o b s e r vings a r t e m e the earth observer

In this issue . . .

Ear

thObserving System

THE EARTH OBSERVER

A Bimonthly EOS Publication November/December 2000 Vol. 12 No. 6

EDITOR’S CORNERMichael KingEOS Senior Project ScientistSCIENCE MEETINGS

Aura Fall Annual Meeting ........................... 3

Clouds and the Earth’s Radiant EnergySystem Science Team Meeting ............. 5

The Availability and Status of MODIS LandData Products ...................................... 10

Formation of the Science Working Group on Data (SWGD) .................................. 18

Terra Science Working Group (SWAMP)Meeting ................................................ 20

2nd Science Team Meeting for the SolarRadiation and Climate Experiment—SORCE ................................................. 22

DAAC Managers’ Quarterly Meeting ........ 28

The Emerging Committee on EarthObservation Satellite (CEOS) WorkingGroup on Information Systems andServices (WGISS) Test Facility (WTF) —Global Observation of Forest Cover

(GOFC) ................................................. 31

HDF-EOS Workshop IV Meeting Summary................................................................. 35

National Snow and Ice Data Center (NSIDC) User Working Group Meeting .............. 40

SCIENCE ARTICLESWest Antarctic Ice Sheet May Be a Smaller

Source of Current Sea-Level Rise ....... 30

Global Diagnosis Completed of OceanRegions Most Sensitive To an Iron-RichDiet ...................................................... 34

NASA, FEMA Partner to Use Science andSpace Technology for Disaster Prevention............................................................. 39

ANNOUNCEMENTSEOS Scientists in News .............................. 2

Science Calendars .................................... 43

Information/Inquiries ................. Back cover

The year 2000 marked a major milestone in the development of the EarthObserving System, with unprecedented new global measurement capabili-ties and science results from the EOS satellite Terra, complemented by those

from Landsat 7, QuikScat, ACRIMSAT and EO-1. As exciting as this year hasbeen, it is merely the beginning of an aggressive and comprehensivespaceborne global monitoring campaign, and will greatly increase thefrequency and significance of new Earth science discoveries.

Many of these new science discoveries were presented at the Fall AmericanGeophysical Union conference held in San Francisco in December. Severalimportant Earth science findings were published in a variety of disciplines,

including hydrology, atmospheric chemistry, oceanography, and glaciology.Many of these findings were presented during two dedicated Terra sessions.A Terra-biosphere-results session was chaired by Terra Project Scientist JonRanson, and an atmospheric results session was chaired by John Gille and

Yoram Kaufman. This diverse suite of new research applications demon-strates the comprehensive global monitoring capabilities of the EOS pro-gram, and is indicative of its certain contributions in the near future.

Also during the AGU meeting, a special NASA Town Hall meeting was heldto discuss NASA’s New Strategic Plan for Earth Science Research, and tosolicit feedback from the diverse audience. Associate Administrator for EarthScience, Ghassem Asrar, and Director of the Research Division, Jack Kaye,

presented the new Research Strategy that will be used to prioritize andstructure solicitations for new measurement capabilities. The ResearchStrategy is a living document and will continue to be refined with the inputof the broader Earth science community.

I’m pleased to announce the availability of a new Terra Data Sampler,produced by the EOS Data and Information System. It contains a variety ofdata from Terra instruments obtained for the Western U.S. wildfires that

occurred in the fall of 2000. All data on the CD-ROM are in HDF-EOSformat, and include comprehensive documentation and viewing tools. It is

(Continued on page 2)

THE EARTH OBSERVER

2

compatible with all computer systemscapable of reading an ISO 9600 CD-

ROM, and its imbedded HTML links willprovide real-time access to external dataand reference materials. This packagewill aid in familiarizing data users with

new data applications, formats, andanalysis tools. To obtain a copy of theTerra Data Sampler CD-ROM, contactthe EOSDIS Science Operations Office by

e-mail at [email protected].

The EOS Project Science Office has alsoproduced the 2001 volume of the EOSGlobal Change Media Directory. Thisvaluable reference provides journalists

and other media professionals with aready source of international expertiseon global change science and policy. The343 scientists included in this directory

represent over 30 scientific disciplines atthe center of NASA’s EOS program, fromozone chemistry and natural hazards toglobal warming and ecosystems. The

scientists listed in the Directory haveexpressed an interest and willingness towork with the media, and many haveoffered to respond to inquiries within 24

hours. The Global Change MediaDirectory is available online from theNewsroom of the Earth ObservatoryWeb site at earthobservatory.nasa.gov/

Newsroom/.

Finally, I want to express my congratula-tions to all of those who have contrib-uted to an extremely successful year forNASA’s Earth Science Enterprise. Now

we are eagerly anticipating the launchesof METEOR 3M/SAGE III, Jason-1,QuikTOMS, Aqua, ESSP/GRACE,ADEOS II, and ICESat. May 2001 be just

as successful as 2000.

“Rain Maps Show Air Pollution

Hinders Rainfall,” (December 19)

CNN.com, CNN cable, United Press

International

Daniel Rosenfeld (Hebrew University ofJerusalem), Robert Adler (NASAGoddard) and Edward Zipser (Univer-sity of Utah) discussed new rain maps

from the TRMM satellite that show howsmoke can reduce rainfall from a stormby as much as half compared to the samekind of storm over the ocean.

“Ozone Layer Over Arctic is Not

Healing Quickly,” (December 18)

CNN.com, CNN cable, Associated Press

NASA Goddard’s Paul Newman andDale Hurst (NOAA) discussed thatbromine may be preventing the Arctic

ozone loss from recovering, even thoughchlorine levels are dropping.

“Satellite Images Show Africa’s River

of Smoke,” (November 30) CNNfn,

AllAfrica Inc.

Robert Swap (University of Virginia) and

Peter Hobbs (University of Washington)were interviewed about their involve-ment in the SAFARI 2000 field campaign,which monitored pollution from biomass

burning over southern Africa this

summer.

“Robots in the Sky,” (November 2000)

Scientific American

Judy Curry (University of Colorado) is

using unmanned aircraft in Alaska togather weather data and fill in data-poorareas in the Arctic. Curry hopes to usethem to enhance global weather forecasts

and monitor the movement of Arctic seaice.

“Satellite Eyes Coral Reefs,” (October

31) Discovery.com

Frank Muller-Karger (University of

South Florida) and Darrel Williams(NASA Goddard) discuss how NASA’sLandsat 7 imagery is providing a quickerand more cost-effective way of monitor-

ing coral reefs around the world.

