et al., 1996), as is the case at Masaya.

At both Poás and Masaya, there appear to be elevation thresholds to above which

most of the vegetation kill, as detected by NDVI, is restricted. In the case of Poás, little

marked unhealthiness in vegetation exists below 2000 m a.s.l. (Figure 4-10, Figure 4-15c). At

Masaya, the zone of vegetation kill-off exists in two discrete locations. In the El Panama

region (Figure 4-8c), only a few kilometers from the degassing vent, the most damaged

vegetation lies above 500 m a.s.l., whereas on the Llano Pacaya ridge, vegetation appears

largely unaffected up to elevations of 700 m a.s.l. (Figure 4-9, Figure 4-16a, b). An exception

114

Figure 4-17: The area exhibiting the most severe degree of vegetation damage at Masaya, as determined from NDVI residual data from 2001 (background image shown here). The prominent zone of vegetation kill is delineated by the striped area, and the red caldera outline is included for reference.




115

is a relatively narrow strip of damaged vegetation that extends down from the Llano Pacaya

to approximately 400 m a.s.l. on the leeward side of the ridge, which may be a result of lee

eddies causing turbulent downward mixing of the plume as it accelerates over the rise in

topography (Chapter 3; Oke, 1987).

Results of the profiles from the DEM and NDVI (or residuals) reinforce the

assessment that the most severe vegetation kill is restricted to high altitudes at both

volcanoes (Figure 4-18). Each of Masaya’s numbered profiles begins at the degassing

Santiago crater and extends for approximately 35 km to the west. On each profile, though

noisy, the NDVI residual data average near zero (as on the control profile, Figure 4-18b),

with the exception of three distinct areas that display marked negative departures from the

regional trend. The crater itself, comprising mostly barren lava flows is the first. The

second two segments of the profiles with negative residuals are clearly the El Panama and

Llano Pacaya highlands. The degree of drop in NDVI varies from profile to profile, with

the values appearing to remain closer to the average of zero on the more southern profiles,

where the relief of the Llano Pacaya ridge is not so extreme.

The five profiles of NDVI and elevation at Poás display many of the same

characteristics as those at Masaya. As with the control ridge at Masaya, which exhibits an

NDVI dependence on elevation only, Poás profiles 1 and 5, adjacent to the main area of

vegetation kill, remain near-constant at ‘healthy’ NDVI values regardless of elevation.

Profiles 2, 3, and 4 show relatively high NDVI values near 0.3 in the area upwind of the

crater, but values drop drastically in the vicinity of the crater and persist at constant, low

levels until the elevation of the ground surface drops below ~2000 m a.s.l., at which point

NDVI values return to near 0.3. The extremely low NDVI near -0.7 in profile 3 is due to

pixels representing Laguna Caliente; water reflects very little radiation in the near-infrared

portion of the electromagnetic spectrum. Therefore, very negative values of NDVI are

characteristic of any body of water and do not relate to vegetation kill.

4.6 Discussion

Simple observation of the NDVI residuals from 2001 indicates that the area of most

extreme vegetation kill coincides well with the area identified by Delmelle et al. (2002)

through the use of air photo analysis (Figure 4-17). The zone is largely confined to the

116

Figure 4-18: Results of profiles of DEM and NDVI residual (or NDVI, as at Poás) datasets. Elevation profiles are shown by a black line, and NDVI data by a gray line. See Figures 4-13 and 4-14 for location of profiles. All profiles are shown with the eastern end at the origin. (a) 8 Masaya profiles (b) Masaya control profile (c) 5 Poás profiles

a.)



117



118



119



120

b.)

c.)



121



122




123

regions of highest altitude in the path of the plume’s propagation, as noted by Delmelle et al.

(2001; 2002) and confirmed by the inspection of the profiles of NDVI residuals and

elevation. Compared to the general trend of NDVI increasing with elevation, the top of the

Llano Pacaya ridge exhibited a marked departure from the trend on all of the profiles,

though to differing degrees depending on the distance from the average location of the

center of the plume. The effect of elevation is even more pronounced when the relationship

between elevation and NDVI is examined (Figure 4-19). A positive relationship is the

general trend in the region, with vegetation largely becoming healthier at higher altitudes due

to cooler temperatures and an increase in ambient moisture. However, both the El Panama

region and the Llano Pacaya ridge display marked departures from the regional trend, with

NDVI decreasing as the land surface rises to higher elevations (Figure 4-20), which forces

increased levels of interaction with the volcanic plume. Accordingly, the areas of the marked

kill-off coincide with the highest ground-level concentrations of SO2 and enhanced rates of S

dry deposition. At Poás, while the geometry of the topography is reversed, with the plume

initially in contact with the ground surface rather than only encountering it at a distance

from the vent, the relationship is quite similar to that at Masaya. Profiles adjacent to the

main kill area clearly exhibit a general healthy state of vegetation, whereas areas in which the

plume is in direct contact with the topography are areas of marked vegetation kill. The

effect of topography and elevation holds at Poás as at Masaya; once the slope at Poás

becomes steeper and drops away from the plume, such that the plume is no longer in direct

contact with the ground surface, apparently healthy vegetation dominates the whole of the

landscape.

While elevation may appear to be the clear factor in inducing interaction with the

plume aloft, the plume at Masaya is shown to be present over a much wider area than the

small ridge-top Llano Pacaya region denoted by extremely low NDVI. Profiles of the SO2

ground-level concentrations and S dry deposition show the persistence of the plume to

regions beyond the NDVI-based kill-off zone (Figure 4-21). SO2 is above background levels

over an approximately 1470 km2 region of land, and dry deposition exhibits a similar spatial

distribution (~1285 km2). Both ground-based datasets have maxima that do not coincide

with the largest drop in NDVI; this may be an artifact of the irregular and coarse sampling

scheme utilized by Delmelle et al. (2001, 2002), a result of the method of contouring of the

124

Figure 4-19: Plot of the relationship between NDVI and elevation along Masaya profile 4. Small and large arrows highlight departures from the positive region trend and correspond to the El Panama and Llano Pacaya regions, respectively.




125

Figure 4-20: Perspective view of 2001 Masaya NDVI draped over DEM. Note excessively damaged areas on the highlands of El Panama and Llano Pacaya. Vertical exaggeration is 7.5x.


126

initial datasets, or due to the fact that the ground-based data were collected over short

periods of time, under specific degassing and environmental conditions. As such, values

obtained for ground-level SO2 concentrations and S dry deposition velocities would not

necessarily coincide spatially with NDVI, the values of which are the result of years of

chronic exposure to Masaya’s plume. The spatial extent of the plume as determined by the

contoured surfaces of ground-based data is nevertheless a valid approximation of the area

that may suffer the detrimental effects of fumigation, as widespread damage to vegetation

has been noted beyond the border of the main kill area (Delmelle et al., 2001; 2002), and is

likely also an effect of the SO2 in the plume. While the central kill area on the Llano Pacaya

ridge is exposed to some of the highest ambient concentrations of SO2, chlorotic and

necrotic damage to a variety of plants has been found at SO2 concentrations as low as

60 ppb after only two weeks (Mandl et al., 1975).

Acidic gases such as SO2, HCl, and HF, all of which are constituents of volcanic

plumes at Masaya and elsewhere, have long been known to have detrimental effects on

vegetation with which they come in contact (e.g., Mandl et al., 1975; Ziegler, 1975; Hällgren,

1978; Linzon et al., 1979). Various amounts and combinations of these gases can have a

range of effects, from an increase in stored S, to damage of the leaves, to death of the plant,

with the likely source of SO2-induced damage being the associated disturbance of

intracellular pH regulation (Kropff, 1991). Much of the previous research in this discipline

has been restricted to laboratory settings, or field settings concerned with industrial

emissions of such gases. Little has been done to determine the effects of volcanic degassing

on vegetation, though some studies have previously been conducted to assess the damaging

effects of volcanic emissions on cultivated crops (Kratky et al., 1974) and wild vegetation

(Winner and Mooney, 1980). Other studies involve the investigation of periodic exposure to

SO2 of horticultural species in a laboratory setting (e.g., Dodd and Doley, 1998). However,

all of the aforementioned studies dealt predominantly either with acid rain and acute

fumigation events. In the case of Masaya, acute fumigation events do occur in association

with the dissipation of nocturnal temperature inversions, which increases vertical mixing of

air, including the plume (Delmelle et al., 2002). Persistent fumigation over longer periods of

time also likely exists when the plume is simply at a low altitude in the atmosphere, at which

point it is obstructed by the height of the Llano Pacaya ridge. Constant fumigation by SO2,

such as that on Llano Pacaya, has been shown to significantly affect plant growth in some

127

Figure 4-21: Data along Masaya profile 4, showing elevated dry deposition velocities and ground-level concentrations of SO2 persisting beyond the area marked by vegetation damage as reflected by NDVI residual values.




128

species at concentrations as low as 20 ppb (Mejstrik, 1980). In such cases, visible chlorotic

and necrotic damage is absent, but growth is impaired, with smaller leaves and roots.

In all cases of damage to vegetation by SO2, the damage occurs largely due to the

direct uptake of SO2 through a plant’s stomata (e.g., Pfanz et al., 1987). However, the

stomatal conductivity, or relative openness of the stomata to the atmosphere, is not

consistent with time; a number of factors can affect the degree of uptake of SO2. Studies

have shown that one of the most important factors governing the stomatal conductivity of

certain plant species is ambient relative humidity (Heck and Dunning, 1978; McLaughlin and

Taylor, 1981). The two variables are directly related, such that higher relative humidity is

correlated with an increased uptake of SO2. Masaya is generally quite dry during daytime

conditions, but relative humidity in the region exhibits diurnal oscillation during the dry

season (~December to April), with relative humidity increasing significantly at night.

Nocturnal temperature inversions in the Masaya area are common (Delmelle et al., 2002) and

often impede the propagation of the plume over the Llano Pacaya ridge, thereby causing

transient fumigation events, especially on the windward side of Llano Pacaya. While daytime

conditions normally favor a dry atmosphere and the plume’s clearing of Llano Pacaya, the

increased relative humidity (and, therefore, stomatal conductance and SO2 uptake) at night,

coupled with the temperature inversions, likely exacerbates the devastating effects on

vegetation. This is also of consequence at Poás, where the wet, humid climate of the

surrounding cloud forest is conducive to an increase in stomatal conductivity. Accordingly,

the humid climate of Poás may be the reason that the region immediately downwind of the

active crater is largely devoid of any vegetation as opposed to the highly damaged zone

downwind of Masaya despite the lower SO2 fluxes at Poás. Another factor which may

further aggravate the effects on vegetation at both volcanoes is the tendency of SO2 itself to

increase stomatal conductivity (Majernik and Mansfield, 1971; Biscoe et al., 1973). Effects of

low concentrations of SO2 have also been shown to be more damaging when present in

conjunction with HF (Mandl et al., 1975), as is often the case in volcanic emissions. Based

on the information regarding humidity and transient fumigation events associated with

nocturnal temperature inversions, it may be that the most severe kill atop Llano Pacaya is a

result of chronic fumigation and interaction with the plume during the day. The more subtle

damage to plants at lower elevations may be the effect of the acute exposures associated with

temperature inversions at night. Depending on the frequency of such inversion-induced


129

fumigation events, the plants at lower elevations may have more time to repair themselves in

response to the individual acute exposure prior to the next fumigation event.

In addition to the detrimental effects of volcanic fumigation on vegetation, there is

also the issue of the human population downwind of volcanoes like Masaya and Poás. Near-

constant exposure to SO2 and other acidic constituents of a volcanic plume can have a wide

range of ramifications on the health of the affected population, and the potential for

detrimental effects on human health at Masaya has been noted (Baxter et al., 1982). Acute

exposure to SO2 levels as low as 6 ppm causes noticeable irritation and inflammation of the

eyes and respiratory tract (Williams-Jones and Rymer, 2000); concentrations as low as 1 ppm

can lead to bronchoconstriction, decreases in maximum expiration, and decreased

respiratory rates (Environmental Protection Agency, 1982). Thresholds may be even lower

for populations at higher risk, such as children, senior citizens, asthmatics, and persons with

other pre-existing cardiopulmonary conditions. Eye irritation is evident at concentrations as

low as 20 ppm, and skin irritation occurs at concentrations of approximately 10,000 ppm

(Williams-Jones and Rymer, 2000). Levels of SO2 in the areas surrounding Masaya and Poás

do not generally exceed 230 ppbv and 500 ppb, respectively, except in areas immediately

adjacent to the active craters (Nicholson et al., 1996; Delmelle et al., 2002), but both

volcanoes have been continuously degassing for many years (Chapter 3; Rymer et al., 2005).

Few studies have been done on the effects of chronic exposure to low levels of SO2, but

anecdotal evidence from both volcanoes indicates a prevalence of dry throats, aggravated

respiratory problems, eye sensitivity, headaches, and bronchitis (Nicholson et al, 1996;

Williams-Jones and Rymer, 2000; Delmelle et al., 2002). Based on population data for

Nicaragua in 2000 and the extent of the severe vegetation kill area at Masaya, approximately

30,000 Nicaraguans are subjected to the most severe conditions. An additional 540,000

people inhabit the area in which elevated ground level SO2 concentrations are detectable.

Further, the acidic HCl and HF gases associated with the plume can also have detrimental

health effects. For example, excess fluoride consumed in food and water can lead to dental

fluorosis (e.g., Thórarinsson, 1979; Durand and Grattan, 1999; Delmelle et al., 2002).

Beyond the direct effects of persistent fumigation on human health, there is also the

issue of the impact that the volcanic fumigation of both Poás and Masaya has on the

livelihood of affected citizens. Concrete and metal are materials especially susceptible to


130

corrosion by acids such as those found in volcanic plumes, and many homes and other

structures around the two volcanoes are constructed of those materials. Roofs, fences,

telephone wires, and other metal mechanical equipment are repeatedly damaged by the

persistent fumigation and must often be replaced. Further, even in areas removed from the

direct fumigation that has led to the destruction of numerous lucrative coffee plantations,

individual instances of acid rain can have devastating effects on the agricultural community

of both countries. In Nicaragua, heavy rains from Hurricane Mitch in late 1998 mixed with

the acidic plume at Masaya to cause the complete destruction of a soya plantation over the

course of only one day (Williams-Jones and Rymer, 2000).

4.7 Conclusions

The use of ground-truth data and the integration of remotely sensed vegetation

indices (NDVI) and digital elevation models has identified areas exhibiting the most severe

vegetation kill at Masaya and Poás volcanoes. The severe kill appears restricted to above

‘threshold’ elevations at both volcanoes. This is interpreted to be an effect of the plumes

interacting directly with the ground surface at high elevations, as large topographic features

have been shown to affect plume behavior (Chapter 3). However, the presence of SO2 and

dry-deposited sulfur at Masaya exists beyond the area of most severe vegetation kill as

defined by NDVI data. Partial chlorotic and necrotic damage to vegetation also persists

beyond the kill zone identified by the NDVI (Delmelle et al., 2002). Significant detrimental

effects on vegetation, as well as the human population, can exist at relatively low

concentrations of SO2 and other acidic gases in volcanic plumes. Landsat NDVI alone is

not sufficient to identify all of the vegetative effects of the plumes at Masaya and Poás.

In the future, studies of environmental effects at volcanoes should utilize a multi-

parameter approach, incorporating a variety of ground-based and remotely-sensed datasets.

Epidemiological surveys would identify the effects felt by human populations, and more

extensive studies of the biological implications of plume fumigation on vegetation would

more accurately identify the aerial extent of the plume-damaged vegetation. At volcanoes

where remote sensing is the only feasible approach, the use of sensors with greater spectral

or spatial resolution than Landsat, such as hyperspectral sensors AVIRIS and Hyperion, may

provide more detailed information concerning the healthy or unhealthy state of vegetation.


131

4.8 References

Anderson, G.L., Hanson, J.D., and Haas, R.H., 1993. Evaluating Landsat thematic mapper derived vegetation indices for estimating above-ground biomass on semiarid rangelands. Remote Sensing of the Environment, 45: 165-175.