(Now the Ebola Forecast,” (October 19)

New Scientist

Compton Tucker (NASA/Goddard) and

Jim Wilson (World Health Organization)used satellite data to monitor wet anddry conditions and believe a rare climatepattern precedes outbreaks of the deadly

Ebola disease.


November/December 2000 • Vol. 12 No. 6

3

The Aura Fall Annual Meeting (AFAM)for the Aura project was held Tuesdayand Wednesday, October 17-18, 2000, inEaston Maryland. The AFAM replaced

the Chemistry Annual Project SteeringMeeting (CHAPS) for Fall 2000. After thedelta critical design review (CDR) for thespacecraft on September 11, meetings of

the Mission Operations Working Groupand the Integration and Test WorkingGroup occurred. Thus the Projectdecided not to schedule CHAPS as suchand to have a reduced meeting that

allowed more time for discussions ofAura validation. The focus of the AFAMwas limited to the remaining workinggroups (Algorithms, Education and

Outreach, Data Systems, Aerosols, andValidation).

Mark Schoeberl (Project Scientist)

introduced formation flying to thescience team and discussed the benefitsof acquiring MODIS cloud and aerosolproducts very close in time to an Aura

fly over. In general, each of the instru-ment PIs saw value in acquiring nearlysimultaneous MODIS data with forma-tion flying but also saw some complica-

tions and felt more analysis was needed.A second formation flying groupmeeting was held at Goddard onNovember 28 with the PIs of CloudSat,

PICASSO-CENA, PARASOL, and Aquamaking presentations.

Peg Luce, Aura Project Manager, gave anupdate on project activities and the latestactivities on the spacecraft buildup atTRW. Although Aqua will pave the way

for Aura, Aqua difficulties and scheduledelays could have an impact on Aura’sschedule.

John Liacono (Aura Deputy ProjectManager) gave a report on the instru-ment and spacecraft status. His presenta-tion focussed on delivery of subsystems;there has been rapid tangible progress

during the last year, which is expectedfor a project that is in this part of its lifecycle relative to the expected launch dateof summer 2003.

John Gille gave a status report onHIRDLS development and discussed thealgorithm for dealing with variability in

temperature and constituent mixing ratioalong the line of site.

Joe Waters provided an update on MLS

development and discussed retrievals forCH3CN (methyl cyanide) and posed theproblem on how these data would bevalidated.

Bert Van de Oord gave an update on thestatus of the OMI instrument. Thedemonstration model, with only UV

channels, is undergoing tests at thecontractor facility, which will demon-

strate its spectral imaging capability.OMI Level 1 requirements and the

primary data deliverables were officiallypresented to the instrument teams andthe Project.

Reinhard Beer reviewedTES development and

discussed TES’sability to detect

troposphericozone. Cal-

culations utilize ozone sonde profiles tosimulate TES observations. These

simulated observations are utilized inthe TES retrievals, and results arecompared with the original profiles.Results of this study indicate that ozone

will be detectable at several levels in thetroposphere.

Dolly Perkins, the ESDIS Project Man-

ager, discussed the EOSDIS fundingstatus. EOSDIS was under a lot of stressbecause of budget constraints set byNASA HQ. Aura is expected to getfunding and is not likely to suffer as a

result of Terra or Aqua pressures formore funding.

Aerosol Working Group

Representatives from the HIRDLS, OMI,and TES science teams attended aworking group meeting (Steve Massie,chair). The OMI representative stated

that the AERONET network would be aprimary source of correlative data.Information on characteristics of aerosols(size distributions, composition) in the

upper troposphere is needed by all threeteams. It was also clear that the groupneeds to obtain (or modify) existingcodes so that the influence of multiple

scattering can be examined.

Aura Fall Annual Meeting— Anne Douglass ([email protected]),

NASA Goddard Space Flight Center

THE EARTH OBSERVER

4

Algorithm Working Group

A brief meeting was held on Tuesdayevening (Nathaniel Livesey, chair). Thereare plans to utilize constituent output

from the MOZART model to testalgorithms and compare results amongthe instrument teams. Although only oneday will be considered in detail, it is

important that atmospheric variability berealistic. The integration can be com-pleted using input winds from the Terrasystem of the NASA GSFC Data Assimi-

lation Office.

Education and Public Outreach

Aura’s outreach program includes the

American Chemical Society which willreach 30,000 high school chemistryteachers on atmospheric chemistry andmeasurements from space. The

Smithsonian’s National Museum ofNatural History will develop a displayon atmospheric chemistry in conjunctionwith its Global Links program. DrexelUniversity is developing a low cost

instrument and protocol to measure UVAand UVB, along with aerosols, for theGLOBE program (public schools, K-12),which has international participation.

Aura’s new Website was presented to theattendees for their comments. Scheduledlaunch of the Website is January 2001.

The main subject of the Fall AnnualMeeting concerned validation of theAura data products. Prior to thismeeting, Lucien Froidevaux (Chair,

Validation Working Group) had as-sembled a comprehensive version of avalidation document. At this meeting, asubgroup attempted to set priorities for

validation. Criteria used for the strato-sphere were data continuity andimportance of the constituent measure-ment to understanding changes in the

ozone layer. For the troposphere, the

criteria were key species that arefundamental to ozone production.

Discussion on Wednesday morning

concerned lessons learned from thecorrelative measurement campaign forthe Upper Atmosphere Research Satellite(UARS). Mike Kurylo (NASA HQ)

commented that, in the case of UARS,correlative measurements were made inthe first years after launch. Later, scienceapplications contributed to the overall

validation effort. Phil DeCola (NASAHQ Aura Program Scientist) and VickiConnors (NASA HQ Acting Head of theTropospheric Chemistry Program) also

participated in the meeting discussion.

Lynn Sparling (NASA GSFC) showed ananalysis of aircraft observations of ozone

to highlight the differences betweengeophysical variability and instrumenterrors. Such analyses are of importanceto the Aura platform because of theemphasis on the lower stratosphere and

upper troposphere, where synopticvariability plays a significant role. It isnot obvious how to compare in situ

observations made along an aircraft

flight track with remote measurementswhen there is significant geophysicalvariability along the flight path thatdefines the footprint for a single remote

observation.

The remainder of the meeting centeredon reports from the Snowmass working

groups. These groups were formed todefine aircraft missions that could fulfillthe following roles: (a) test realisticscientific hypotheses; (b) be part of the

satellite validation; and (c) utilize bothsatellite and in situ measurements toaddress science goals and questionsrelated to hypotheses in (a).