Baxter, P.J., Stoiber, R.E., and Williams, S.N., 1982. Volcanic gases and health: Masaya Volcano, Nicaragua. The Lancet, 2: 150-151.

Bice, D.C., 1980. Tephra strigraphy and physical aspects of recent volcanism near Managua, Nicaragua. PhD Thesis, University of California, Berkeley, CA, USA.

Biscoe, R.V., Unsworth, M.H., and Pinckney, H.R., 1973. The effects of low concentrations of sulphur dioxide on stomatal behaviour in Vicia faba. New Phytologist, 72: 1299-1306.

Casadevall, T.J., Doukas, M.P., Neal, C.A., McGimsey, R.G., and Gardner, C.A., 1994. Emission rates of sulfur dioxide and carbon dioxide from Redoubt Volcano, Alaska during the 1989-1990 eruptions. Journal of Volcanology and Geothermal Research, 62: 519-530.

Delfosse, T., Delmelle, P., Givron, C., and Delvaux, B., 2005a. Inorganic sulphate extraction from SO2-impacted Andosols. European Journal of Soil Science, 56: 127-133.

Delfosse, T., Delmelle, P., Iserentant, A., and Delvaux, B., 2005b. Contribution of SO3 to the acid neutralizing capacity of Andosols exposed to strong volcanogenic acid and SO2 deposition. European Journal of Soil Science, 56: 113-125.

Delmelle, P., Baxter, P., Beaulieu, A., Burton, M., Francis, P., Garcia-Alvarez, J., Horrocks, L., Navarro, M., Oppenheimer, C., Rothery, D., Rymer, H., St. Amand, K., Stix, J., Strauch, W., Williams-Jones, G., 1999. Origin, effects of Masaya Volcano’s continued unrest probed in Nicaragua. EOS, Transactions, American Geophysical Union 80, 575- 581.

Delmelle, P., Delfosse, T., and Delvaux, B., 2003. Sulfate, chloride and fluoride retention in Andosols exposed to volcanic acid emissions. Environmental Pollution, 126: 445-457.

Delmelle, P., Stix, J., Baxter, P.J., Garcia-Alvarez, J., and Barquero, J., 2002. Atmospheric dispersion, environmental effects and potential health hazard associated with the low-altitude gas plume of Masaya volcano, Nicaragua. Bulletin of Volcanology, 64: 423-434.



132

Dodd, I.C., and Doley, D., 1998. Growth responses of cucumber seedlings to sulphur dioxide fumigation in a tropical environment. Environmental and Experimental Botany, 39: 41-47.

Durand, D., and Grattan, J.P., 1999. Extensive respiratory health effects of volcanogenic dry fog in 1783 inferred from European documentary sources. Environmental Geochemistry and Health, 21: 371-376.

Elias, T., Sutton, A.J., Stokes, J.B., and Casadevall, T.J., 1998. Sulfur dioxide emission rates of Kilauea Volcano, Hawaii, 1979-1997. U.S. Geological Survey Open-File Report 98-462.

Environmental Protection Agency, 1982. Air Quality Criteria for Particulate Matter and Sulfur Oxides, Volume I. Environmental Protection Agency Report EPA-600/8-82-029a.

Fournier, N., Rymer, H., Williams-Jones, G., Brenes, J., and Markiewicz, S. Persistent degassing at Poás volcano, Costa Rica: multidisciplinary study of hydrothermal and shallow magmatic processes. Journal of Volcanology and Geothermal Research, submitted.


Garrec, J.P., Plebin, R., and Faivre-Pierret, R.X., 1984. Impact of Volcanic Fluoride and SO2 Emissions from Moderated Activity Volcanoes on the Surrounding Vegetation. Bulletin of Volcanology, 47: 491-496.

Gerstell, M.F., Crisp, J., and Crisp, D., 1993. Radiative forcing of the stratosphere by SO2, ash, and H2SO4 aerosols, during the first 3 months after the El Chichon eruption. Eos, Transactions, American Geophysical Union, vol. 74, no. 43, Suppl., pp. 105.

Hällgren J.E., 1978. Physiological and biochemical effects of sulfur dioxide on plants. In: Nriagu, J.O. (Editor) Sulfur in the Environment Part II. Ecological Impacts. Wiley and Sons, New York: 163-209.

Hansen, J., Lacis, A., Ruedy, R., and Sato, M., 1992. Potential climate impact of Mount Pinatubo eruption. Geophysical Research Letters, 19: 215-218

Heck, W.W., and Dunning, J.A., 1978. Response of oats to sulfur dioxide: interactions of growth temperature with exposure to temperature or humidity. Journal of the Air Pollution Control Association, 28: 241-246.

Johnson, N., and Parnell Jr., R.A., 1986. Composition, distribution and neutralization of ‘acid rain’ derived from Masaya volcano, Nicaragua. Tellus, 38B: 106-117.

Koslowski, T.T., 1980. Impacts of air pollution on forest ecosystems. Bioscience, 30: 88-93.

Kratky, B.A., Fukunaga, E.T., Hylin, J.W., and Nakano, R.T., 1974. Volcanic Air Pollution: Deleterious Effects on Tomatoes. Journal of Environmental Quality, 3: 138-140.


133

Kropff, M.J., 1991. Long-term effects of SO2 on plants, SO2 metabolism and regulation of intracellular pH. Plant and Soil, 131: 235-245.

Lawrence, R.L., and Ripple, W.J., 1998. Comparisons among vegetation indices and bandwise regression in a highly disturbed, heterogeneous landscape: Mount St. Helens, Washington. Remote Sensing of the Environment, 64: 91-102.

Likens, G.E., and Bormann, F.H., 1974. Acid rain: a serious regional environmental problem. Science, 184: 1176-1179.

Linzon, S.N., Temple, P.J., and Pearson, R.G., 1979. Sulfur concentrations in plant foliage and related effects. Journal of the Air Pollution Control Association, 29: 520-525.

Luhr, J.F., 1991. Mount Pinatubo: volcanic shade causes cooling. Nature, 354: 104-105.

Majernik, O., and Mansfield, T.A., 1971. Effects of SO2 pollution on stomatal movements in Vicia faba. Phytopathologische Zeitschrift, 71: 123-128.

Mandl, R.H., 1974. Effects of hydrogen fluoride and sulphur dioxide alone and in combination on several species of plants. Environmental Pollution, 9, 133-143.



McLaughlin, S.B., and Taylor, G.E., 1981. Relative humidity: important modifier of pollutant uptake by plants. Science, 211: 167-169.

Mejstrik, V., 1980. The influence of low SO2 concentrations on growth reduction of Nicotiana tabacum L. cv. Samsun and Cucumus sativus cv. Unikate. Environmental Pollution, 21:73-76.

Nicholson, R.A., Roberts, P.D., and Baxter, P.J., 1996. Preliminary studies of acid and gas contamination at Poás volcano, Costa Rica. In: Appleton, J.D., Fuge, R., and McCall, G.J.H. (Editors), Environmental Geochemistry and Health, Geological Society, Special publications, 113: 239-244.

Pfanz, H., Martinoia, E., Lange, O.-L., and Heber, U., 1987. Flux of SO2 into leaf cells and cellular acidification by SO2. Plant Physiology, 85: 928-933.

Pollack, J.B., Toon, O.B., Sagan, C., Summers, A., Baldwin, B., and Van Camp, W., 1976. Volcanic explosions and climate change, a theoretical assessment. Journal of Geophysical Research, 81: 1071-1083.

Rouse, J.W., Haas, R.H., Schell, J.A., Deering, D.W., and Harlan, J.C., 1973. Monitoring the vernal advancement and retrogradation (greenwave effect) of natural vegetation. NASA/GFSC Type III Final Report, Greenbelt, MD.


134


Rymer, H., Locke, C.A., Brenes, J., and Williams-Jones, G., 2005. Magma plumbing processes for persistent activity at Poás volcano, Costa Rica. Geophysical Research Letters, 32, L08307, doi:10.1029/2004GL022284.

Sandoval, L., Valdés, J., Martínez, M., Barquero, J., and Fernández, E., 1996. Efecto de las emisiones del Volcán Poás sobre la vegetación, Costa Rica. Ingenieria y Ciencia Quimica, 16: 62-64.

Sebesta, J., 1997. Dynamic development of the relief in the Managua area, Nicaragua. Acta Universitatis Carolinae/Geographica, 2: 93-109.













Smithsonian Institution, 2003. Poás. Bulletin of the Global Volcanism Network, 28(9).

Smithsonian Institution, 2006a. Masaya. Bulletin of the Global Volcanism Network, 31(4).

Smithsonian Institution, 2006b. Poás. Bulletin of the Global Volcanism Network, 31(3).


Thórarinsson, S., 1979. On the damage cuased by volcanic eruptions with special reference to tephra and gases. In: Sheets, P.D., and Grayson, D.K. (Editors), Volcanic activity and human ecology, Academic Press, New York: 125-159.

Tucker, C.J., 1979. Red and photographic infrared linear combinations for monitoring vegetation. Remote Sensing of the Environment, 35: 105-119.


135

van Wyk de Vries, B., 1993. Tectonic and magma evolution of Nicaraguan volcanic systems. PhD Thesis, Open University, Milton Keynes, UK.

Williams, W.T., Brady, M., and Williamson, S.C., 1977. Air pollution damage to the forests of the Sierra Nevada mountains of California. Journal of the Air Pollution Control Association, 27: 230-234.

Williams-Jones, G, and Rymer, H., 2000. Hazards of volcanic gases. In: Sigurdsson, H., Houghton, B., Mcnutt, S.R., Rymer, H., and Stix, J. (Editors), Encyclopedia of Volcanoes, Academic Press, San Diego: 997-1004.


Winner, W.E., and Mooney, H.A., 1980. Responses of Hawaiian Plants to Volcanic Sulfur dioxide: Stomatal Behavior and Foliar Injury. Science, 210: 789-791.

Ziegler, I., 1975. The effect of SO2-pollution on plant metabolism. Residue Review, 56: 79-105.

Chapter 5 – Conclusions

136

Chapter 5: CONCLUSIONS

5.1 General Conclusions

Masaya volcano, Nicaragua has long been a frequent study site for volcanologists,

especially with regard to the study of volcanic SO2. However, with the evolution of gas-

sensing technology, methods, and the reduction of measurement error, it was felt that

revisiting the issue of SO2 plume monitoring at Masaya was warranted. Two month-long

campaigns to measure SO2 fluxes and the conversion of SO2 to sulfate aerosols were

conducted in tandem with an evaluation of remote and ground-based data. These studies

resulted in important findings with respect to the behavior of Masaya’s plume and ensuing

environmental effects.

Over 700 SO2 flux measurements were made over the course of the two field

campaigns. The average coefficient of variation of a given day’s measurements is commonly

~30%, consistent with results since 1972. A substantial discrepancy in the values for flux, as

measured at different distances downwind, was noted. Previously, the resolution of the SO2

flux dataset at Masaya was too sparse, temporally, to exhibit this discrepancy. In 2005 and

2006, however, fluxes measured on the Llano Pacaya road, ~15 km downwind of the vent,

were consistently ~30-50% lower than those recorded on the Ticuantepe road, only ~5 km

from the vent.

Both conversion of SO2 to sulfate aerosols and dry deposition of sulfur were

discounted as possible causes of the lower fluxes measured on Llano Pacaya. Rather, it is

interpreted that the complex topography of the Masaya region indirectly induces the

variation in flux magnitudes from Ticuantepe to Llano Pacaya. The existence of the Llano

Pacaya ridge forces local acceleration of regional trade winds and, accordingly, acceleration

of the plume itself, which dilutes the portion of the plume measured over the Llano Pacaya

ridge. Wind acceleration factors were determined to be, at minimum, approximately 1.4,

though factors approaching 2 are likely to be more realistic. Acceleration of the plume over

topographic barriers, therefore, is interpreted to be the main source of the apparent loss

downwind. These findings highlight the need to measure plume speed as accurately as


137

possible at the same location at which SO2 measurements are being made in order to obtain

the most representative flux of volcanic SO2. Failure to do so in the past may have

significantly skewed estimates of global volcanic SO2 output as well as estimates of volumes

of degassed magma.

Spatially mapped Normalized Differential Vegetation Indices (NDVI) derived from

Landsat ETM+ data identified areas of the most severe vegetation damage at Masaya as well

as at Poás volcano, Costa Rica. The areas displaying the greatest extent of vegetation kill

correspond to regions of locally high relief, where the acidic plumes interact with the ground

surface. However, NDVI is not a sufficient tool in detecting all of the negative impacts of

the plume on vegetation. SO2, S, and Cl, as well as visible damage to plants, are detectable

over a much wider area than the small area identified by NDVI as severely damaged.

Nonetheless, the integration of DEMs and Landsat imagery, which are often readily

available, allows for a rapid first order evaluation of the areas subject to the most intense

fumigation near volcanoes, which may be beneficial for hazard mitigation.

5.2 Recommendations for Future Work

In future studies at Masaya, plume speed measurements need to be made

simultaneously on both the Llano Pacaya and Ticuantepe roads, using the multi-

spectrometer method to accurately determine the speed of the plume above each of the

measurement sites. Completion of such measurements will confirm whether or not the 30-

50% loss seen between the two roads can in fact be accounted for by different wind speeds

at the different measurement sites. Atmospheric modeling should be investigated in future

studies, as plume speed data in tandem with plume transport models will confirm

acceleration of the wind and entrained plume over the Llano Pacaya ridge. Further, all flux

measurements made should be accompanied by a plume speed measurement made at the

same distance downwind.

With respect to SO2 monitoring at volcanoes worldwide, this study shows that

greater care must be taken when making assumptions about the behavior of a given volcanic

plume. Atmospheric conditions such as wind speed are not necessarily constant over the

entire reach of a volcanic plume, such that SO2 fluxes measured at different locations are not

the same, and neither is necessarily representative of the true rate of SO2 degassing from the


138

volcano in question. Assuming otherwise may lead to apparent inconsistencies in time-series

datasets that are not real, which has far-reaching implications given that volcanologists may

use SO2 time series data in tandem with seismic, ground deformation, or gravity data in

order to interpret volcanic behavior. SO2 flux variations that are merely artifacts of

measurement methods, therefore, may be wrongly interpreted. Future SO2 monitoring

campaigns at Masaya and elsewhere should be rigorous in conducting plume speed

measurements at the same site as SO2 measurements for far greater accuracy in determining

SO2 flux than has previously been attained.

To assess the true extent of environmental ramifications of downwind fumigation by

volcanic plumes, more diversity and magnitude is required for datasets involved. Effects of

persistently degassing volcanoes on human populations have not been evaluated to date;

epidemiological surveys would help identify such effects. Biological research on plant

species specific to volcanic regions and their responses to volcanic plumes would yield

insight on the full ramifications of incessant fumigation. Such research could shed light on

species of plants and crops more suited to cultivation in areas susceptible to the deleterious

effects of volcanic degassing. Also, incorporation of data from hyperspectral remote sensors

such as AVIRIS and Hyperion may provide more detailed and accurate information

concerning the healthy or unhealthy state of the vegetation near degassing volcanoes.

Use of diverse and extensive datasets, with care taken to constrain all variables, such

as plume speed, will offer a greater understanding of the volcanic plume at Masaya, and,

consequently, of the plumes at volcanoes worldwide.

Appendix A – Comprehensive Methodologies

139

APPENDIX A: COMPREHENSIVE METHODOLOGIES

A.1 SO2 Traverses

All SO2 measurements made for this study involved the use of ultraviolet

spectrometers and principles of differential optical absorption spectroscopy (DOAS).

DOAS techniques for the evaluation of concentrations of trace gases in the atmosphere rely

on the Beer-Lambert law, which dictates the extent of absorption of electromagnetic

radiation by matter:

I(λ) = I0(λ) exp(-Lσ(λ)c) [A-1]

I(λ) is the measured radiation at a specific wavelength, I0(λ) is the initial intensity of the

radiation prior to passing through a gas column of thickness L, σ(λ) is an intrinsic property

of the desired trace gas called the absorption cross section or absorption coefficient, which is

measured in a laboratory. The concentration of the trace gas of choice is denoted as c. Any

gas has a specific spectral absorption signature, whereby it blocks varying degrees of

incoming radiation at different wavelengths. SO2 has strong absorption features in the 200-

230 nm and the 290-310 nm ranges (Platt, 1994, and references therein). It is the above

principle upon which the methods utilized in this study are based.

The FLYSPEC, described in Horton et al. (2006) and Elias et al. (2006), is composed

mainly of an Ocean Optics USB2000 spectrometer (or USB4000, as with the FLYSPEC v2).

A 25–μm slit results in a spectral resolution of 0.25 nm over a range of wavelengths from

177-330 nm. The FLYSPEC utilizes a laptop computer for data processing and as a power

source, and weighs less than two kilograms, including the small, sub-notebook computer.

Unlike some other small spectrometers in use (e.g., mini-DOAS; Galle et al., 2002),

the FLYSPEC configuration does not include a fiber optic cable with which to attach an

external telescope. Rather the telescope is affixed directly to the spectrometer to reduce light

losses. The FLYSPEC has a field of view of 44 mrad. Additionally, the FLYSPEC has a UV

filter to reduce stray light, and includes SO2 calibration cells within its configuration. All


140

components of the instrument are contained within a robust, weather-proof case, well-suited

for field deployment.

The FLYSPEC relies on its two calibration cells of known SO2 concentration

(approximately 400 ppm-m and 1600 ppm-m for the low and high calibration cells,

respectively) in order to determine the amount of SO2 in the column of atmosphere

overhead. Instead of comparing final spectra to a laboratory spectrum to determine SO2

concentration, as do other currently used spectrometer systems, the final spectra from the

FLYSPEC are calibrated to reference spectra produced by the insertion of calibration cells

of known concentration into the instrument’s field of view. Calibrations are performed in

the field prior to every second measurement and in the same atmospheric conditions as

collection of the larger, primary dataset. As such, the FLYSPEC methodology avoids

otherwise obligatory corrections for atmospheric effects in the spectra, as with Fraunhofer

lines or Ring-effect Raman spectra, which can present problems when laboratory spectra are

utilized for calibrations. By using calibration-cell reference spectra obtained in the field, any

atmospheric effects will be present in all spectra, negating the need for data corrections

(Horton et al., 2006). The FLYSPEC also records and stores full-spectra, allowing for re-

processing and comparison to spectra obtained at other times, or in other physical settings.

FLYSPEC results in field trials are comparable to those of other spectrometers used to

measure volcanic SO2, the COSPEC and the mini-DOAS (Elias et al., 2006; Horton et al.,

2006).

Rather than reporting simple concentrations of SO2, such as ppm, the FLYSPEC

and other spectrometers report values in units of concentration-path length, or ppm-m.

Because the spectrometers measure the amount of SO2 in an “infinite” column between the

sensor and the sun, the true concentration of SO2 in a given parcel of air cannot be

determined. A concentration-path length of 100 ppm-m could indicate either an average

concentration of 100 ppm for a plume thickness of 1 m, or an average concentration of

10 ppm over a plume of a thickness equal to 10 m. It is for this reason that any vertical

dispersion of the plume with propagation from the vent is not of concern for this method.

In the field at Masaya, initial set-up of the FLYSPEC equipment was performed

approximately 5 km from the general location of the plume, as determined from archived

data. For each FLYSPEC, the external Pelican case was opened and the spectrometer,


141

calibration cells, and electronic connections checked to ensure that the equipment was in

working order. The original FLYSPEC was then mounted to the vehicle, wedged between

the front passenger-side door frame and the side mirror. With the window completely rolled

down, the instrument was secured with duct tape. The external GPS antenna was duct-taped

to the roof of the vehicle. All wires were passed through the open window to a small laptop

computer, which was operated by the passenger. The FLYSPEC v2, with its horizontally-

configured Pelican case, was placed on the roof of the car and secured either by duct tape or

by two compression straps that were fed through the rear passenger- and driver-side doors

of the car. When the compression straps were used, duct tape supplemented the two straps.

The single data cable from the FLYSPEC v2 was then fed through the front passenger

window to the laptop computer. With both instruments, a small bubble level was utilized to

ensure that the fields of view on the FLYSPECs were vertical.

Once the instruments were in place and a clear view of the plume-free sky obtained,

set-up continued with the preparation of the software. For the original FLYSPEC, the

software used was LapFly (or LapFly v2 in the case of the FLYSPEC v2), which was

developed at the Hawaii Institue of Geophysics and Planetology (HIGP), University of

Hawaii at Manoa. After opening LapFly, a parameter file containing the concentration path

lengths of the two calibration cells in the instrument was loaded. A preliminary scan was

conducted to determine the intensity of the incoming solar radiation. To avoid saturation of

the spectrometer during measurements, a range of sampling intervals were tested, and an

appropriate sampling scheme chosen (generally 1, 2, or 3 scans per second, averaged to an

interval of one measurement per second). An initial set of calibrations was then taken and

stored using the file structure MDDYYYY\calibration type-road name-traverse number (for

example, 3022006\highlp1 for the first high calibration on the Llano Pacaya road on March

2, 2006). In the case of the FLYSPEC 2 and the LapFly v2 software, the calibrations were

automated, and after naming the files, the instrument subsequently completed a dark (no

incoming radiation) scan, a high calibration cell scan, a low calibration cell scan, and a clear

sky reference scan. The procedure was nearly identical for the original FLYSPEC, with the

only difference being that each calibration scan had to be initiated individually by the

instrument operator and prior to each scan, the correct calibration cell, or lack thereof, had

to be manually placed into the field of view of the instrument. During the calibration scans,

the measured incoming radiation was monitored to verify that the correct measurement was


142

being conducted for each scan. For example, little to no radiation sensed by the

spectrometer during the dark, and radiation intensities that increased from the high

concentration gas cell, to the low concentration gas cell, to the clear sky were indicative of a

proper set of calibration scans (Figure A-1). If, for some reason, either due to mechanical or

operator error, incorrect radiation intensities were recorded during some part of the

calibration, the scan was aborted and re-started with the correct gas cell configuration.

Following successful calibration, the first SO2 profile was started. Once a filename

was designated (MDDYYYY\road name-traverse number.txt; 2242006\t1.txt for the first

Ticuantepe traverse on February 24, 2006), data collection was initiated by the operator. The

car proceeded along either the Ticuantepe or Llano Pacaya road, with the FLYSPEC

displaying the data collected in real time. Spectra of incoming radiation intensities over a

range of wavelengths as well as the calculated concentration path length (in ppm-m) were

output graphically at one-second intervals. During the course of the traverse, the operator

made note of the concentration path length, determining when the vehicle was beneath the

plume and measuring SO2. Once the concentration path lengths returned to background

levels, as determined by the operator, the traverse was complete. The car was then turned

around, and another traverse scan was initiated. New calibration scans were completed in

the clear sky outside the plume after every two traverses to account for any changes in

background atmospheric conditions. Traverses were continued as necessary until the sun

angle grew too low for adequate UV input into the spectrometer (generally around 4 pm

local time in Nicaragua).

Post-processing of the SO2 traverse scans involved the utilization of a second piece

of software, FluxCalc (also developed by the HIGP, University of Hawaii at Manoa), in

order to determine the average daily flux of SO2 represented by each individual traverse. In

FluxCalc, any of the road traverse .txt files created by LapFly could be opened and displayed

graphically as a plot of concentration path length of SO2 versus time. Based on the graph,

minor adjustments could be made to correct for any flaws within the data: spikes in the data

due to trees overhead would be ignored by clipping the data to a subset that included only

the plume; and baseline drift due to changing sun angles over the course of long traverses

could be accounted for by setting a new, subhorizontal baseline. Once the SO2 profile was

deemed adequate, the coordinates of Masaya’s summit vent (the source of SO2) were


143

Figure A-1: An example of raw spectral radiances from a set of calibration scans from FLYSPEC, showing the clear sky reference and low and high calibration cell spectra. These spectra have all been dark-subtracted to reduce electronic and dark-current noise.


144

entered into the program to allow for perpendicularity corrections based on the positional

data for each SO2 measurement. Subsequently, a measured plume speed, wind speed, or

normalizing factor of 1 m/s was entered into FluxCalc, which then calculated the average

daily SO2 flux based on any one particular traverse. Fluxes were calculated in this manner

for each of the traverses performed, both at 1 m/s to normalize the data, and at the

appropriate calculated plume speed (or wind speed measured by hand held anemometer in

the absence of spectrometer-based plume speed data).

A.2 Aerosol Traverses

The Microtops II is a 5-channel Volz-type sun photometer (Morys et al., 1996;

Porter et al., 2001). When the Microtops is aimed at the sun for a measurement, the on-

board microprocessor records incoming voltages at 5 wavelengths. Multiple measurements

in clear sky at various times are utilized in performing a Langley calibration, whereby the

intensity of solar radiation prior to attenuation by Earth’s atmosphere is determined (Shaw,

1983; Tomasi et al., 1997). When the natural log of background voltage measurements are

plotted against the air mass factor (a function of the length of the atmosphere through which

incoming solar radiation passes at a given time of day, on a given date, at a given location),

the y-intercept (at an air mass of zero, or the top of Earth’s atmosphere) of the best-fit line is

representative of initial incoming radiation, I0(λ). The slope of the same line is equal to τb, or

the background optical thickness of the atmosphere. To implement these steps,

measurements in the field were made before and after each traverse in clear, plume-free sky.

A Langley calibration may be performed based on field measurements or, alternatively,

previous calibrations stored in the Microtops II instrument may be used. As the use of

previously obtained calibrations does not cause significant variations in the measurements

(Ichoku et al., 2002), we accepted the internal calibration of the Microtops and did not

perform Langley calibrations.

If atmospheric conditions were promising (i.e., cloud-free and likely to remain cloud-

free for the duration of the traverse), either the FLYSPEC operator or a second passenger in

the backseat of the field vehicle would begin measurements with the sun photometer. The

photometer’s lens was first cleaned using a Kimwipe, turned on, and the instrument time

adjusted to coincide with that being recorded by the FLYSPEC’s GPS. Once the SO2 data


145

collection had been initiated in the clear sky, the vehicle would not immediately proceed

along the road as with a normal SO2 traverse. Rather, the initial location of the car was

recorded and marked using a hand-held GPS unit, and five sun photometer measurements

were taken. Five measurements were made for averaging purposes, as well as to allow for

deletion of erroneous measurements, such as when the sun was obscured by a passing truck.

Once the first data point’s measurements were taken, the vehicle then proceeded

approximately 300 m further down the road, where the process was repeated, with the

marking of the location with the handheld GPS and the taking of 5 measurements with the

sun photometer. This photometer data collection continued until the FLYSPEC indicated

that the car was no longer beneath the plume, at which point a final clear sky background

aerosol measurement for the traverse would be taken for use in the accounting for

atmospheric scattering and absorption unrelated to plume aerosols.

Aerosol measurements in tandem with SO2 traverses were completed as frequently as

the oft-cloudy skies permitted. In the most ideal of situations, with cloud-free skies above

both the Llano Pacaya and Ticuantepe roads, aerosol traverses were conducted in a quasi-

simultaneous fashion, with each vehicle outfitted with a FLYSPEC and sun photometer and

collecting data over the course of an entire day.

Following field measurements, post-processing was done using IDL (ITT Industries,

Inc.) code after Watson and Oppenheimer (2000) and Mather et al. (2004). From the

measured incoming radiation of a given measurement, I(λ), along with the calculated air

mass factor, m, and the initial radiation, I0(λ), the total optical thickness of a measurement

within the plume, -τtot, can be calculated using the Beer-Lambert law:

I(λ) = I0(λ) exp(-τtotm) [A-2]

The aerosol optical thickness due to the constituents of a volcanic plume, τp, can

then be determined by simply subtracting τb, the background optical thickness (as

determined either from a Langley calibration or from application of Equation A-2 to a

Microtops II measurement made in clear sky), from τtot. Once τp is determined (given that

τM, the Mie optical depth, is equivalent to τp, which is assumed because of the


146

aforementioned correction by subtraction of τb from τtot), a particle size distribution can be

found through the inversion of a Fredholm integral of the first kind:

τM(λ) = ∫ πr2QEXT(2πr/λ, mi)nc(r)dr [A-3]

where QEXT is the extinction efficiency factor calculated from Mie scattering theory, r is the

particle radius, mi is the complex index of refraction, nc(r) is the particle size distribution, and

the limits of the integral are equivalent to assumed maximum and minimum particle sizes.

Inversion of Equation A-3 yields:

nc(r) = dNc/dr = cr-v-1 [A-4]

where Nc is the particle number per unit area of the measured plume, c is the concentration

parameter, and v is the Junge exponent, which determines the gradient of the particle size

distribution.

Once the particle number per unit area was determined for all radii for all

measurement sites within a traverse, the total number of particles for that measurement site

was found by summing the particle numbers of all of the size bins. The totals were then

averaged for all of the sites within the traverse. The average was then multiplied by the

length of the traverse and a normalized plume speed of 1 m/s in order to find the particle

flux for each traverse in particles/second.

A.3 Plume Speed

As a first-order approximation of the speed of the plume’s propagation, hand-held

anemometer measurements were taken, with the ground-level wind speed acting as a proxy

for the true plume speed. The anemometer was held up perpendicular to the wind’s

direction, at arms length above the operator’s head to a height of just over 2 m above

ground level. Each measurement was an approximately five-minute average taken either

between traverses in the general vicinity of the plume’s center on the Llano Pacaya or

Ticuantepe roads, or during the stationary spectrometer-based plume speed measurements

taken once daily on either the Ticuantepe or the Llano Pacaya road.

A substantially better estimate of the plume’s speed was achieved by aligning two

FLYSPECs along the axis of propagation of the plume and tracking the puffs of the plume


147

over a known distance. Initially, each spectrometer was taken to an area away from the SO2

plume and a full calibration was performed on each instrument. Subsequently, the

instruments were returned to the location of the center of the plume (as determined from

previous SO2 traverses). Each instrument was removed from the car and fastened to a metal

plate that had been fitted to a small camera tripod. The external GPS antennae for the

original FLYSPECs were also attached with duct tape to the metal plate. The tripods were

then placed approximately 30 – 50 m apart beneath the plume and parallel to the plume’s

axis of propagation. Each spectrometer was leveled to ensure a vertical field of view. The

laptop computers also remained part of the configuration, set on the ground in the shade of

the tripod. In some instances the tripod configuration was omitted in favor of leaving the

FLYSPECs attached to the cars. In these cases, the instruments’ positions on the cars were

carefully adjusted to again ensure a vertical field of view. Regardless of the physical

platform, once the FLYSPECs were level, the LapFly software on each computer was set to

record data for 30 minutes, and the data collection was initiated simultaneously for the two

time-synchronized FLYSPECs. In some instances measurements were stopped early,

resulting in a 20 minute dataset.

Following the half-hour of data collection, the recorded concentration path lengths

would reflect the puffing and turbulence within the plume, seen as sharp peaks and valleys

within the data set. Similar structures would have been recorded by each of the two

FLYSPECs, though with a time lag between them. This time lag, in combination with the

known distance between the instruments, as measured by the GPS on each tripod, could

then be exploited to determine the speed of the plume as it moved over the two tripods.

An IDL code called ‘plumespeed.pro’ (Williams-Jones et al., 2006), was used for the

calculation of the plume’s speed. Both the upwind FLYSPEC and downwind FLYSPEC

data files were imported into IDL. Based on the structure of the SO2 concentration path

length variations within the two files, the program iteratively matched the scans to each

other, calculating a correlation coefficient for each temporal offset attempted. The offset

that yielded the best correlation between the upwind and downwind files was then assumed

to be the actual time lag between the instruments, and using either the distance between the

spectrometers as calculated from the GPS data within the files or a manually entered

distance (e.g., as measured on the ground with a tape measure), the program then calculated


148

a single plume speed for the whole of the measurement. On average, one spectrometer-

based measurement was conducted each day.