Darryn Waugh (Johns Hopkins) pre-sented arguments for investigations of

the tropical upper troposphere/lowerstratosphere. The physical and chemicalprocesses in this region have a stronginfluence on both ozone and water vapor

and represent key couplings betweenatmospheric composition and climate.Ross Salawitch (JPL) argued for mea-surements of a variety of constituents at

the tropical tropopause. Jim Holton(University of Washington) argued fortropical measurements to clarify themechanisms for entry of dry air into the

lower tropical stratosphere.

Daniel Jacob (Harvard) presented abroad agenda for aircraft observations in

the troposphere. The missions said tohave the most support from the aircraftcommunity are the IntercontinentalChemical Transport Experiment (INTEX)

and LARS TRACE B. Both of these areconsidered excellent missions; LARSTRACE B takes advantage of the LBAProject in the Amazon basin and musttake place while the ground-based

observations that are part of the LargeBiosphere-Atmosphere Experiment inAmazonia (LBA) Project are in progress(through 2004).

Mark Schoeberl charged Steve Massie(Aerosol Working Group) with the taskof developing a validation plan for

aerosols. Each of the four instruments onAura has some products that are relatedto aerosols.

Writing assignments were made withdeliveries due in November. Most ofthese have been received. The ValidationWorking group will now be finishing up

version 1.0 of the document. For furtherinformation on the Aura project,instruments, and working groups, visitthe EOS Aura Project Home Page,

(http://eos-aura.gsfc.nasa.gov).


5

The 22nd Clouds and the Earth’s RadiantEnergy System (CERES) Science Teammeeting was hosted by Ron Welch at theBevill Conference Center, University of

Alabama in Huntsville (UAH), onSeptember 20-22, 2000. The Science Teamapproved release of validated TerraCERES Level 1 radiance data products as

well as the global Earth RadiationBudget Experiment (ERBE)-like top-of-atmosphere (TOA) fluxes. Early “Beta”products were already available at theLangley Atmospheric Sciences Data

Center (ASDC), and the new CERES datawill be the first validated Level 1-3 TerraData Products available to the sciencecommunity for research and publication.

Data Quality Summaries will be com-pleted this Fall and the Edition 1 dataproducts will appear at the ASDC byDecember. The merged Tropical Rainfall

Measuring Mission (TRMM) cloudproperty/radiation data product (SingleScanner Footprint, SSF) was alsoapproved. The next Science Team

meeting is scheduled for January 23-25,2001 at the NASA Langley ResearchCenter (LaRC).

The next meeting will focus on science

team results, geostationary diurnalcorrections, early Angular DistributionModel (ADM) results from SSF dataproducts, Surface and Atmospheric

Radiation Budget (SARB), Chesapeake

Lighthouse andAircraft Mea-surements forSatellites (CLAMS)

plans, instrumentdeep space deter-mined offsets, andimproved spectral corrections.

Bruce Wielicki (LaRC), CERES Co-Principal Investigator, opened themeeting with an Earth Observing System(EOS) program status report. The Aqua

launch date has slipped to July 2001.Wielicki is continuing to work with thescience community, including modelers,to evolve the EOS science strategy. Bruce

Barkstrom (LaRC), CERES Co-PrincipalInvestigator, discussed the scienceimpact of the present bounds on EarthObserving System Data and Information

System (EOSDIS) hardware capacities.

CERES Instrument Status

Kory Priestley (LaRC) presented the

instrument status report. The Aquainstruments were delivered on scheduleand have completed mechanical andelectrical integration on the spacecraft.

While no instrument problems havebeen encountered, there are outstandingspacecraft interface issues, includingsuspect traceability of spacecraft

supplied bolts to mount the instruments

as well as spacecraft supplied deploy-ment drives that continue to leakpenzane oil. Priestley gave a ‘postmortem’ report on the CERES instru-

ment on TRMM. The likely culprit forthe demise of this instrument is thefailure of an interpoint voltage converter.The Terra instruments have continued to

function flawlessly since their launch inDecember 1999. These two instruments

are currently in an opposing three-month operational cycle of rotating

azimuthal plane and fixed azi-muthal plane scanning. We expectthat in early 2001 the Terra spacecraftwill perform a Deep Space Calibra-

tion maneuver such that finalmeasurements of the scan angledependent offsets can be made.

Terra Radiometric Validation

Kory Priestley (LaRC) led the sessionthat reported results of 11 studies whichquantify the radiometric performance of

the CERES/Terra Flight Model 1 (FM1)and Flight Model 2 (FM2) instruments aswell as their relative relationship withthe CERES/TRMM and ERBE scanner

instruments. These studies utilize dataproducts ranging from instantaneousraw counts to temporally and spatiallyaveraged TOA fluxes and are statistically

robust at the sub 1% confidence level.The ground-to-flight traceability of theradiometric calibration is better than0.3% for the total (TOT) and shortwave

(SW) channels. The window channelcalibrations both shifted by about 0.5%,but in opposite directions. Pointingaccuracy has been established via

coastline detection techniques to bebetter than 1.8 and 2.4 km at nadir forthe FM1 and FM2 instruments, respec-tively. Tropical Deep Convective Cloud

(DCC) investigations were used toestablish both a three-channelintercomparison and SW albedo metric.

Clouds and the Earth’s RadiantEnergy System (CERES) ScienceTeam Meeting— Shashi K. Gupta ([email protected]), and Gary G. Gibson ([email protected]), NASA Langley Research Center

THE EARTH OBSERVER

6

The three-channel intercomparisonsuggests an inconsistency between theFM1 SW channel and SW portion of theTOT channel at the 1.1% level while the

FM2 instrument had no significantinconsistencies in the SW region. TheDCC albedo metric demonstrates thatthe SW channels on all three CERES

instruments agree to within 0.5%,suggesting the inconsistency in FM1resides in the SW portion of the TOTchannel. Results of a direct comparison

of the FM1 and FM2 instruments onTerra conducted by Richard Green(LaRC) were in general agreement withthe other studies. A direct comparison

between the CERES instruments onTRMM and Terra performed by MartialHaeffelin (Virginia Tech) showed thatTerra radiances are within 0.5% of the

CERES/TRMM values.

Terra ERBE-Like TOA Fluxes

Norman Loeb (Hampton University[HU]) conducted theoretical tests thatshowed that unfiltering algorithms aremore robust for Terra SW compared to

TRMM. Unfiltered SW and long wave(LW) radiance differences between FM1and FM2 are small (less than 0.5%) andshow no evidence for any systematic

dependence on scene type or solar angle.