In the case of full files that did not correlate well, either due to varying wind speeds,

lateral movement of the plume away from the dual-spectrometer array, or some other factor,

the plume speed program also allowed for the fitting of specific windows of time (10 mins, 5

min, and 3 mins) from within the 20 or 30 minute period. A specific subset of the

measurement could be designated, or a moving window of any length of time could be set in

order to divide the 20 or 30 minutes in to equal segments. Each of the segments would then

be treated as individual measurements, each with their own calculated plume speed. Final

plume speeds for the full measurement were then determined by averaging the plume speeds

calculated for only those segments that yielded R2 values greater than 0.9.


149

A.4 References

Elias, T., Sutton, A.J., Oppenheimer, C., Horton, K.A., Garbeil, H., Tsanev, V., McGonigle, A.J.S., and Williams-Jones, G., 2006. Intercomparison of COSPEC and two miniature ultraviolet spectrometer systems for SO2 measurements using scattered sunlight. Bulletin of volcanology, 68: 313-322.

Galle, B., Oppenheimer, C., Geyer, A., McGonigle, A.J.S., Edmonds, M., and Horrocks, L., 2002. A miniaturised ultraviolet spectrometer for remote sensing of SO2 fluxes; a new tool for volcano surveillance. Journal of volcanology and geothermal research, 119(1-4): 241-254.

Horton, K.A., Williams-Jones, G., Garbeil, H., Elias, T., Sutton, A.J., Mouginis-Mark, P., Porter, J.N., and Clegg, S., 2006. Real-time measurement of volcanic SO2 emissions: Validation of a new UV correlation spectrometer (FLYSPEC). Bulletin of volcanology, 68: 323-327.

Ichoku, C., Levy, R., Kaufman, Y.J., Remer, L.A., Li, R.-R., Martins, V.J., Holben, B.N., Abuhassan, N., Slutsker, I., Eck., T.F., and Pietras, C., 2002. Analysis of the performance characteristics of the five-channel Microtops II sun photometer for measuring aerosol optical thickness and precipitable water vapor. Journal of Geophysical Research, 107(D13), 4179, doi:10.1029/2001JD001302.

Platt, U., 1994. Differential Optical Absorption Spectroscopy (DOAS). In: Sigrist, M.W. (Editor), Air Monitoring by Spectroscopic Techniques, Chemical Analysis Series, 127. Wiley, New York: 27-84.

Shaw, G.E., 1983. Sun Photometry. Bulletin of the American Meteorological Society, 64(1): 4-10.

Tomasi, C., Vitale, V., and Tarrozi, L., 1997. Sun-photometric measurements of atmospheric turbidity caused by the Pinatubo aerosol cloud in the Himalayan region during the summer periods of 1991 and 1992. Nuovo Cimento, 20C(1): 61-88.


Appendix B – Data

150

APPENDIX B: DATA

B.1 2005 and 2006 SO2 traverses along the Ticuantepe and Llano Pacaya roads. (See Figure B-1 for reference; 12 records on 2/26/06 marked by an asterisk indicate 6 runs

that were completed with two spectrometers simultaneously)

Date Road Spectrometer Run # Flux @ 1 m/s

2/25/05 Ticuantepe FLYSPEC 1 164 2/25/05 Ticuantepe FLYSPEC 2 188 2/25/05 Ticuantepe FLYSPEC 3 233 2/25/05 Ticuantepe FLYSPEC 4 157 2/25/05 Ticuantepe FLYSPEC 5 149 2/25/05 Ticuantepe FLYSPEC 6 90 2/25/05 Llano Pacaya FLYSPEC 1 220 2/25/05 Llano Pacaya FLYSPEC 2 202 2/25/05 Llano Pacaya FLYSPEC 3 178 2/25/05 Llano Pacaya FLYSPEC 4 221 2/25/05 Llano Pacaya FLYSPEC 5 145 2/25/05 Llano Pacaya FLYSPEC 6 256 2/26/05 Ticuantepe FLYSPEC 1 393 2/26/05 Ticuantepe FLYSPEC 2 316 2/26/05 Ticuantepe FLYSPEC 3 345 2/26/05 Ticuantepe FLYSPEC 4 345 2/26/05 Ticuantepe FLYSPEC 5 298 2/26/05 Ticuantepe FLYSPEC 6 184 2/26/05 Ticuantepe FLYSPEC 7 417 2/26/05 Ticuantepe FLYSPEC 8 505 2/26/05 Llano Pacaya FLYSPEC 1 230 2/26/05 Llano Pacaya FLYSPEC 2 215 2/26/05 Llano Pacaya FLYSPEC 3 261 2/26/05 Llano Pacaya FLYSPEC 4 289 2/26/05 Llano Pacaya FLYSPEC 5 237 2/26/05 Llano Pacaya FLYSPEC 6 385 2/26/05 Llano Pacaya FLYSPEC 7 474 2/27/05 Ticuantepe FLYSPEC 1 353 2/27/05 Ticuantepe FLYSPEC 2 470 2/27/05 Ticuantepe FLYSPEC 3 415 2/27/05 Ticuantepe FLYSPEC 4 430 2/27/05 Ticuantepe FLYSPEC 5 448 2/27/05 Llano Pacaya FLYSPEC 1 315 2/27/05 Llano Pacaya FLYSPEC 2 291 2/27/05 Llano Pacaya FLYSPEC 3 285 2/27/05 Llano Pacaya FLYSPEC 4 206

Appendix B – Data

151


2/27/05 Llano Pacaya FLYSPEC 5 182 2/27/05 Llano Pacaya FLYSPEC 6 230 3/3/05 Ticuantepe FLYSPEC 1 110 3/3/05 Ticuantepe FLYSPEC 2 189 3/3/05 Ticuantepe FLYSPEC 3 161 3/3/05 Ticuantepe FLYSPEC 4 132 3/3/05 Ticuantepe FLYSPEC 5 73 3/3/05 Ticuantepe FLYSPEC 6 138 3/3/05 Ticuantepe FLYSPEC 7 185 3/5/05 Ticuantepe FLYSPEC 1 37 3/5/05 Ticuantepe FLYSPEC 2 63 3/5/05 Ticuantepe FLYSPEC 3 55 3/5/05 Ticuantepe FLYSPEC 4 62 3/5/05 Ticuantepe FLYSPEC 5 44 3/5/05 Ticuantepe FLYSPEC 6 75 3/5/05 Ticuantepe FLYSPEC 7 62 3/5/05 Ticuantepe FLYSPEC 8 49 3/5/05 Ticuantepe FLYSPEC 9 33 3/5/05 Ticuantepe FLYSPEC 11 55 3/6/05 Ticuantepe FLYSPEC 1 67 3/6/05 Ticuantepe FLYSPEC 2 69 3/6/05 Ticuantepe FLYSPEC 3 74 3/6/05 Ticuantepe FLYSPEC 4 50 3/6/05 Ticuantepe FLYSPEC 5 49 3/6/05 Ticuantepe FLYSPEC 6 52 3/6/05 Ticuantepe FLYSPEC 7 49 3/6/05 Ticuantepe FLYSPEC 8 37 3/6/05 Ticuantepe FLYSPEC 9 55 3/6/05 Ticuantepe FLYSPEC 10 46 3/6/05 Ticuantepe FLYSPEC 11 47 3/6/05 Ticuantepe FLYSPEC 12 47 3/6/05 Ticuantepe FLYSPEC 13 101 3/6/05 Ticuantepe FLYSPEC 14 105 3/6/05 Ticuantepe FLYSPEC 15 121 3/7/05 Ticuantepe FLYSPEC 1 43 3/7/05 Ticuantepe FLYSPEC 2 57 3/7/05 Ticuantepe FLYSPEC 3 45 3/7/05 Ticuantepe FLYSPEC 4 45 3/7/05 Ticuantepe FLYSPEC 5 58 3/7/05 Ticuantepe FLYSPEC 6 104 3/7/05 Ticuantepe FLYSPEC 7 73 3/7/05 Ticuantepe FLYSPEC 8 98 3/7/05 Ticuantepe FLYSPEC 9 85 3/7/05 Llano Pacaya FLYSPEC 1 34 3/7/05 Llano Pacaya FLYSPEC 2 72 3/11/05 Ticuantepe FLYSPEC 1 42 3/11/05 Ticuantepe FLYSPEC 3 94 3/11/05 Ticuantepe FLYSPEC 4 56

Appendix B – Data

152


3/11/05 Ticuantepe FLYSPEC 5 77 3/11/05 Ticuantepe FLYSPEC 6 81 3/11/05 Ticuantepe FLYSPEC 7 108 3/11/05 Ticuantepe FLYSPEC 8 36 3/11/05 Ticuantepe FLYSPEC 9 39 3/11/05 Ticuantepe FLYSPEC 10 42 3/11/05 Ticuantepe FLYSPEC 11 53 3/11/05 Ticuantepe FLYSPEC 12 88 3/12/05 Ticuantepe FLYSPEC 1 90 3/12/05 Ticuantepe FLYSPEC 2 54 3/12/05 Ticuantepe FLYSPEC 3 59 3/12/05 Ticuantepe FLYSPEC 4 82 3/12/05 Ticuantepe FLYSPEC 5 47 3/12/05 Ticuantepe FLYSPEC 6 69 3/12/05 Ticuantepe FLYSPEC 7 62 3/12/05 Ticuantepe FLYSPEC 8 105 3/12/05 Ticuantepe FLYSPEC 9 72 3/12/05 Llano Pacaya FLYSPEC 1 48 3/12/05 Llano Pacaya FLYSPEC 2 40 3/12/05 Llano Pacaya FLYSPEC 3 57 3/12/05 Llano Pacaya FLYSPEC 4 38 3/12/05 Llano Pacaya FLYSPEC 5 33 3/12/05 Llano Pacaya FLYSPEC 6 60 3/14/05 Ticuantepe FLYSPEC 1 49 3/14/05 Ticuantepe FLYSPEC 2 151 3/14/05 Ticuantepe FLYSPEC 3 44 3/14/05 Ticuantepe FLYSPEC 4 119 3/14/05 Ticuantepe FLYSPEC 5 79 3/14/05 Ticuantepe FLYSPEC 6 84 3/14/05 Ticuantepe FLYSPEC 7 34 3/14/05 Ticuantepe FLYSPEC 8 61 3/14/05 Ticuantepe FLYSPEC 9 120 3/14/05 Ticuantepe FLYSPEC 10 100 3/14/05 Ticuantepe FLYSPEC 11 88 3/14/05 Ticuantepe FLYSPEC 12 88 3/14/05 Llano Pacaya FLYSPEC 1 44 3/14/05 Llano Pacaya FLYSPEC 2 27 3/14/05 Llano Pacaya FLYSPEC 3 52 3/14/05 Llano Pacaya FLYSPEC 4 60 3/14/05 Llano Pacaya FLYSPEC 5 30 3/14/05 Llano Pacaya FLYSPEC 6 43 3/15/05 Ticuantepe FLYSPEC 1 61 3/15/05 Ticuantepe FLYSPEC 2 73 3/15/05 Ticuantepe FLYSPEC 3 126 3/15/05 Ticuantepe FLYSPEC 4 54 3/15/05 Ticuantepe FLYSPEC 5 66 3/15/05 Ticuantepe FLYSPEC 6 83 3/15/05 Ticuantepe FLYSPEC 7 82

Appendix B – Data

153


3/15/05 Llano Pacaya FLYSPEC 1 50 3/15/05 Llano Pacaya FLYSPEC 2 69 3/15/05 Llano Pacaya FLYSPEC 3 60 3/15/05 Llano Pacaya FLYSPEC 4 35 3/16/05 Ticuantepe FLYSPEC 1 209 3/16/05 Ticuantepe FLYSPEC 2 133 3/16/05 Ticuantepe FLYSPEC 3 106 3/16/05 Ticuantepe FLYSPEC 4 169 3/16/05 Ticuantepe FLYSPEC 5 48 3/16/05 Ticuantepe FLYSPEC 6 74 3/16/05 Ticuantepe FLYSPEC 7 65 3/16/05 Ticuantepe FLYSPEC 8 112 3/16/05 Ticuantepe FLYSPEC 9 99 3/16/05 Ticuantepe FLYSPEC 10 144 2/16/06 Ticuantepe FLYSPEC 1 126 2/16/06 Ticuantepe FLYSPEC 2 111 2/16/06 Ticuantepe FLYSPEC 3 61 2/16/06 Llano Pacaya FLYSPEC 1 88 2/16/06 Llano Pacaya FLYSPEC 2 46 2/17/06 Ticuantepe FLYSPEC 1 166 2/17/06 Ticuantepe FLYSPEC 2 152 2/17/06 Ticuantepe FLYSPEC 3 97 2/17/06 Ticuantepe FLYSPEC 4 156 2/17/06 Ticuantepe FLYSPEC 5 97 2/17/06 Ticuantepe FLYSPEC 6 169 2/17/06 Ticuantepe FLYSPEC 7 75 2/17/06 Ticuantepe FLYSPEC 8 81 2/17/06 Ticuantepe FLYSPEC 9 123 2/17/06 Llano Pacaya FLYSPEC 1 37 2/17/06 Llano Pacaya FLYSPEC 2 84 2/17/06 Llano Pacaya FLYSPEC 3 41 2/17/06 Llano Pacaya FLYSPEC 4 33 2/17/06 Llano Pacaya FLYSPEC 5 97 2/17/06 Llano Pacaya FLYSPEC 6 59 2/17/06 Llano Pacaya FLYSPEC 7 78 2/17/06 Llano Pacaya FLYSPEC 8 27 2/17/06 Llano Pacaya FLYSPEC 9 52 2/17/06 Llano Pacaya FLYSPEC 10 55 2/17/06 Ticuantepe FLYSPEC 10 79 2/17/06 Ticuantepe FLYSPEC 11 101 2/17/06 Ticuantepe FLYSPEC 12 93 2/17/06 Ticuantepe FLYSPEC 13 82 2/17/06 Ticuantepe FLYSPEC 14 117 2/18/06 Ticuantepe FLYSPEC 1 85 2/18/06 Ticuantepe FLYSPEC 2 118 2/18/06 Ticuantepe FLYSPEC 3 71 2/18/06 Ticuantepe FLYSPEC 4 57 2/18/06 Ticuantepe FLYSPEC 5 77

Appendix B – Data

154


2/18/06 Ticuantepe FLYSPEC 6 119 2/18/06 Ticuantepe FLYSPEC 7 61 2/18/06 Ticuantepe FLYSPEC 8 77 2/18/06 Ticuantepe FLYSPEC 9 106 2/18/06 Ticuantepe FLYSPEC 10 53 2/18/06 Ticuantepe FLYSPEC 11 119 2/18/06 Llano Pacaya FLYSPEC 1 48 2/18/06 Llano Pacaya FLYSPEC 2 37 2/18/06 Llano Pacaya FLYSPEC 3 54 2/18/06 Llano Pacaya FLYSPEC 4 26 2/18/06 Llano Pacaya FLYSPEC 5 52 2/18/06 Llano Pacaya FLYSPEC 6 40 2/18/06 Llano Pacaya FLYSPEC 7 35 2/18/06 Llano Pacaya FLYSPEC 8 41 2/18/06 Llano Pacaya FLYSPEC 9 55 2/18/06 Llano Pacaya FLYSPEC 10 53 2/18/06 Llano Pacaya FLYSPEC 11 41 2/18/06 Llano Pacaya FLYSPEC 12 50 2/19/06 Ticuantepe FLYSPEC 1 78 2/19/06 Ticuantepe FLYSPEC 2 144 2/19/06 Ticuantepe FLYSPEC 3 80 2/19/06 Ticuantepe FLYSPEC 4 80 2/19/06 Ticuantepe FLYSPEC 5 107 2/19/06 Ticuantepe FLYSPEC 6 32 2/19/06 Ticuantepe FLYSPEC 7 72 2/19/06 Ticuantepe FLYSPEC 8 65 2/19/06 Ticuantepe FLYSPEC 9 64 2/19/06 Ticuantepe FLYSPEC 10 96 2/19/06 Ticuantepe FLYSPEC 11 49 2/19/06 Llano Pacaya FLYSPEC 1 40 2/19/06 Llano Pacaya FLYSPEC 2 36 2/19/06 Llano Pacaya FLYSPEC 3 68 2/19/06 Llano Pacaya FLYSPEC 4 40 2/19/06 Llano Pacaya FLYSPEC 6 29 2/19/06 Llano Pacaya FLYSPEC 7 57 2/19/06 Llano Pacaya FLYSPEC 8 44 2/19/06 Llano Pacaya FLYSPEC 9 62 2/19/06 Llano Pacaya FLYSPEC 10 51 2/19/06 Llano Pacaya FLYSPEC 11 61 2/19/06 Llano Pacaya FLYSPEC 12 39 2/19/06 Ticuantepe FLYSPEC 12 117 2/20/06 Ticuantepe FLYSPEC 1 75 2/20/06 Ticuantepe FLYSPEC 2 64 2/20/06 Ticuantepe FLYSPEC 3 51 2/20/06 Llano Pacaya FLYSPEC 1 51 2/20/06 Llano Pacaya FLYSPEC 2 61 2/20/06 Llano Pacaya FLYSPEC 3 61 2/20/06 Llano Pacaya FLYSPEC 4 32

Appendix B – Data

155


2/20/06 Llano Pacaya FLYSPEC 5 72 2/20/06 Llano Pacaya FLYSPEC 6 40 2/20/06 Llano Pacaya FLYSPEC 7 66 2/20/06 Llano Pacaya FLYSPEC 8 51 2/20/06 Ticuantepe FLYSPEC 4 141 2/20/06 Ticuantepe FLYSPEC 5 43 2/20/06 Ticuantepe FLYSPEC 6 58 2/20/06 Ticuantepe FLYSPEC 7 104 2/20/06 Ticuantepe FLYSPEC 8 122 2/20/06 Ticuantepe FLYSPEC 9 93 2/20/06 Ticuantepe FLYSPEC 10 91 2/21/06 Ticuantepe FLYSPEC 1 102 2/21/06 Ticuantepe FLYSPEC 2 86 2/21/06 Ticuantepe FLYSPEC 3 108 2/21/06 Ticuantepe FLYSPEC 4 164 2/21/06 Ticuantepe FLYSPEC 5 111 2/21/06 Ticuantepe FLYSPEC 6 98 2/21/06 Ticuantepe FLYSPEC 7 94 2/21/06 Ticuantepe FLYSPEC 8 75 2/21/06 Ticuantepe FLYSPEC 9 100 2/21/06 Ticuantepe FLYSPEC 10 108 2/21/06 Ticuantepe FLYSPEC 11 78 2/21/06 Llano Pacaya FLYSPEC 1 51 2/21/06 Llano Pacaya FLYSPEC 2 47 2/21/06 Llano Pacaya FLYSPEC 3 79 2/21/06 Llano Pacaya FLYSPEC 4 52 2/21/06 Llano Pacaya FLYSPEC 5 53 2/21/06 Llano Pacaya FLYSPEC 6 38 2/21/06 Llano Pacaya FLYSPEC 7 53 2/21/06 Llano Pacaya FLYSPEC 8 59 2/21/06 Llano Pacaya FLYSPEC 9 66 2/21/06 Llano Pacaya FLYSPEC 10 40 2/21/06 Llano Pacaya FLYSPEC 11 29 2/21/06 Llano Pacaya FLYSPEC 12 43 2/21/06 Ticuantepe FLYSPEC 12 165 2/21/06 Ticuantepe FLYSPEC 13 107 2/21/06 Ticuantepe FLYSPEC 14 72 2/23/06 Ticuantepe FLYSPEC 1 164 2/23/06 Ticuantepe FLYSPEC 2 115 2/23/06 Ticuantepe FLYSPEC 3 192 2/23/06 Ticuantepe FLYSPEC 4 214 2/23/06 Ticuantepe FLYSPEC 5 122 2/23/06 Ticuantepe FLYSPEC 6 193 2/23/06 Ticuantepe FLYSPEC 7 147 2/23/06 Llano Pacaya FLYSPEC 1 132 2/23/06 Llano Pacaya FLYSPEC 2 128 2/23/06 Llano Pacaya FLYSPEC 3 50 2/23/06 Llano Pacaya FLYSPEC 4 91

Appendix B – Data

156


2/23/06 Llano Pacaya FLYSPEC 5 33 2/23/06 Llano Pacaya FLYSPEC 6 79 2/24/06 Ticuantepe FLYSPEC 1 152 2/24/06 Ticuantepe FLYSPEC 2 124 2/24/06 Ticuantepe FLYSPEC 3 84 2/24/06 Ticuantepe FLYSPEC 4 272 2/24/06 Ticuantepe FLYSPEC 5 100 2/24/06 Ticuantepe FLYSPEC 6 153 2/24/06 Ticuantepe FLYSPEC 7 158 2/24/06 Ticuantepe FLYSPEC 8 98 2/24/06 Ticuantepe FLYSPEC 9 85 2/24/06 Ticuantepe FLYSPEC 10 115 2/24/06 Ticuantepe FLYSPEC 11 86 2/24/06 Llano Pacaya FLYSPEC 3 94 2/24/06 Llano Pacaya FLYSPEC 4 59 2/24/06 Llano Pacaya FLYSPEC 5 60 2/24/06 Llano Pacaya FLYSPEC 6 50 2/24/06 Llano Pacaya FLYSPEC 7 73 2/24/06 Llano Pacaya FLYSPEC 8 87 2/24/06 Llano Pacaya FLYSPEC 9 54 2/24/06 Llano Pacaya FLYSPEC 10 48 2/24/06 Llano Pacaya FLYSPEC 11 113 2/24/06 Llano Pacaya FLYSPEC 12 90 2/24/06 Llano Pacaya FLYSPEC 13 52 2/24/06 Llano Pacaya FLYSPEC 14 83 2/26/06 Ticuantepe FLYSPEC V2 1* 117 2/26/06 Ticuantepe FLYSPEC V2 2* 380 2/26/06 Ticuantepe FLYSPEC V2 3* 200 2/26/06 Ticuantepe FLYSPEC V2 4* 197 2/26/06 Ticuantepe FLYSPEC V2 5* 50 2/26/06 Ticuantepe FLYSPEC V2 6* 124 2/26/06 Llano Pacaya FLYSPEC V2 1 53 2/26/06 Llano Pacaya FLYSPEC V2 2 66 2/26/06 Llano Pacaya FLYSPEC V2 3 108 2/26/06 Llano Pacaya FLYSPEC V2 4 137 2/26/06 Llano Pacaya FLYSPEC V2 5 110 2/26/06 Llano Pacaya FLYSPEC V2 6 87 2/26/06 Llano Pacaya FLYSPEC V2 7 174 2/26/06 Llano Pacaya FLYSPEC V2 8 57 2/26/06 Ticuantepe FLYSPEC 1* 116 2/26/06 Ticuantepe FLYSPEC 2* 377 2/26/06 Ticuantepe FLYSPEC 3* 186 2/26/06 Ticuantepe FLYSPEC 4* 190 2/26/06 Ticuantepe FLYSPEC 5* 54 2/26/06 Ticuantepe FLYSPEC 6* 134 2/26/06 Ticuantepe FLYSPEC 8 199 2/26/06 Ticuantepe FLYSPEC 9 86 2/26/06 Ticuantepe FLYSPEC 10 178

Appendix B – Data

157


2/26/06 Ticuantepe FLYSPEC 11 124 2/26/06 Ticuantepe FLYSPEC 12 133 2/26/06 Ticuantepe FLYSPEC 13 221 2/26/06 Ticuantepe FLYSPEC 14 75 2/26/06 Ticuantepe FLYSPEC 15 29 2/26/06 Ticuantepe FLYSPEC 17 167 2/26/06 Ticuantepe FLYSPEC 18 110 2/26/06 Ticuantepe FLYSPEC 19 184 2/27/06 Llano Pacaya FLYSPEC 1 56 2/27/06 Llano Pacaya FLYSPEC 2 43 2/27/06 Llano Pacaya FLYSPEC 3 71 2/27/06 Llano Pacaya FLYSPEC 4 60 2/27/06 Llano Pacaya FLYSPEC 6 44 2/27/06 Llano Pacaya FLYSPEC 7 80 2/27/06 Llano Pacaya FLYSPEC 12 54 2/27/06 Llano Pacaya FLYSPEC 13 51 2/27/06 Llano Pacaya FLYSPEC 15 60 2/27/06 Llano Pacaya FLYSPEC 16 109 2/27/06 Llano Pacaya FLYSPEC 17 62 2/27/06 Ticuantepe FLYSPEC V2 1 119 2/27/06 Ticuantepe FLYSPEC V2 2 240 2/27/06 Ticuantepe FLYSPEC V2 3 87 2/27/06 Ticuantepe FLYSPEC V2 4 103 2/27/06 Ticuantepe FLYSPEC V2 5 51 2/27/06 Ticuantepe FLYSPEC V2 6 72 2/27/06 Ticuantepe FLYSPEC V2 7 42 2/27/06 Ticuantepe FLYSPEC V2 8 75 2/27/06 Ticuantepe FLYSPEC V2 9 152 2/27/06 Ticuantepe FLYSPEC V2 10 116 2/27/06 Ticuantepe FLYSPEC V2 12 113 2/27/06 Ticuantepe FLYSPEC V2 13 164 2/27/06 Ticuantepe FLYSPEC V2 14 134 2/27/06 Ticuantepe FLYSPEC V2 15 79 2/27/06 Ticuantepe FLYSPEC V2 16 73 2/27/06 Ticuantepe FLYSPEC V2 17 66 2/27/06 Ticuantepe FLYSPEC V2 18 66 2/27/06 Ticuantepe FLYSPEC V2 19 47 2/27/06 Ticuantepe FLYSPEC V2 20 100 2/27/06 Ticuantepe FLYSPEC V2 21 56 2/28/06 Ticuantepe FLYSPEC 1 227 2/28/06 Ticuantepe FLYSPEC 2 177 2/28/06 Ticuantepe FLYSPEC 3 149 2/28/06 Ticuantepe FLYSPEC 4 167 2/28/06 Ticuantepe FLYSPEC 5 140 2/28/06 Ticuantepe FLYSPEC 6 140 2/28/06 Ticuantepe FLYSPEC 7 125 2/28/06 Ticuantepe FLYSPEC 8 164 2/28/06 Ticuantepe FLYSPEC 9 99

Appendix B – Data

158


2/28/06 Ticuantepe FLYSPEC 10 211 2/28/06 Llano Pacaya FLYSPEC 1 71 2/28/06 Llano Pacaya FLYSPEC 2 34 2/28/06 Llano Pacaya FLYSPEC 3 80 2/28/06 Llano Pacaya FLYSPEC 4 84 2/28/06 Llano Pacaya FLYSPEC 5 41 2/28/06 Llano Pacaya FLYSPEC 6 64 2/28/06 Llano Pacaya FLYSPEC 8 69 2/28/06 Llano Pacaya FLYSPEC 9 44 2/28/06 Llano Pacaya FLYSPEC 10 59 2/28/06 Llano Pacaya FLYSPEC 11 56 2/28/06 Ticuantepe FLYSPEC 12 92 2/28/06 Ticuantepe FLYSPEC 13 138 2/28/06 Ticuantepe FLYSPEC 14 139 3/1/06 Ticuantepe FLYSPEC 1 195 3/1/06 Ticuantepe FLYSPEC 2 193 3/1/06 Ticuantepe FLYSPEC 3 125 3/1/06 Ticuantepe FLYSPEC 4 112 3/1/06 Ticuantepe FLYSPEC 5 161 3/1/06 Ticuantepe FLYSPEC 6 273 3/1/06 Ticuantepe FLYSPEC 7 101 3/1/06 Ticuantepe FLYSPEC 8 101 3/1/06 Ticuantepe FLYSPEC 9 174 3/1/06 Ticuantepe FLYSPEC 10 150 3/1/06 Ticuantepe FLYSPEC 11 176 3/1/06 Ticuantepe FLYSPEC 12 229 3/1/06 Ticuantepe FLYSPEC 13 128 3/1/06 Ticuantepe FLYSPEC 14 150 3/1/06 Ticuantepe FLYSPEC 15 70 3/1/06 Ticuantepe FLYSPEC 16 240 3/1/06 Ticuantepe FLYSPEC 17 43 3/1/06 Ticuantepe FLYSPEC 18 132 3/1/06 Ticuantepe FLYSPEC 20 140 3/1/06 Ticuantepe FLYSPEC 21 187 3/1/06 Ticuantepe FLYSPEC 22 181 3/1/06 Ticuantepe FLYSPEC 23 131 3/1/06 Ticuantepe FLYSPEC 24 190 3/1/06 Llano Pacaya FLYSPEC V2 1 95 3/1/06 Llano Pacaya FLYSPEC V2 2 81 3/1/06 Llano Pacaya FLYSPEC V2 3 40 3/1/06 Llano Pacaya FLYSPEC V2 4 79 3/1/06 Llano Pacaya FLYSPEC V2 5 58 3/1/06 Llano Pacaya FLYSPEC V2 6 70 3/1/06 Llano Pacaya FLYSPEC V2 7 79 3/1/06 Llano Pacaya FLYSPEC V2 8 87 3/2/06 Llano Pacaya FLYSPEC V2 1 68 3/2/06 Llano Pacaya FLYSPEC V2 2 64 3/2/06 Llano Pacaya FLYSPEC V2 3 90

Appendix B – Data

159


3/2/06 Llano Pacaya FLYSPEC V2 4 53 3/2/06 Llano Pacaya FLYSPEC V2 5 73 3/2/06 Llano Pacaya FLYSPEC V2 6 85 3/2/06 Llano Pacaya FLYSPEC V2 7 84 3/2/06 Llano Pacaya FLYSPEC V2 8 49 3/2/06 Llano Pacaya FLYSPEC V2 9 64 3/2/06 Ticuantepe FLYSPEC V2 1 95 3/2/06 Ticuantepe FLYSPEC V2 2 128 3/2/06 Ticuantepe FLYSPEC V2 3 206 3/2/06 Ticuantepe FLYSPEC V2 4 133 3/2/06 Ticuantepe FLYSPEC V2 5 210 3/2/06 Ticuantepe FLYSPEC V2 6 150 3/2/06 Ticuantepe FLYSPEC V2 7 99 3/2/06 Ticuantepe FLYSPEC V2 8 200 3/2/06 Ticuantepe FLYSPEC V2 9 144 3/2/06 Ticuantepe FLYSPEC 1 243 3/2/06 Ticuantepe FLYSPEC 2 179 3/2/06 Ticuantepe FLYSPEC 3 195 3/2/06 Llano Pacaya FLYSPEC 2 78 3/2/06 Llano Pacaya FLYSPEC 3 174 3/2/06 Llano Pacaya FLYSPEC 4 36 3/2/06 Llano Pacaya FLYSPEC 5 77 3/2/06 Llano Pacaya FLYSPEC 6 129 3/2/06 Llano Pacaya FLYSPEC 7 106 3/2/06 Llano Pacaya FLYSPEC 8 136 3/2/06 Llano Pacaya FLYSPEC 9 115 3/2/06 Llano Pacaya FLYSPEC 10 118 3/2/06 Llano Pacaya FLYSPEC 11 137 3/2/06 Llano Pacaya FLYSPEC 12 101 3/2/06 Llano Pacaya FLYSPEC 13 111 3/2/06 Llano Pacaya FLYSPEC 14 107 3/2/06 Llano Pacaya FLYSPEC 15 97 3/3/06 Ticuantepe FLYSPEC V2 3 167 3/3/06 Ticuantepe FLYSPEC V2 5 182 3/3/06 Ticuantepe FLYSPEC V2 6 70 3/3/06 Ticuantepe FLYSPEC V2 7 164 3/3/06 Ticuantepe FLYSPEC V2 8 232 3/3/06 Ticuantepe FLYSPEC V2 9 130 3/3/06 Ticuantepe FLYSPEC V2 10 103 3/3/06 Ticuantepe FLYSPEC V2 11 126 3/3/06 Ticuantepe FLYSPEC V2 12 161 3/3/06 Ticuantepe FLYSPEC V2 13 184 3/3/06 Ticuantepe FLYSPEC V2 14 169 3/3/06 Ticuantepe FLYSPEC V2 15 106 3/3/06 Ticuantepe FLYSPEC V2 16 142 3/3/06 Ticuantepe FLYSPEC V2 17 136 3/3/06 Ticuantepe FLYSPEC V2 18 128 3/3/06 Ticuantepe FLYSPEC V2 19 130