David Young (LaRC) updated the statusof ERBE-like product validation forTRMM and Terra. He compared tropical

means from 1998 CERES data with EarthRadiation Budget Satellite (ERBS)scanner five-year (1985-89) averages.Comparison of Terra results with the

ERBE climatology was consistent with asimilar comparison using TRMM data. Acomparison of results from the crosstrackand rotating azimuth plane scan modes

showed zonal mean differences less than2 Wm-2 except at the Poles.

Data Production

Erika Geier (LaRC) updated the team onCERES data product status, major SSFchanges, subsetting data products(Langley ASDC can subset SSF, CRS, and

ERBE-like instantaneous products), andCERES data distributed from theLangley ASDC. There have been someproblems in receiving Terra data. Most

data sources are available for March-May2000, but June-August data are onlyavailable sporadically. Terra countconversion offsets previously set to zero

have been updated to the ground-determined values. Reprocessing will becompleted this fall.

TRMM SSF Edition 1 Status andEarly MODIS/TRMM comparisons

Patrick Minnis (LaRC) summarized

recent improvements to the cloudalgorithm. The 1.6-µm calibration wasrevised, sub-pixel cloud contaminationwas minimized with a spatial variabilitytest for aerosol optical depth (AOD)

retrievals, a 10’ land/water mask wasincorporated, three new parameters wereadded to SSF (fraction of pixels, and themean 0.6- and 1.6-µm reflectances used

for AOD retrievals), and three different3.7-µm reflectance models were com-pared to help explain differences inparticle size. Preliminary cloud retrievals

from the Moderate-Resolution ImagingSpectroradiometer (MODIS) werecompleted and compared to VisibleInfrared Scanner (VIRS) results.

Angular Modeling

Norman Loeb discussed ADM-relatedactivities. LW ADMs were developed forclear and overcast scenes stratified by

cloud properties. Progress was made onvalidating cloud property averagingover CERES footprints. A new methodwas developed for reducing angle-

dependent biases in satellite cloudoptical depth retrievals. The approachwill be used to develop CERES ADMoptical depth classes, but will require

more than the eight months of VIRSdata.

CERES Validation Experiments

Ken Rutledge (Analytical Services &Materials, Inc. [AS&M]) briefed the teamon the status of the CERES OceanValidation Experiment (COVE) site.

Chesapeake Lighthouse is an oceanplatform located 25 km from VirginiaBeach, VA. COVE measures insolation,albedo, net LW flux, and AOD. The basic

radiometric instrumentation has beeninstalled using the recommendations ofthe Baseline Surface Radiation Network(BSRN).

Tom Charlock discussed plans andobjectives for the summer 2001 CLAMSmission. The CLAMS validation missionincludes clear-sky measurements of

aerosols, SW broadband fluxes, andspectral radiances at the sea surface andwithin the atmosphere. CLAMS willinclude COVE measurements of radia-

tion, meteorology, and ocean waves, alow-level aircraft to measure radiation atthe sea surface, a mid-level aircraft toprofile aerosol properties in situ, and a

high-level aircraft to measure radiationand sense aerosols remotely with lidar.

Advances in TemporalInterpolation

David Young reviewed the status of datasets and algorithms for incorporatingthree-hourly visible and infrared datafrom four geostationary satellites into the

CERES temporal interpolation process.Preliminary results show a significantimprovement in daily and monthlymean fluxes. Using geostationary data to


7

interpolate cloud properties furtherreduced the instantaneous SW fluxinterpolation errors.

CERES/ERBE Decadal Variability

Bruce Wielicki showed an update on thedecadal variability of tropical broadbandLW radiation budget using both observa-

tions and general circulation modelresults. Observational analysis usingdata from five independent broadbandradiation instruments between 1985 and

2000 shows good agreement amongmeasurements and significant variabilityin the tropical broadband radiationbudget. This decadal variability in the

tropical broadband radiation budget,however, is not captured by the currentgeneral circulation models (GCMs).These models tend to under-predict the

tropical mean interannual variability inboth radiation and precipitation.

Cloud Working Group

Patrick Minnis led discussions of cloud

retrieval, archival, and validation issues.Minnis, Jim Coakley (Oregon StateUniversity [OSU]), and David Young willcontinue conducting comparisons of

particle size and optical depth over theocean to isolate the differences inretrieval techniques. Other action itemsincluded fixing a problem in the optical

depth interpolation routine and addingthe mean viewing zenith angle of imagerAOD retrieval pixels. SSF Edition 1 hascloud properties, aerosols, and clear-sky

fluxes, but cloudy sky fluxes will not beavailable until new ADMs are generated.

Larry Stowe (NOAA/National Environ-

mental Satellite, Data, and InformationService [NESDIS]) discussed the effect ofrecent changes in VIRS calibration onretrieved AOD. The 1.6 µm channel

calibration is still unsettled. Alexander

Ignatov (NESDIS) discussed the depen-dence of AOD retrievals on scattering,solar zenith, and viewing zenith angles.Xuepeng Zhao (NESDIS) reported on

efforts to validate VIRS-derived AODusing surface data collected at AerosolRobotic Network (AERONET) sites.

Bryan Baum (LaRC) reported that theMODIS channels from 1.23 to 4.55 µm

were having serious problems due toelectronic cross talk. He comparedMODIS Airborne Simulator-derived icecrystal habits and size distributions with

model calculations. Xiquan Dong(University of Utah) compared VIRS-and MODIS-derived cloud propertieswith surface-derived cloud properties

from ARM and other surface sites. RonWelch (UAH) demonstrated a newversion of their MODIS cloud classifierand presented results from an improved

polar cloud classifier.

SARB Working Group

The meeting was jointly chaired by

Thomas Charlock and David Kratz(LaRC). Action items included: continueCLAMS field experiment planning,determine the status of meteorological

products from the Data AssimilationOffice and the European Center forMedium-range Weather Forecasts(ECMWF), and conduct early CRS

(Clouds and Radiative Swath) dataproduct testing with new SSF dataproducts.

David Rutan (AS&M) presented an

update on the CERES ARM ValidationExperiment (CAVE) database that wasdeveloped to facilitate the validation ofCERES-derived surface fluxes. CAVE

consists of flux measurements fromARM, BSRN, SURFRAD, and othersurface sites from around the world thatare matched in space and time with

CERES retrievals.

Ellsworth Dutton (NOAA/ClimateMonitoring & Diagnostics Laboratory[CMDL]) presented the status of surfaceradiometric observations at CMDL and

the 23 BSRN sites. Spectral AODmeasurements are now being made atseveral CMDL sites. Martial Haeffelinpresented a procedure to eliminate large

errors in measured irradiance caused bytemperature differences between thebody and the dome of pyranometers.Bernardo Carnicero-Dominguez (Vir-

ginia Tech) showed that there is a linearrelationship between the net LW fluxmeasured by a pyrgeometer and thethermal offset of a pyranometer.