Appendix B – Data

160


3/3/06 Llano Pacaya FLYSPEC 1 125 3/3/06 Llano Pacaya FLYSPEC 2 99 3/3/06 Llano Pacaya FLYSPEC 3 99 3/3/06 Llano Pacaya FLYSPEC 4 85 3/3/06 Llano Pacaya FLYSPEC 5 81 3/3/06 Llano Pacaya FLYSPEC 6 76 3/3/06 Llano Pacaya FLYSPEC 7 83 3/3/06 Llano Pacaya FLYSPEC 8 85 3/3/06 Llano Pacaya FLYSPEC 9 78 3/3/06 Llano Pacaya FLYSPEC 10 104 3/3/06 Llano Pacaya FLYSPEC 11 93 3/3/06 Llano Pacaya FLYSPEC 12 82 3/3/06 Llano Pacaya FLYSPEC 13 100 3/3/06 Llano Pacaya FLYSPEC 14 106 3/3/06 Llano Pacaya FLYSPEC 15 162 3/3/06 Llano Pacaya FLYSPEC 16 111 3/3/06 Llano Pacaya FLYSPEC 17 117 3/3/06 Llano Pacaya FLYSPEC 18 82 3/3/06 Llano Pacaya FLYSPEC V2 1 106 3/3/06 Llano Pacaya FLYSPEC V2 2 105 3/3/06 Llano Pacaya FLYSPEC V2 3 55 3/3/06 Llano Pacaya FLYSPEC V2 4 103 3/3/06 Llano Pacaya FLYSPEC V2 5 130 3/3/06 Llano Pacaya FLYSPEC V2 7 124 3/3/06 Llano Pacaya FLYSPEC V2 8 98 3/3/06 Ticuantepe FLYSPEC 1 220 3/3/06 Ticuantepe FLYSPEC 2 119 3/3/06 Ticuantepe FLYSPEC 3 66 3/3/06 Ticuantepe FLYSPEC 4 142 3/3/06 Ticuantepe FLYSPEC 5 219 3/3/06 Ticuantepe FLYSPEC 6 244 3/3/06 Ticuantepe FLYSPEC 7 140 3/3/06 Ticuantepe FLYSPEC 8 188 3/3/06 Ticuantepe FLYSPEC 9 192 3/4/06 Llano Pacaya FLYSPEC V2 1 117 3/4/06 Llano Pacaya FLYSPEC V2 2 52 3/4/06 Llano Pacaya FLYSPEC V2 3 51 3/4/06 Llano Pacaya FLYSPEC V2 4 60 3/4/06 Llano Pacaya FLYSPEC V2 5 74 3/4/06 Llano Pacaya FLYSPEC V2 6 66 3/4/06 Llano Pacaya FLYSPEC V2 8 93 3/4/06 Llano Pacaya FLYSPEC V2 9 95 3/4/06 Llano Pacaya FLYSPEC V2 10 83 3/4/06 Llano Pacaya FLYSPEC V2 11 64 3/4/06 Llano Pacaya FLYSPEC V2 12 99 3/4/06 Llano Pacaya FLYSPEC V2 13 54 3/4/06 Llano Pacaya FLYSPEC V2 14 91 3/4/06 Llano Pacaya FLYSPEC V2 15 92

Appendix B – Data

161


3/4/06 Ticuantepe FLYSPEC 1 160 3/4/06 Ticuantepe FLYSPEC 2 167 3/4/06 Ticuantepe FLYSPEC 3 190 3/4/06 Ticuantepe FLYSPEC 4 118 3/4/06 Ticuantepe FLYSPEC 5 113 3/4/06 Ticuantepe FLYSPEC 6 114 3/4/06 Ticuantepe FLYSPEC 7 89 3/4/06 Ticuantepe FLYSPEC 8 142 3/4/06 Ticuantepe FLYSPEC 9 199 3/4/06 Ticuantepe FLYSPEC 10 137 3/4/06 Ticuantepe FLYSPEC V2 1 81 3/4/06 Ticuantepe FLYSPEC V2 2 61 3/4/06 Ticuantepe FLYSPEC V2 3 131 3/4/06 Ticuantepe FLYSPEC V2 4 77 3/4/06 Ticuantepe FLYSPEC V2 5 75 3/4/06 Ticuantepe FLYSPEC V2 6 99 3/4/06 Ticuantepe FLYSPEC V2 7 53 3/4/06 Ticuantepe FLYSPEC V2 8 68 3/4/06 Ticuantepe FLYSPEC V2 9 118 3/4/06 Ticuantepe FLYSPEC V2 10 157 3/4/06 Ticuantepe FLYSPEC V2 11 208 3/4/06 Ticuantepe FLYSPEC V2 12 109 3/4/06 Ticuantepe FLYSPEC V2 13 143 3/4/06 Ticuantepe FLYSPEC V2 14 101 3/4/06 Ticuantepe FLYSPEC V2 15 111 3/4/06 Ticuantepe FLYSPEC V2 16 107 3/4/06 Ticuantepe FLYSPEC V2 17 104 3/4/06 Ticuantepe FLYSPEC V2 18 163 3/4/06 Ticuantepe FLYSPEC V2 19 65 3/4/06 Llano Pacaya FLYSPEC 1 102 3/4/06 Llano Pacaya FLYSPEC 3 75 3/4/06 Llano Pacaya FLYSPEC 4 88 3/4/06 Llano Pacaya FLYSPEC 5 70 3/4/06 Llano Pacaya FLYSPEC 6 37 3/4/06 Llano Pacaya FLYSPEC 7 36 3/4/06 Llano Pacaya FLYSPEC 8 43 3/4/06 Llano Pacaya FLYSPEC 9 60 3/4/06 Llano Pacaya FLYSPEC 10 134 3/4/06 Llano Pacaya FLYSPEC 11 154 3/6/06 Llano Pacaya FLYSPEC V2 1 65 3/6/06 Llano Pacaya FLYSPEC V2 2 54 3/6/06 Llano Pacaya FLYSPEC V2 3 29 3/6/06 Llano Pacaya FLYSPEC V2 4 49 3/6/06 Llano Pacaya FLYSPEC V2 5 40 3/6/06 Llano Pacaya FLYSPEC V2 6 60 3/6/06 Llano Pacaya FLYSPEC V2 7 45 3/6/06 Llano Pacaya FLYSPEC V2 8 27 3/6/06 Llano Pacaya FLYSPEC V2 9 47

Appendix B – Data

162


3/6/06 Llano Pacaya FLYSPEC V2 10 29 3/6/06 Llano Pacaya FLYSPEC V2 11 72 3/6/06 Llano Pacaya FLYSPEC V2 12 25 3/6/06 Llano Pacaya FLYSPEC V2 13 37 3/6/06 Llano Pacaya FLYSPEC V2 14 49 3/6/06 Llano Pacaya FLYSPEC V2 15 69 3/6/06 Llano Pacaya FLYSPEC V2 16 63 3/6/06 Llano Pacaya FLYSPEC V2 17 33 3/6/06 Llano Pacaya FLYSPEC V2 18 54 3/6/06 Llano Pacaya FLYSPEC V2 19 61 3/6/06 Llano Pacaya FLYSPEC V2 20 57 3/6/06 Llano Pacaya FLYSPEC V2 21 26 3/6/06 Llano Pacaya FLYSPEC V2 22 93 3/6/06 Ticuantepe FLYSPEC 1 82 3/6/06 Ticuantepe FLYSPEC V2 1 47 3/6/06 Ticuantepe FLYSPEC V2 2 90 3/6/06 Ticuantepe FLYSPEC V2 3 140 3/6/06 Ticuantepe FLYSPEC V2 4 139 3/6/06 Ticuantepe FLYSPEC V2 5 113 3/6/06 Ticuantepe FLYSPEC V2 6 121 3/6/06 Ticuantepe FLYSPEC V2 7 92 3/6/06 Llano Pacaya FLYSPEC 2 67 3/6/06 Llano Pacaya FLYSPEC 3 53 3/6/06 Llano Pacaya FLYSPEC 4 82 3/6/06 Llano Pacaya FLYSPEC 5 91 3/6/06 Llano Pacaya FLYSPEC 6 28 3/6/06 Llano Pacaya FLYSPEC 7 33 3/6/06 Llano Pacaya FLYSPEC 8 100 3/6/06 Llano Pacaya FLYSPEC 9 109 3/6/06 Llano Pacaya FLYSPEC 11 41 3/7/06 Llano Pacaya FLYSPEC 1 95 3/7/06 Llano Pacaya FLYSPEC 2 61 3/7/06 Llano Pacaya FLYSPEC 3 74 3/7/06 Llano Pacaya FLYSPEC 4 72 3/7/06 Llano Pacaya FLYSPEC 5 48 3/7/06 Llano Pacaya FLYSPEC 6 59 3/7/06 Llano Pacaya FLYSPEC 7 71 3/7/06 Llano Pacaya FLYSPEC 8 56 3/7/06 Llano Pacaya FLYSPEC 10 63 3/7/06 Ticuantepe FLYSPEC 1 148 3/7/06 Ticuantepe FLYSPEC 2 64 3/7/06 Ticuantepe FLYSPEC 3 121 3/7/06 Ticuantepe FLYSPEC 4 98 3/7/06 Ticuantepe FLYSPEC 5 159 3/7/06 Ticuantepe FLYSPEC 6 127 3/8/06 Ticuantepe FLYSPEC V2 1 97 3/8/06 Ticuantepe FLYSPEC V2 2 116 3/8/06 Ticuantepe FLYSPEC V2 3 97

Appendix B – Data

163


3/8/06 Ticuantepe FLYSPEC V2 4 111 3/8/06 Ticuantepe FLYSPEC V2 5 85 3/8/06 Ticuantepe FLYSPEC V2 6 99 3/8/06 Ticuantepe FLYSPEC V2 7 87 3/8/06 Ticuantepe FLYSPEC V2 8 123 3/8/06 Ticuantepe FLYSPEC V2 9 97 3/8/06 Ticuantepe FLYSPEC V2 10 91 3/8/06 Ticuantepe FLYSPEC V2 11 88 3/8/06 Llano Pacaya FLYSPEC 1 63 3/8/06 Llano Pacaya FLYSPEC 2 71 3/8/06 Llano Pacaya FLYSPEC 3 95 3/8/06 Llano Pacaya FLYSPEC 4 105 3/8/06 Llano Pacaya FLYSPEC 5 45 3/8/06 Llano Pacaya FLYSPEC 6 28 3/8/06 Llano Pacaya FLYSPEC 7 45 3/8/06 Llano Pacaya FLYSPEC 8 42 3/8/06 Llano Pacaya FLYSPEC 9 74 3/8/06 Llano Pacaya FLYSPEC 10 56 3/8/06 Llano Pacaya FLYSPEC 11 64 3/8/06 Llano Pacaya FLYSPEC 12 64 3/8/06 Llano Pacaya FLYSPEC 13 54 3/8/06 Llano Pacaya FLYSPEC 14 55 3/8/06 Llano Pacaya FLYSPEC V2 1 69 3/8/06 Llano Pacaya FLYSPEC V2 2 67 3/8/06 Llano Pacaya FLYSPEC V2 3 77 3/8/06 Llano Pacaya FLYSPEC V2 4 78 3/8/06 Llano Pacaya FLYSPEC V2 5 70 3/8/06 Llano Pacaya FLYSPEC V2 6 54 3/8/06 Ticuantepe FLYSPEC 1 181 3/8/06 Ticuantepe FLYSPEC 2 88 3/8/06 Ticuantepe FLYSPEC 3 177 3/8/06 Ticuantepe FLYSPEC 4 127 3/8/06 Ticuantepe FLYSPEC 6 157 3/8/06 Ticuantepe FLYSPEC 7 114 3/8/06 Ticuantepe FLYSPEC 8 111 3/8/06 Ticuantepe FLYSPEC 9 60 3/8/06 Ticuantepe FLYSPEC 10 238 3/8/06 Ticuantepe FLYSPEC 11 145 3/8/06 Ticuantepe FLYSPEC 12 116 3/8/06 Ticuantepe FLYSPEC 13 97 3/8/06 Ticuantepe FLYSPEC 14 113 3/9/06 Ticuantepe FLYSPEC 1 170 3/9/06 Ticuantepe FLYSPEC 2 130 3/9/06 Ticuantepe FLYSPEC 3 190 3/9/06 Ticuantepe FLYSPEC 4 162 3/9/06 Ticuantepe FLYSPEC 5 126 3/9/06 Ticuantepe FLYSPEC 6 195 3/9/06 Ticuantepe FLYSPEC 7 174

Appendix B – Data

164


3/9/06 Llano Pacaya FLYSPEC 1 94 3/9/06 Llano Pacaya FLYSPEC 2 96 3/9/06 Llano Pacaya FLYSPEC 3 74 3/9/06 Llano Pacaya FLYSPEC 4 90 3/9/06 Llano Pacaya FLYSPEC 6 71 3/9/06 Ticuantepe FLYSPEC 8 204 3/9/06 Ticuantepe FLYSPEC 9 94 3/9/06 Ticuantepe FLYSPEC 10 137 3/9/06 Ticuantepe FLYSPEC 11 172 3/9/06 Ticuantepe FLYSPEC 12 173 3/10/06 Llano Pacaya FLYSPEC 1 146 3/10/06 Llano Pacaya FLYSPEC 2 131 3/10/06 Llano Pacaya FLYSPEC 3 105 3/10/06 Llano Pacaya FLYSPEC 4 123 3/10/06 Llano Pacaya FLYSPEC 5 119 3/10/06 Llano Pacaya FLYSPEC 6 178 3/10/06 Ticuantepe FLYSPEC 1 193 3/10/06 Ticuantepe FLYSPEC 2 158 3/10/06 Ticuantepe FLYSPEC 3 129 3/10/06 Ticuantepe FLYSPEC 4 111 3/10/06 Ticuantepe FLYSPEC 5 190 3/10/06 Ticuantepe FLYSPEC 6 248 3/10/06 Ticuantepe FLYSPEC 7 118 3/10/06 Ticuantepe FLYSPEC 8 233 3/10/06 Ticuantepe FLYSPEC 9 201 3/12/06 Llano Pacaya FLYSPEC 1 67 3/12/06 Llano Pacaya FLYSPEC 2 113 3/12/06 Llano Pacaya FLYSPEC 3 66 3/12/06 Llano Pacaya FLYSPEC 4 96 3/12/06 Llano Pacaya FLYSPEC 5 86 3/12/06 Llano Pacaya FLYSPEC 6 74 3/12/06 Llano Pacaya FLYSPEC 7 82 3/12/06 Llano Pacaya FLYSPEC 8 91 3/12/06 Llano Pacaya FLYSPEC 9 65 3/12/06 Ticuantepe FLYSPEC 1 176 3/12/06 Ticuantepe FLYSPEC 2 263 3/12/06 Ticuantepe FLYSPEC 3 174 3/12/06 Ticuantepe FLYSPEC 4 113 3/12/06 Ticuantepe FLYSPEC 5 236 3/12/06 Ticuantepe FLYSPEC 6 106 3/12/06 Ticuantepe FLYSPEC 7 188

Appendix B – Data

165

Figure B-1: Map of road network in the vicinity of Masaya volcano. Roads utilized for measuring SO2 flux, the Ticuantepe and Llano Pacaya roads, are shown.

Appendix B – Data

166

B.2 Previous SO2 fluxes (1972-2004) (Data compiled from various studies made using a range of instruments, methods, and locations.)

Mean Daily Flux, presented by day

Date

Mean Daily Flux (t/d)

1 Std. Dev.

Average Wind

Speed

Normalized Mean Flux @ 1 m/s

(t/d)

1 Std. Dev.

n Method Location

10/1/1972 180 12/8/1976 660 215 5.4 122 40 3 Scan 11/26/1977 400 50 3.9 102 13 4 Scan 6/17/1978 326 100 5.7 57 18 10 Scan 6/18/1978 307 63 5.0 61 13 6 Scan 2/7/1980 1557 315 11.5 135 27 3 Road LP 2/10/1980 1252 217 7.7 165 34 3 Road LP 6/24/1980 930 115 4 Road LP 6/28/1980 614 1 Airplane 8/1/1980 1038 352 5 Road LP 8/2/1980 1174 4.5 261 1 Airplane 8/5/1980 255 134 2.7 94 50 2 Road LP 8/6/1980 1433 104 4.4 326 24 2 Airplane 8/15/1980 1533 343 2 Road LP 11/17/1980 729 160 3.4 214 47 4 Road LP 1/20/1981 878 189 3.0 309 52 8 Road LP 1/21/1981 1205 403 4.2 327 116 7 Road LP 1/22/1981 973 805 3.9 256 222 4 Road LP 1/23/1981 1353 153 4.3 318 11 3 Road LP 1/24/1981 4138 1116 1.8 2369 32 2 Road LP 1/25/1981 1127 225 3.9 295 80 5 Road LP 1/27/1981 1062 254 3.3 326 17 4 Road LP 1/28/1981 1350 741 3.6 352 124 12 Road LP 1/29/1981 1115 344 7.2 157 40 9 Road LP 1/30/1981 672 198 5.6 120 36 6 Road LP 1/31/1981 938 210 6.3 150 25 6 Road LP 2/1/1981 583 39 3.7 160 28 3 Road LP 2/3/1981 839 142 7.2 119 16 4 Road LP 2/5/1981 1125 12 9.6 117 1 2 Road LP 2/12/1981 724 355 8.5 87 44 4 Road LP 2/13/1981 817 193 7.8 106 26 4 Road LP 2/17/1981 535 54 4.3 124 12 2 Road LP 3/1/1981 1279 292 9.2 141 31 5 Road LP 3/19/1981 2015 260 4.9 411 53 2 Road LP 11/26/1981 764 281 4.6 174 51 3 Road 11/27/1981 542 51 2.7 204 39 4 Road 2/1/1982 597 338 3.5 168 16 5 Road LP 2/8/1982 838 639 3.4 227 67 4 Road LP 4/26/1992 10

n = number of measurements, LP = Llano Pacaya, T = Ticuantepe, MM = Masaya-Managua Highway

Appendix B – Data

167

Date


1 Std. Dev.

Average Wind

Speed


(t/d)

1 Std. Dev.

n Method Location

3/16/1996 603 288 3.9 154 74 8 Road MM 2/12/1997 159 73 3.5 45 21 4 Road T 2/13/1997 474 338 2.4 189 106 7 Road T 2/14/1997 312 65 3.9 80 17 4 Road T 3/7/1997 407 54 9.2 45 7 4 Road T 3/7/1997 334 91 10.0 34 12 4 Road LP 3/12/1997 437 186 7.5 57 18 2 Road T 3/12/1997 327 75 8.1 41 10 12 Road LP 3/25/1997 777 390 5.9 127 45 2 Road T 3/25/1997 349 30 6.2 57 6 3 Road LP 3/27/1997 479 343 4.8 96 55 4 Road T 3/27/1997 351 104 5.1 69 21 2 Road LP 3/28/1997 488 66 6.6 75 11 3 Road T 3/28/1997 545 94 7.4 74 13 3 Road LP 2/21/1998 1518 467 8.8 144 62 6 Road LP 2/22/1998 1897 859 6.3 293 109 8 Road LP 2/23/1998 3663 275 9.1 403 30 4 Road LP 2/24/1998 1517 400 9.2 117 22 9 Road LP 2/25/1998 1279 236 7.9 165 42 10 Road LP 3/1/1998 6118 1568 4.9 4 Road LP 3/2/1998 1963 547 8.0 218 76 4 Road LP 3/3/1998 675 143 4.7 144 31 3 Road LP 3/7/1998 953 276 5.8 164 48 8 Road LP 3/8/1998 2876 757 5.8 496 130 3 Road LP 3/10/1998 752 266 6.4 118 36 4 Road LP 3/13/1998 699 258 10.5 67 25 12 Road LP 3/14/1998 1800 409 16.2 105 22 6 Road LP 3/17/1998 3114 964 10.6 222 61 20 Road T 3/17/1998 2527 514 10.5 240 43 8 Road LP 3/18/1998 1627 425 8.1 153 33 10 Road LP 3/21/1998 1536 731 4.6 322 128 11 Road LP 3/25/1998 1092 345 8.8 125 40 10 Road LP 4/11/1998 1989 917 10.8 183 65 11 Road LP 4/17/1998 3016 551 11.2 270 53 9 Road LP 4/18/1998 2489 286 11.0 227 33 7 Road LP 4/20/1998 2369 450 12.3 193 36 6 Road LP 4/23/1998 1578 485 12.8 124 40 6 Road LP 4/27/1998 1388 225 12.8 108 18 7 Road LP 9/7/1998 1054 231 2.4 439 96 2 Road MM 9/8/1998 1422 284 3.4 418 84 4 Road MM 9/9/1998 316 142 1.2 265 117 6 Road MM 9/10/1998 1074 461 3.4 332 169 4 Road MM 9/11/1998 1155 643 2.1 550 306 4 Road MM 9/16/1998 231 37 0.5 462 75 8 Road MM


Appendix B – Data

168

Date


1 Std. Dev.

Average Wind

Speed


(t/d)

1 Std. Dev.

n Method Location

9/17/1998 429 100 1.0 429 100 7 Road MM 2/18/1999 4287 919 6.1 703 151 8 Road LP 2/19/1999 1861 502 3.1 675 413 10 Road LP 2/22/1999 1460 491 9.9 146 42 11 Road LP 2/23/1999 1700 357 9.2 186 36 10 Road LP 2/27/1999 2184 852 10.2 214 87 10 Road LP 3/2/1999 1939 618 5.7 340 117 10 Road LP 3/4/1999 1318 316 11.2 118 29 10 Road LP 3/5/1999 1335 314 10.9 123 27 16 Road LP 3/7/1999 1928 670 11.6 164 52 10 Road LP 3/10/1999 1251 279 10.5 120 31 11 Road LP 3/11/1999 1475 270 9.4 157 24 10 Road LP 3/13/1999 2085 406 7.4 287 73 7 Road LP 3/17/2000 1404 389 8.6 164 49 8 Road LP 3/18/2000 1282 229 8.6 148 25 8 Road LP 3/22/2000 1367 243 9.6 143 29 8 Road LP 3/24/2000 1850 456 10.0 164 40 8 Road LP 3/31/2000 1170 334 8.3 126 36 6 Road LP 4/1/2000 1846 305 12.5 133 20 6 Road LP 4/2/2000 987 242 10.2 88 23 8 Road LP 4/4/2000 1146 370 10.2 170 49 8 Road LP 4/6/2000 1112 233 11.4 99 26 10 Road LP 4/7/2000 1166 544 10.4 113 54 6 Road LP 4/10/2000 742 204 10.8 68 13 2 Road LP 4/11/2000 1469 8.7 169 1 Road LP 2/3/2001 421 0 3.9 108 0 2 Road LP 2/8/2001 261 10 5.3 49 2 2 Road LP 3/11/2001 757 284 4.8 158 59 8 Road LP 3/12/2001 526 141 3.4 155 42 10 Road LP 5/5/2001 320 3 Road T

12/14/2001 380 1 Road T

12/16/2001 372 4 Walking ~1 km from vent

12/18/2001 570 2 Road T 12/18/2001 484 1 Road LP 2/22/2002 224 122 4.1 55 30 4 Road LP 2/24/2002 656 294 9.2 72 25 3 Road LP 2/25/2002 572 116 10.0 57 12 4 Road LP 2/26/2002 468 187 7.3 65 26 5 Road LP 3/1/2002 475 146 11.4 42 13 6 Road T 3/1/2002 759 11.4 67 1 Road LP 3/24/2003 225 50 1.6 140 31 3 Road T 3/24/2003 190 8 1.6 119 5 3 Road LP 3/25/2003 863 401 7.1 122 25 4 Road T


Appendix B – Data

169

Date


1 Std. Dev.

Average Wind

Speed


(t/d)

1 Std. Dev.

n Method Location

3/25/2003 177 32 3.0 59 11 3 Road LP 3/28/2003 2651 943 10.5 253 90 2 Road T 3/29/2003 517 88 3.2 162 28 2 Road T 1/25/2004 682 316 3.3 209 97 12 Road T 1/25/2004 658 16 3.3 202 5 2 Road LP 1/31/2004 2024 3.4 596 1 Road T 1/31/2004 1919 643 3.4 564 189 5 Road LP


*1972 data from Stoiber and Jepsen (1973); 1976-1982 data from Stoiber and Williams (1986); 1992 data from Smithsonian Insitution (1992); 1996 data from Rymer et al. (1998); 1997-March 2001 data from Williams-Jones et al. (2003); May 2001 data from Galle et al. (2002); December 2001 data from McGonigle et al. (2002); 2002-2004 data from Williams-Jones (unpublished)

Appendix B – Data

170

Mean Daily Flux, presented by month

Month Mean Daily Flux (t/d)

1 Std. Dev. (t/d) n

October 1972 180 December 1976 660 215 3 November 1977 400 50 4

June 1978 319 86 16 February 1980 1405 294 6

June 1980 867 173 5 August 1980 1067 493 12

November 1980 729 160 4 January 1981 1174 714 66 February 1981 768 243 19

March 1981 1489 444 7 November 1981 637 204 7 February 1982 704 476 9

April 1992 10 March 1996 603 288 8

February 1997 347 263 15 March 1997 412 175 39

February 1998 1775 841 37 March 1998 1936 1361 103 April 1998 2171 788 46

September 1998 670 520 35 February 1999 2209 1154 49

March 1999 1572 517 74 March 2000 1428 399 38 April 2000 1202 429 41

February 2001 341 93 4 March 2001 629 240 18 May 2001 320 23 3

December 2001 452 94 8 February 2002 468 229 16

March 2002 516 171 7 March 2003 680 846 17

January 2004 1056 704 20

*1972 data from Stoiber and Jepsen (1973); 1976-1982 data from Stoiber and Williams (1986); 1992 data from Smithsonian Insitution (1992); 1996 data from Rymer et al. (1998); 1997-March 2001 data from Williams-Jones et al. (2003); May 2001 data from Galle et al. (2002); December 2001 data from McGonigle et al. (2002); 2002-2004 data from Williams-Jones (unpublished)

Appendix B – Data

171

B.3 2006 comparison of FLYSPEC and FLYSPEC v2 on 2/26/2006

Traverse FLYSPEC Flux @ 1m/s (tonnes/day)

FLYSPEC v2 Flux @ 1m/s (tonnes/day)

% Difference from FLYSPEC to FLYSPEC v2

1 116 117 0.86 2 377 380 0.80 3 186 200 7.53 4 190 197 3.68 5 54 50 7.41 6 134 124 7.46

Average 176.2 178 1.02

Appendix B – Data

172

B.4 2005 and 2006 spectrometer-based plume speed data as determined from dual spectrometer method (Various time windows from iterative processing are presented. See Appendix A.3 for details. Asterisk indicates spectrometer separation measured manually with tape measure.)

Date Duration (minutes)

Spectrometer Separation

(m)

Plume Speed based on full scan

(m/s) R2

2/25/2005 30 49.9 +/- 1.2 10.2 0.91 2/26/2005 30 50.2 +/- 0.9 10.7 0.97 3/3/2005 30 18.1 +/- 4.9 2.4 0.80 3/5/2005 30 26.1 +/- 2.7 -----no good correlations----- 3/6/2005 30 20.7 +/- 1.5 4.5 0.87 3/7/2005 30 22.1 +/- 3.8 12.3 0.99 3/11/2005 30 20.1 +/- 3.2 6.0 0.68 3/12/2005 30 21* 4.8 0.95 3/15/2005 30 29.7 +/- 1.8 29.7 0.96 3/16/2005 30 33.2 +/- 11.7 9.8 0.97 2/23/2006 20 73.2 +/- 2.7 11.6 0.96 2/26/2006 30 54.5 +/- 2.3 21.8 0.99 2/27/2006 30 37.7 +/- 3.9 19.9 0.91 2/28/2006 20 46.3 +/- 1.3 17.8 0.98 3/1/2006 20 71.2 +/- 2.9 20.3 0.99 3/3/2006 30 39.9 +/- 2.3 8.9 0.99 3/7/2006 20 22.2 +/- 1.1 6.9 0.96 3/12/2006 20 83.3 +/- 1.4 17.7 0.99

Appendix B – Data

173

Date Plume speed based on 10-

minute moving window (m/s)

Average R2of windows with

R2 >0.9

Number of windows with R2 >0.9 (out of total

number of windows)

2/25/2005 10.2 +/- 0.9 0.96 +/- 0.03 3/3 2/26/2005 11.5 +/- 2.6 0.98 +/- 0.03 3/3 3/3/2005 2.7 0.96 1/3 3/5/2005 3.9 0.95 1/3 3/6/2005 1.9 +/- 0.6 0.96 +/- 0.04 2/3 3/7/2005 12.4 +/- 2.0 0.99 +/- 0.01 2/3 3/11/2005 -------------no good correlations------------- 3/12/2005 3.4 +/- 0.2 0.99 +/- 0.01 3/3 3/15/2005 11.0 0.96 1/3 3/16/2005 13.3 +/- 6.5 0.97 +/- 0.01 3/3 2/23/2006 11.7 +/- 0.4 0.97 +/- 0.01 2/2 2/26/2006 21.9 +/- 3.7 0.99 +/- 0.01 3/3 2/27/2006 12.2 0.94 1/3 2/28/2006 17.9 +/- 2.0 0.99 +/- 0.01 2/2 3/1/2006 20.1 +/- 1.2 0.98 +/- 0.01 2/2 3/3/2006 7.5 +/- 1.1 0.99 +/- 0.02 3/3 3/7/2006 6.97 +/- 0.8 0.96 +/- 0.02 2/2 3/12/2006 17.2 +/- 0.8 0.99 +/- 0.01 2/2

Appendix B – Data

174

Date Plume speed based on 5-minute moving window

(m/s)

Average R2of windows with

R2 >0.9


number of windows)

2/25/2005 10.3 +/- 1.9 0.95 +/- 0.01 5/6 2/26/2005 9.0 0.97 1/6 3/3/2005 2.4 +/- 0.1 0.97 +/- 0.02 2/6 3/5/2005 3.4 +/- 1.3 0.96 +/- 0.01 2/6 3/6/2005 2.1 +/- 0.4 0.96 +/- 0.02 2/6 3/7/2005 12.0 +/- 3.1 0.99 +/- 0.03 4/6 3/11/2005 2.1 +/- 1.3 0.96 +/- 0.01 2/6 3/12/2005 3.5 +/- 0.5 0.99 +/- 0.02 6/6 3/15/2005 12.4 +/- 7.2 0.99 +/- 0.01 2/6 3/16/2005 10.5 +/- 1.6 0.98 +/- 0.02 5/6 2/23/2006 12.1 +/- 0.9 0.97 +/- 0.06 4/4 2/26/2006 22.5 +/- 3.2 0.99 +/- 0.02 6/6 2/27/2006 12.4 +/- 3.8 0.95 +/- 0.02 2/6 2/28/2006 19.0 +/- 3.0 0.99 +/- 0.01 4/4 3/1/2006 20.6 +/- 2.7 0.99 +/- 0.02 4/4 3/3/2006 7.3 +/- 1.2 0.99 +/- 0.02 5/6 3/7/2006 7.0 +/- 0.9 0.97 +/- 0.03 2/4 3/12/2006 16.8 +/- 1.8 0.99 +/- 0.02 4/4