David Kratz presented validation resultsfor CERES/TRMM surface-only SW andLW fluxes. CERES fluxes were compared

with measurements from ARM, BSRN,and CMDL ground sites. CERES SWresults showed excellent agreement withARM data, but large positive biasesrelative to CMDL and BSRN measure-

ments.

Anand Inamdar (Scripps Institution ofOceanography, SIO) found good

agreement between model outgoing LWresults and CERES monthly averagevalues. He also showed the seasonalvariation of ocean surface temperature

and the normalized atmosphericgreenhouse parameter.

Fred Rose (AS&M) presented a study of

the effects of using sea surface tempera-ture (SST) and column precipitable watervapor (PW) retrieved from TRMMMicrowave Imager (TMI) measurements

on SARB results. Use of TMI PW inSARB computations provided betteragreement with CERES.

Qingyuan Han (UAH) suggested twodifferent approaches for monitoringindirect aerosol radiative forcing from

THE EARTH OBSERVER

8

satellites. The snapshot method usescloud properties from a region andcorrelates them with aerosol propertiesfrom adjacent regions. His one-month

regression method correlates cloud andaerosol properties acquired from thesame region.

TOA Fluxes/ADM Working Group

Norman Loeb led the working groupmeeting with a general overview ofcritical ADM/Inversion research issues.

The group concluded that we shouldavoid screening out “no cloud propertyretrieval cases” or flux biases will occur.In addition, optical depth frequency

distribution normalization should helpminimize flux retrieval biases.

Nitchie Manalo-Smith (AS&M) pre-

sented LW and window limb darkeningfunctions obtained from CERES mea-surements under overcast conditions.She examined how stratifications of themeasurements based on scene param-

eters, such as precipitable water, cloud-surface temperature difference, andcloud emissivity, influence the anisot-ropy of the scene.

Richard Green compared SWreflectances and albedos for ERBS andCERES/TRMM. His results demonstrate

how the difference in footprint sizebetween these two instruments causeslarge differences in scene type frequency-of-occurrence. If the same ADMs are

used to estimate albedo from bothinstruments, the differences in scene typefrequency-of-occurrence lead to largebiases in the all-sky mean albedo despite

the fact that the mean reflectivities areconsistent. Separate ADMs are neededfor ERBS and CERES/TRMM to accountfor the differences in footprint size.

Lin Chambers (LaRC) demonstrated the

diurnal variability in cloud optical depthretrievals from three months of ground-based radar over the SGP and TropicalWestern Pacific ARM sites. She found a

much smaller diurnal trend than issuggested by satellite retrievals (e.g.,VIRS). Her theoretical simulationsshowed that 1D cloud optical depth

retrievals showed considerable bias forsolar zenith angles beyond 45°.

Invited Presentations

Pete Robertson (Marshall Space FlightCenter [MSFC]) presented a comparisonof decadal variability of tropical-meanoutgoing LW in the ERBE/CERES data

set, and in the TIROS (Television andInfrared Observation Satellite) Opera-tional Vertical Sounder (TOVS) Path-Adata set derived from the High Resolu-

tion Infrared Radiation Sounder (HIRS-2)/Microwave Sounding Unit (MSU)measurements. ERBE/CERES datashowed a much stronger variability thanPath-A, though the differences for

corresponding clear-sky fluxes weresmaller.

Roy Spencer (MSFC) made a presenta-

tion on the retrieval of cloud liquid waterpath (LWP) and rain rates from passivemicrowave satellite measurements.Areas of high LWP are often indistin-

guishable from those of light rain. Theadvantages of microwave techniques arethat ice clouds are transparent tomicrowaves and the absorption is

insensitive to drop size distribution.Notable disadvantages are the largefootprint of microwave instruments, andthat the absorption is temperature-

dependent.

John Christy (UAH) presented a 20-yeartime-series of atmospheric temperatures

derived from the MSU. This recordagrees well with balloon measurements

from 30 highly reliable stations in theU.S. Christy also presented the workdone for the IPCC 2000 where long time-series of atmospheric variables (e.g.,

temperature, precipitation) wereexamined to detect long-term trends. Hefound no evidence of global warmingand related effects in the data sets.

Investigator PresentationHighlights

Baijun Tian (SIO) presented a studyexamining the role of tropical cloud

radiative forcing (CRF) in maintainingHadley and Walker circulations. WhileGCM studies show that Hadley andWalker circulations are maintained by

tropical CRF, the observational basis ofthis hypothesis is still missing. From anobservational perspective under clearskies, radiative and latent heat processes

do not provide the necessary energygradients. He showed that LW CRF ofthe atmosphere establishes the necessaryenergy gradients for maintaining thesecirculations.

Leo Donner (NOAA, Geophysical FluidDynamics Laboratory) presented resultsof a GCM study examining the effects of

the treatment of convective shields andcumulus-scale vertical velocities ongeneral circulation. When convectiveshields are included and vertical

velocities are allowed to vary in cumulusparameterizations, the convectivesystems are much larger and modelresults match better with observations.

David Randall (Colorado State Univer-sity) presented a comparative study ofconvective and stratiform precipitation

with particular emphasis on the Tropics.He presented TRMM data that show thatprecipitation types are not limited toconvective and stratiform. He stressed

the need to develop and use similar


9

parameterizations for convective andstratiform precipitation.

Shi-Keng Yang (NOAA/National

Centers for Environmental Prediction[NCEP]) analyzed the downward trendin LW CRF in the new NCEP/NCAR 50-year reanalysis data set and concluded

that the trend is caused by changes inmid-tropospheric relative humidity.

Helene Chepfer (Laboratoire de

Meteorologie Dynamique [LMD],France) presented results of a study ofice-particle shapes in cirrus clouds whenthe same cloud is observed from two

different directions. Her results can beused to validate the phase functions forCERES and MODIS processing and toimprove estimates of ice water content.

Michel Viollier (LMD) presented resultsof a new temporal interpolation proce-dure for SW fluxes. He applied monthly,regional climatological diurnal albedo

curves derived from five years of ERBSdata to fill all local hours for March 2000CERES/Terra data. Comparisons withERBE-like data showed regional differ-

ences up to 10 Wm-2, but no significantdifferences for tropical and global means.

Nicolas Clerbaux (Royal Meteorological

Institute of Belgium) compared theGeostationary Earth Radiation Budget(GERB) Spinning Enhanced Visible andInfrared Imager (SEVIRI) scene identifi-

cation algorithm with the correspondingCERES algorithm. GERB processing willuse ADMs developed by CERES.

Bryan Baum (LaRC) compared MODIScloud fraction, optical depth, height, andphase retrievals with data from othersatellite instruments/algorithms and

field experiments.

Larry Stowe and Alexander Ignatov

(NESDIS) presented results of SSFEdition 1 data testing. Stowe showedthat Edition 1 data were much improvedover earlier SSF versions. Ignatov

outlined the steps underway for devel-oping a third generation aerosol retrievalalgorithm.

Thomas P. Charlock (LaRC) discussedthe aerosol effect in the residual clear-skyinsolation discrepancy. This discrepancysurfaced a few years ago when several

well-regarded theoretical simulationsproduced values for clear-sky SWinsolation that exceeded measurementsby 20 to 30 Wm-2. By both carefully

screening the radiometer observationsand adjusting them with sound physicalprinciples, Charlock found that thetheory exceeded observations by only 5

to 10 Wm-2. Also, for moderate values ofAOD, the aerosol forcing to surfaceinsolation is considerably greater thanthe (now reduced) discrepancy of theoryand observations.

Seiji Kato (HU) presented a radiativetransfer algorithm that can treat horizon-tal inhomogeneities of cloud optical

thickness in a CERES footprint. Thealgorithm assumes a gamma distributionfor optical thickness and uses thediscrete ordinate method to compute the

footprint-averaged irradiance.

James Coakley (OSU) presented esti-mates of direct aerosol radiative forcing

over cloud-free ocean regions derivedusing CERES/TRMM SW reflectanceswith models of AOD. Analyses of IndianOcean Experiment (INDOEX) data

suggested that model AOD valuesprovided reasonable results.

Robert Cess (State University of New

York at Stony Brook) presented algo-rithm development strategies forretrieving downward LW flux (DLF) at

the surface. Model results showed thateven for clear skies, the relationshipbetween DLF and outgoing LW was notgood. Cess suggested that averaging

cloud-base height over a grid box isinappropriate, and that cloud LWP couldbe used in the regression as a surrogatefor cloud-base height.

Lou Smith (Virginia Tech) presented themethodology and results for limb-darkening functions derived using

CERES/TRMM along-track radiancesfrom the window channel.

Bing Lin (LaRC) presented a comparison

study of cloud LWP derived from in situ

and microwave radiometer data takenduring the Surface Heat Budget of theArctic (SHEBA) experiment. Mean cloud

LWP derived using the standard ARMretrieval technique is nearly twice aslarge as coincident in situ aircraft data.Lin’s new microwave radiative transfermodel accounts for atmospheric gas and

cloud water absorption. This methodresults in a 25 to 45% reduction in LWPvalues relative to the ARM estimates. Ifpossible precipitation cases are excluded,

the difference is reduced to only 3%.

Educational Outreach

Lin Chambers (LaRC) reported that over

600 schools from all 50 states and over 44nations are now participating in theStudents’ Cloud Observations On-Line(S’COOL) program.

THE EARTH OBSERVER

10

The Availability and Status of MODISLand Data Products— Chris Justice ([email protected]), University of Virginia— Robert Wolfe ([email protected]), Raytheon— Nazmi El-Saleous (SSAI)— Jacques Descloitres (SSAI)— Eric Vermote, David Roy ([email protected]), University of Maryland— John Owens ([email protected]), University of Maryland— Ed Masuoka ([email protected]), NASA Goddard Space Flight Center,— MODIS Land Science Team (MODLAND)

Over the past few months there havebeen a number of enquiries as to the

status and availability of MODIS landdata products. This article provides asummary of the status as of the time ofpublishing.

Background on MODIS Productionand Option A+

The Terra spacecraft was launched onDecember 18, 1999. After a period of

instrument checkout the first imagesfrom MODIS were obtained on February24, 2000. Since that time MODIS has beenacquiring data with short breaks for

orbital maneuvers, and a period fromAugust 5-18 when the instrument wentinto safe-hold and no Earth-view datawere collected, due to a formatter reset

problem.

In 1998, a NASA management decisionwas made based on the 1997 biennial

review to hold the MODIS Level 2/3data production to 50% of full capacityduring the first year of MODIS produc-tion, followed by a ramp up in subse-

quent years to 0.75 and 1.0X. In 1998, asNASA Headquarters approved the so-called Option A+ budget for EOSDIS, thehardware budget allocations were

determined using the 1996 baseline ofestimated processing loads and datavolumes and applying the ramp-up rules(0.5, 0.75, 1.0).” Budget constraints

dictated this limitation in capacity. Theplan was to increase capacity in thesecond year after launch to 0.75Xproduction capacity with an additional

0.75X capacity for data reprocessing. Inaddition, the distribution capacity wascapped at 1X of the same baseline.Concerns of the instrument team in

terms of the serious impact of thisdecision on production were raised atthat time, but were not addressed. Thebaseline estimates were developed prior

to most of the algorithms being run intest mode providing a best-guess oflikely volumes and loads. As expected,since that time the needed volumes and

loads have increased. A summary andexplanation of this growth is provided inthe recent SWAMP Working Group onData Report (ftp://typhoon.larc.nasa.

gov/pub/brb/SWGD_Final_Final.doc).

The 1X (L1) and 0.5X (L2/3) productioncapacity has been a significant constrain-

ing factor in our ability to establishsteady data production during this firstyear since launch.

In 1998, a decision was made to movethe production of the MODIS L2/3 out ofthe ECS system, and generate theseproducts on a PI Processing system

called MODAPS (MOdis DAta Process-ing System). MODAPS operations areoverseen by the MODIS Science Teamand are managed and run by the MODIS

Science Data Support Team (SDST). Thisapproach was seen as a cost-effectivemeans to generate the MODIS higher-level data sets, and this, indeed, has been

the case. However MODAPS was scopedonly to meet the Option A+ levels ofproduction because of the budgetconstraints previously described.