Appendix B – Data

175

Date Plume speed based on 3-minute moving window

(m/s)

Average R2of windows with R2

>0.9


number of windows)

2/25/2005 11.6 +/- 1.8 0.95 +/- 0.02 4/10 2/26/2005 13.2 0.95 1/10 3/3/2005 2.4 0.97 1/10 3/5/2005 3.3 +/- 1.5 0.98 +/- 0.02 3/10 3/6/2005 2.1 +/- 0.6 0.95 +/- 0.02 2/10 3/7/2005 11.3 +/- 4.6 0.99 +/- 0.02 6/10 3/11/2005 1.4 +/- 0.2 0.97 +/- 0.02 3/10 3/12/2005 3.6 +/- 0.8 0.99 +/- 0.02 9/10 3/15/2005 9.4 +/- 1.2 0.97 +/- 0.03 2/10 3/16/2005 12.9 +/- 5.6 0.98 +/- 0.01 8/10 2/23/2006 11.4 +/- 2.5 0.96 +/- 0.06 3/6 2/26/2006 23.3 +/- 3.8 0.99 +/- 0.02 10/10 2/27/2006 13.0 0.96 1/10 2/28/2006 18.6 +/- 0.9 0.99 +/- 0.02 6/6 3/1/2006 21.4 +/- 2.8 0.99 +/- 0.01 6/6 3/3/2006 7.7 +/- 1.2 0.99 +/- 0.04 9/10 3/7/2006 7.4 +/- 1.3 0.96 +/- 0.02 3/6 3/12/2006 17.4 +/- 2.1 0.99 +/- 0.03 4/6

Appendix B – Data

176

B.5 2005 and 2006 wind speed data as measured by hand-held anemometer (Reported are the recorded averages of continuous measurements with a 1 second sampling interval)

Date Road Duration (mins)

Average (m/s)

Maximum Gust (m/s)

2/25/05 Ticuantepe 5 2 4 2/25/05 Llano Pacaya 15 3.7 7 2/26/05 Ticuantepe 5 1.3 3.9 2/26/05 Llano Pacaya 15 2.8 5.9 3/5/05 Ticuantepe 5 1.4 4.4 3/6/05 Ticuantepe 5 1.8 3.7 3/6/05 Ticuantepe 5 1.7 4.1 3/7/05 Ticuantepe 5 2.2 5.2 3/11/05 Ticuantepe 5 3.6 8.5 3/12/05 Ticuantepe 5 2.6 7.5 3/12/05 Llano Pacaya 5 5.7 10 3/12/05 Llano Pacaya 5 5.5 7.8 3/14/05 Ticuantepe 5 3 5 3/15/05 Ticuantepe 5 1.1 2.1 3/15/05 Llano Pacaya 5 5 7.8 3/16/05 Ticuantepe 5 1.1 2.9 2/16/06 Llano Pacaya 5 6.5 11.3 2/17/06 Llano Pacaya 5 7.4 11.7 2/17/06 Llano Pacaya 5 8.4 13.2 2/17/06 Ticuantepe 5 3 6.5 2/18/06 Ticuantepe 5 3.2 7.4 2/18/06 Ticuantepe 5 1.2 2.9 2/18/06 Llano Pacaya 5 2.1 8.2 2/18/06 Llano Pacaya 5 8.1 12 2/18/06 Llano Pacaya 5 13.1 19.3 2/19/06 Ticuantepe 5 1.9 4.9 2/19/06 Ticuantepe 5 0.8 1.8 2/19/06 Llano Pacaya 5 9.9 15.7 2/19/06 Llano Pacaya 5 8.7 14.1 2/20/06 Ticuantepe 5 2.5 5 2/20/06 Llano Pacaya 5 5.3 10.8 2/20/06 Llano Pacaya 5 7.3 12.6 2/20/06 Ticuantepe 5 2.4 5.4 2/21/06 Ticuantepe 5 1.1 3.1 2/21/06 Llano Pacaya 5 6.3 11.9 2/21/06 Llano Pacaya 5 4.1 7 2/23/06 Llano Pacaya 15 7.6 12.5 2/24/06 Ticuantepe 5 2.4 5 2/24/06 Llano Pacaya 15 9.4 14.7 2/26/06 Llano Pacaya 5 6.5 9.9 2/27/06 Llano Pacaya 10 5.6 12.4 2/27/06 Llano Pacaya 5 3 6.2 2/28/06 Llano Pacaya 10 9.3 13.2

Appendix B – Data

177

Date Road Duration (mins)

Average (m/s)

Maximum Gust (m/s)

3/1/06 Llano Pacaya 5 7.2 10.2 3/2/06 Llano Pacaya 10 8 12.6 3/3/06 Llano Pacaya 10 4.9 8.3 3/4/06 Llano Pacaya 10 7.9 11.7 3/6/06 Llano Pacaya 10 9.6 14.7 3/7/06 Llano Pacaya 5 6.1 10.9 3/8/06 Llano Pacaya 5 4.5 8.2 3/12/06 Llano Pacaya 5 6.1 10.2

Appendix B – Data

178

B.6 Simultaneous aerosol/SO2 traverse data as measured on Ticuantepe and Llano Pacaya roads (See Appendix A.2 for details; SO2 is plotted in gray, aerosols are in black; x-axis labels are in the form ‘Road, traverse #, Spectrometer, Microtops, date; SFU represents the Microtops from Simon Fraser University and SH the Microtops from the University of Sherbrooke)

0.00E+00

1.00E+11

2.00E+11

3.00E+11

4.00E+11

5.00E+11

6.00E+11

0 1000 2000 3000 4000 5000 6000

Distance along Traverse (Ticuantepe, # 20, FLYSPEC, SFU Mtops, 3/1/06)

Aer

osol

par

ticle

col

umn

dens

ity

(par

ticle

s/cm

2 )

0

100

200

300

400

500

600

SO2 p

ath

leng

th c

once

ntra

tion

(ppm

-m)

0.00E+00

1.00E+11

2.00E+11

3.00E+11

4.00E+11

5.00E+11

6.00E+11

0 1000 2000 3000 4000 5000 6000 7000

Distance along Traverse (Llano Pacaya, # 6, FLYSPECv2, SH Mtops, 3/1/06)

Aer

osol

par

ticle

col

umn

dens

ity

(par

ticle

s/cm

2 )

0

20

40

60

80

100

120

140

160

SO2 p

ath

leng

th c

once

ntra

tion

(ppm

-m)

Appendix B – Data

179

0.00E+00

1.00E+11

2.00E+11

3.00E+11

4.00E+11

5.00E+11

6.00E+11

0 500 1000 1500 2000 2500 3000 3500 4000

Distance along Traverse (Ticuantepe, # 2, FLYSPECv2, SH Mtops, 3/6/06)

Aer

osol

par

ticle

col

umn

dens

ity

(par

ticle

s/cm

2 )

-100

0

100

200

300

400

500

600

SO2 p

ath

leng

th c

once

ntra

tion

(ppm

-m)

0.00E+00

1.00E+11

2.00E+11

3.00E+11

4.00E+11

5.00E+11

6.00E+11

0 1000 2000 3000 4000 5000 6000 7000 8000

Distance along Traverse (Llano Pacaya, # 5, FLYSPEC, SFU Mtops, 3/6/06)

Aer

osol

par

ticle

col

umn

dens

ity

(par

ticle

s/cm

2 )

-20

0

20

40

60

80

100

120

140

160

180

SO2 p

ath

leng

th c

once

ntra

tion

(ppm

-m)

Appendix B – Data

180

0.00E+00

1.00E+11

2.00E+11

3.00E+11

4.00E+11

5.00E+11

6.00E+11

7.00E+11

0 1000 2000 3000 4000 5000

Distance along Traverse (Ticuantepe, # 6, FLYSPECv2, SH Mtops, 3/6/06)

Aer

osol

par

ticle

col

umn

dens

ity

(par

ticle

s/cm

2 )

-100

0

100

200

300

400

500

600

700

SO2 p

ath

leng

th c

once

ntra

tion

(ppm

-m)

0.00E+00

5.00E+10

1.00E+11

1.50E+11

2.00E+11

2.50E+11

3.00E+11

3.50E+11

4.00E+11

4.50E+11

5.00E+11

0 1000 2000 3000 4000 5000 6000 7000 8000

Distance along Traverse (Llano Pacaya, # 9, FLYSPEC, SFU Mtops, 3/6/06)

Aer

osol

par

ticle

col

umn

dens

ity

(par

ticle

s/cm

2 )

-50

0

50

100

150

200

250

300

SO2 p

ath

leng

th c

once

ntra

tion

(ppm

-m)

Appendix B – Data

181

0.00E+00

1.00E+11

2.00E+11

3.00E+11

4.00E+11

5.00E+11

6.00E+11

7.00E+11

0 500 1000 1500 2000 2500 3000 3500Distance along Traverse (m)

(Ticuantepe, # 4, FLYSPECv2, SH Mtops, 3/8/06)

Aer

osol

par

ticle

col

umn

dens

ity

(par

ticle

s/cm

2 )

-100

0

100

200

300

400

500

600

SO2 p

ath

leng

th c

once

ntra

tion

(ppm

-m)

0.00E+00

5.00E+10

1.00E+11

1.50E+11

2.00E+11

2.50E+11

3.00E+11

3.50E+11

4.00E+11

0 1000 2000 3000 4000 5000 6000 7000

Distance along Traverse (m)(Llano Pacaya, # 4, FLYSPEC, SFU Mtops, 3/8/06)

Aer

osol

par

ticle

col

umn

dens

ity

(par

ticle

s/cm

2 )

-150

-100

-50

0

50

100

150

200

250

SO2 p

ath

leng

th c

once

ntra

tion

(ppm

-m)

Appendix B – Data

182

0.00E+00

1.00E+11

2.00E+11

3.00E+11

4.00E+11

5.00E+11

6.00E+11

7.00E+11

0 500 1000 1500 2000 2500 3000 3500 4000

Distance along Traverse (m) (Ticuantepe, # 6, FLYSPECv2, SH Mtops, 3/8/06)

Aer

osol

par

ticle

col

umn

dens

ity

(par

ticle

s/cm

2 )

-100

0

100

200

300

400

500

600

700

SO2 p

ath

leng

th c

once

ntra

tion

(ppm

-m)

0.00E+00

5.00E+10

1.00E+11

1.50E+11

2.00E+11

2.50E+11

3.00E+11

3.50E+11

0 1000 2000 3000 4000 5000 6000

Distance along Traverse (m) (Llano Pacaya, # 8, FLYSPEC, SFU Mtops, 3/8/06)

Aer

osol

par

ticle

col

umn

dens

ity

(par

ticle

s/cm

2 )

-40

-20

0

20

40

60

80

100

120

140

160

180

SO2 p

ath

leng

th c

once

ntra

tion

(ppm

-m)

Appendix B – Data

183

B.7 Profile data of NDVI, S dry deposition, and SO2 ground-level concentration (Data from profiles described in Chapter 4; see Figures 4-12 and 4-13 for locations of

profiles. Elevation is plotted in gray for reference on all profiles)

B.7.1 NDVI

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000

Distance Along Profile 1 (m)

Elev

atio

n (m

a.s

.l.)

-0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

ND

VI

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

-0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

ND

VI

Appendix B – Data

184

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

-0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

ND

VI

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

-0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

ND

VI

Appendix B – Data

185

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

-0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

ND

VI

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

-0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

ND

VI

Appendix B – Data

186

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

-0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

ND

VI

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

-0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

ND

VI

Appendix B – Data

187

100

150

200

250

300

350

400

450

500

550

600

0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000

Distance Along Control Profile (m)

Elev

atio

n (m

a.s

.l.)

-0.5

-0.4

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

ND

VI

Appendix B – Data

188

B.7.2 Dry deposition of S (Dry deposition data shown in black, elevation in gray; dry deposition data derived from Delmelle et al. (2002))

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

0

60

120

180

240

300

360

420

480

540

600

Dry

Dep

ositi

on (m

g/m

2 /d)

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

0

60

120

180

240

300

360

420

480

540

600

Dry

Dep

ositi

on (m

g/m

2 /d)

Appendix B – Data

189

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

0

60

120

180

240

300

360

420

480

540

600

Dry

Dep

ositi

on (m

g/m

2 /d)

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

0

60

120

180

240

300

360

420

480

540

600

Dry

Dep

ositi

on (m

g/m

2 /d)

Appendix B – Data

190

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

0

65

130

195

260

325

390

455

520

585

650

Dry

Dep

ositi

on (m

g/m

2 /d)

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

0

60

120

180

240

300

360

420

480

540

600

Dry

Dep

ositi

on (m

g/m

2 /d)

Appendix B – Data

191

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

0

60

120

180

240

300

360

420

480

540

600

Dry

Dep

ositi

on (m

g/m

2 /d)

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

0

60

120

180

240

300

360

420

480

540

600

Dry

Dep

ositi

on (m

g/m

2 /d)

Appendix B – Data

192

0

100

200

300

400

500

600

0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000


Elev

atio

n (m

a.s

.l.)

0

30

60

90

120

150

180

Dry

Dep

ositi

on (m

g/m

2 /d)

Appendix B – Data

193

B.7.3 SO2 ground-level concentration (SO2 data shown in black, elevation in gray; SO2 data derived from Delmelle et al. (2002))

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

0

20

40

60

80

100

120

140

160

180

200

SO2 C

once

ntra

tion

(ppb

v)

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

0

20

40

60

80

100

120

140

160

180

200

SO2 C

once

ntra

tion

(ppb

v)

Appendix B – Data

194

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

0

20

40

60

80

100

120

140

160

180

200

SO2 C

once

ntra

tion

(ppb

v)

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

0

20

40

60

80

100

120

140

160

180

200

SO2 C

once

ntra

tion

(ppb

v)

Appendix B – Data

195

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

0

20

40

60

80

100

120

140

160

180

200

SO2 C

once

ntra

tion

(ppb

v)

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

0

20

40

60

80

100

120

140

160

180

200

SO2 C

once

ntra

tion

(ppb

v)

Appendix B – Data

196

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

0

20

40

60

80

100

120

140

160

180

200

SO2 C

once

ntra

tion

(ppb

v)

0

100

200

300

400

500

600

700

800

900

1000

0 5000 10000 15000 20000 25000 30000 35000 40000


Elev

atio

n (m

a.s

.l.)

0

20

40

60

80

100

120

140

160

180

200

SO2 C

once

ntra

tion

(ppb

v)

Appendix B – Data

197

0

100

200

300

400

500

600

0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000


Elev

atio

n (m

a.s

.l.)

0

30

60

90

120

150

180

SO2 C

once

ntra

tion

(ppb

v)

Appendix B – Data

198

B.8 References

Delmelle, P., Stix, J., Baxter, P.J., Garcia-Alvarez, J., and Barquero, J., 2002. Atmospheric dispersion, environmental effects and potential health hazard associated with the low-altitude gas plume of Masaya volcano, Nicaragua. Bulletin of Volcanology, 64: 423-434.








Williams-Jones, G., Rymer, H., and Rothery, D.A., 2003. Gravity changes and passive SO2 degassing at the Masaya caldera complex, Nicaragua. Journal of Volcanology and Geothermal Research, 123: 137-160.

a multi-parameter investigation of olcanic plume … · i also owe a thank you to mr. bob meyer...

Documents