Recognizing the need to generate full-resolution data sets for assessinginstrument performance, quality

assessment, validation, land science, andapplications, the MODIS land group(MODLAND) of the MODIS ScienceTeam approached the 0.5X productioncap with a plan to develop a mix of

sample global data and continuousregional data sets. High priority regionswere selected based on regional empha-sis in the NASA Research and Analysis

Program and Applications Programs. Ata meeting of the EOS InterdisciplinaryWorking Group in June 1999, the sciencecommunity requested that MODIS land

products be made available for the entireglobe at 500 m and 1 km. Recognizingthe budget constraints associated withthe 1996 baseline and in response to the

IWG request, the MODLAND groupdeveloped a work-around solution,supported by Martha Maiden (ESIPSManager) at NASA HQ, for a rapid

development of a PI-based processingand distribution system for the highvolume MODIS 250-m products. Thisapproach was developed outside of the

EOS Core System to maintain theMODIS allocated volumes and loads,whilst enabling global production of 500-


11

and 1-km land products during the firstyear. The land 250-m production systemhas evolved as part of the MODIS PIprocessing system (MODAPS) and is

currently scoped to generate anddistribute 10% of the global 250-m dailyproducts (surface reflectance andvegetation indices). As a result, the 250-m

Vegetative Cover Conversion product(MOD44A) developed by JohnTownshend, University of Maryland(UMD), is currently being generated at

the UMD Earth Science InformationPartnership (ESIP) – the Global LandCover Facility. The current systemcapacity of 10% was a function of funds

available for hardware purchase.

There are several links in the MODISland data production chain, that are

supported by the ESDIS project. TheLevel 0 data are sent from EDOS to theGSFC Distributed Active Archive Center(DAAC). The GSFC DAAC produces andarchives the Level 1A (MOD01), Level 1B

(MOD02), Geolocation (MOD03), and

Cloud Mask (MOD35) products. Thesedata are sent to MODAPS where the landproducts are generated and sent to theEROS Data Center DAAC. The

MODAPS-generated snow and iceproducts are sent to the National Snowand Ice Data Center DAAC. These twoDAACs provide archiving and distribu-

tion for MODIS land products. The stepsin the MODIS land production areshown in Figure 1.

Experience has shown that, with 1X and0.5X capacity and several links in theland production chain, if one link goesdown for an extended period, then it is

very difficult for the production systemto maintain current production and catchup on missed data. As a result, duringthe first six months of production there

were large gaps in the data productionincluding a series of problems associatedwith EDOS tape production and “bitflips.” During the high-speed playbackof the Terra on-board solid-state re-

corder, electronic cross-talk causesrandom and sporadic changes toindividual bits in the real-time instru-

ment data being downloaded to the

White Sands Receiving Station. Itis worth noting that the MODIS

land L2/3 processing andarchiving for year two is

scoped at 0.75 of the ’96baseline. The current global500-m and 1-km productsalone add up to 40% more

than this allocation. At thislevel, the system will beunable to generate andarchive all the peer-

reviewed L2/3 productscurrently provided by the

Science Team.

In addition to the ESDIS

supported activities, NOAA hasa feed of the Level 0 data from the

GSFC DAAC and is preparing togenerate selected products, includingsnow and ice products to supportoperational weather forecasting. There

are also a number of Direct Broadcaststations that are receiving MODIS dataand producing some of the lower-levelproducts.

Data Release

The MODIS Level 1A and Level 1B dataat 1-km, 500-m and 250-m resolutions

and the MODIS Geolocation productwere made available through the GSFCDAAC starting on April 19, 2000. For thefirst six months there were a large

number of gaps in the global productsgenerated by GSFC and MODAPS, dueto incomplete processing from EDOSand inadequate catch-up capacity. We

plan to fill these gaps at the first majordata reprocessing.

The land products are being released asoriginally planned before launch, in aphased approach making them availableas soon as possible following initial

evaluation changes in instrumentcalibration, initial product qualityassessment, and the integration ofnecessary patches to the code, and prior

to validation, i.e., as a Beta release. As aresult, several of the products haveknown problems that are described inthe product specific web sites and the

user guides, which have been developedfor each product. Due to several productinterdependencies, it is important tounderstand the quality of upstream

products that are used in productgeneration.

The first land products were released by

the EDC DAAC on August 4, 2000, forthe period starting June 9. These werethe 8-day 500-m Surface Reflectance, the16-day 500-m and 1-km Vegetation

Indices, and 8-day 1-km Leaf Area Index.Figure 1. MODIS Land Data Flow

Level 0Instrument Data

UserCommunity

Land Snow andIce Cover Products

NSIDC DAAC EDC DAAC

LandNon-cryospheric

Products

MODAPS 250 mProduction

Level 1, Oceanand Atmosphere

ProductsGSFC DAAC

MODAPS

Ocean andAtmosphere

Products

UserCommunity

UserCommunity

THE EARTH OBSERVER

12

Table 1.

* Freguency is 5 min. for Level 2 and daily for Level 3 unless otherwise noted.EDC=EROS Data Center, NSIDC=National Snow and Ice Data Center.

The 1-km Land Surface Temperatureproducts (L2, daily L3, and 8-day L3) areavailable for the period starting June 25.The 16-day 1-km BRDF/Albedo product

is available for the period starting July11. The 8-day 1-km Fire product isavailable for the period starting August20. The 500-m Snow products (L2, daily

L3, and 8-day L3) are available for theperiod starting September 13. Problemsencountered with obtaining a steadydata flow have delayed preliminary

assessment of the monthly products andthose relying on time-series data, e.g.,land-cover, and vegetative-coverconversion. Planned release dates are

summarized in Table 1.

MODIS 250 m Production (JacquesDescloitres, John Owens)

The first 250-m 8-day Surface Reflectanceproduct was made available on August 4for the acquisition period starting June 9.

The complete North America data set(from 20°N to 60°N) was released shortlythereafter, followed by the next 8-dayperiod starting June 17. In the first phase

of this rapid prototyping, productionwas limited to generation of the SurfaceReflectance product for North America.Difficulties in stabilizing the input data

flow led to a suspension in productionuntil the 8-day period starting Septem-ber 13. Data for this compositing periodare available for North America and

Southern Africa for SAFARI 2000. Inearly November, the production of 250-

m daily L2G Surface Reflectance wastransitioned to the MODAPS productionsystem with the 8-day period startingSeptember 21. A steady flow of 10% of

the global production was reachedstarting October 15. Generating thecorresponding MODIS 250-m 16-dayVegetation Indices product is a work in

progress. A 250-m NDVI product is,however, being generated and distrib-uted by the UMD ESIP. Current under-standing is that in the second year of

production and as part of the MODISreprocessing plan, the 250-m productswill be reprocessed and shipped to EDCfor archive and distribution. The speed

at which this is done will depend on theavailable reprocessing and archivingresources.

Product ID Product Name Level Release Start Date of Archive & Dist.(Frequency *) Date data Series Center

MOD09 Surface Reflectance 2G 11/10/00 6/9/00 EDC3(8d) 8/4/00 6/9/00 EDC

MOD10 Snow Cover 2, 3 (1d, 8d) 10/13/00 9/13/00 NSIDC

MOD11 Land Surface Temperature & Emissivity 2, 3 (1d) 9/1/00 6/25/00 EDC3 (8d) 10/13/00 8/28/00 EDC

MOD12 Land Cover/Land Cover Change 3 (96d) 3/30/01 6/1/00 EDC

MOD13 Gridded Vegetation Indices (Max NDVI & Integrated MVI) 3 (16d) 8/4/00 6/9/00 EDC

MOD14 Thermal Anomalies, Fires & Biomass Burning 2, 2G 1/19/01 1/1/00 EDC3 (1d) 2/2/01 11/1/00 EDC3 (8d) 10/13/00 8/20/00 EDC

MOD15 Leaf Area Index & FPAR 4 (8d) 8/4/00 6/9/00 EDC

MOD17 Net Photosynthesis, Net Primary Productivity 4 (8d) 2/2/01 10/31/00 EDC

MOD29 Sea Ice Extent 2, 3 (1d), 3(8d) 2/16/01 12/7/00 NSIDC

MOD43 Albedo 3 (16d) 9/29/00 7/11/00 EDCNadir BRDF-Adjusted Reflectance 3 (16d) 9/29/00 7/11/00 EDCBRDF Ross-Li Model 3 (16d) 2/2/01 10/31/00 EDC

MOD44 Vegetation Cover Conversion and Continuous Fields 4 (32d) 3/30/01 2/28/01 EDC


13

Status of the MODIS LandProducts

Geolocation (R. Wolfe, M.Nishama)

The MODIS geolocation algorithm hasbeen working well since launch. The

geolocation accuracy specification is300 m (2 sigma) and the goal is 100 m(2 sigma) at nadir. This goal is primarilydriven by land 250-m change product

requirements. A global distribution ofGround Control Points from 110 LandsatTM scenes are correlated with MODISdata to measure about 100 control point

residuals per day. The basic geolocationalgorithm corrects off-nadir pixellocations for terrain parallax effects byusing the GTOPO30 1-km terrain model.

The at-launch geolocation error of 1.7 kmRMS was larger than expected whenthe first Earth-view data was received inlate February 2000. We quickly com-

puted the first bias correction primarilyin the roll and pitch axes. By applyingthis correction, the error was reduced to500-m RMS starting in late March 2000. A

second correction was performed in June2000 and included a correction to theyaw and the mirror wedge and tiltangles. This reduced the measured error

to 100 m RMS (200 m 2 sigma). Weexpect to make a correction of the smallremaining biases expected in December2000. We are also currently evaluating

effects of the change to Side B mirrorencoders.

In 2001, we will continue further analysisto characterize along-scan mirror motionand any remaining biases. A longer-term

analysis will look at trends and cyclicalvariations: dependence on temperature,time on orbit, etc. We will be examiningperiods near orbit and attitude maneu-

vers and other events to determine whenthe instrument is outside of accuracyspecifications/goals and developing

processes to identify the events. To dothis we will make the island control-point matching algorithm operational,which will allow almost continuous

measurements of the geolocation error.Later in 2001 we will work towardimproving the accuracy in areas ofsignificant terrain variation by analyzing

the accuracy of the terrain model andterrain correction algorithm. We will alsoexamine cross-instrument geolocationissues by comparing results with MISR

and ASTER instrument teams.

Surface Reflectance (E. Vermote, N.El Saleous)

The first released data set of surface

reflectance (starting June 6, 2000) was apreliminary version that did not includethe correction for aerosol effects, i.e.,comparable to the AVHRR Pathfinder II

and SeaWIFS Land Surface Reflectanceproducts. However, due to the better setof bands for MODIS, the effect of watervapor found in AVHRR data are greatlyreduced in MODIS data and the

“composited” data set should be veryuseful to the scientific community.Starting at the end of September (Julianday 273), after an evaluation of the

aerosol product and evaluation of acorrection prototype, a conservativeaerosol correction was applied to thedata, which greatly improves the

product. The correction is still underevaluation and an extension of thecurrent aerosol correction to brightertargets is being developed and should

eventually give near-global correction.Once a substantial data set of goodquality LAI and BRDF products has beenproduced and evaluated, the BRDF/

atmosphere coupling correction will beassessed and applied. The correction foradjacency effect is currently being givenlower priority as the effect on MODIS is

likely to be small, especially at 500-m

resolution, but the correction will beincluded prior to the first major repro-cessing. The impact of the second ordereffects (BRDF and adjacency) on the

derivation of the aerosol and theassociated feedback will be part offurther development. In parallel,validation activities are on-going which

will enable us to refine/and confirm theerror bars advertised in the ATBD andpossibly lead to subsequent algorithmrefinements.

Land Surface Temperature (Z. Wan)

The MODIS Land-Surface Temperature(LST) algorithm is being refined and thePGE16/31 code has undergone a numberof updates and patches. We have made

several refinements to the LST algorithm,reducing identified problems, such asstriping due to the cloudmask associatedwith the fourth noisy detector element in

band 22. The absolute radiometricaccuracy of MODIS TIR channel datawas evaluated with in situ data collectedin a vicarious calibration field campaign

conducted in Lake Titicaca, Bolivia,during May 26 and June 17, 2000. Thespecified absolute radiometric accuracyis reached or nearly reached in MODIS

bands 29-32. It is difficult to obtain adefinitive estimate in other bandsbecause of the polarization effects andlarger effects of the uncertainties in

atmospheric temperature and watervapor profiles. An error in the differenceof brightness temperatures in bands 31and 32 is estimated as 0.3 K, causing an

error of 0.7-0.9 K in the MODIS LSTproduct. A correction for this effect wasmade in the generalized split-windowLST algorithm in late September and

implemented in the operational PGE16code in early October. The accuracy ofdaily MODIS LST product at 1-kmresolution was validated through in situ

measurement data collected in field

THE EARTH OBSERVER

14

campaigns. The MODIS LST accuracywas better than 0.7 K over Mono Lake,California, in two daytime cases and onenighttime case, over a grassland in

Bridgeport, California, in two nighttimecases, and over a rice field in Chico,California, in two nighttime cases. It isdifficult to validate the daytime LST

product over land sites rather than lakes,because of the high spatial variation inthe in situ LST data. Errors in the 1-kmLST product are expected to be larger in

semi-arid and arid regions where surfaceemissivities in bands 31 and 32 arehighly variable. The MODIS LST productat 5-km resolution generated by the day/

night algorithm is currently beingevaluated. The quality of this LSTproduct is sensitive to the calibrationerrors in seven MODIS TIR bands and to

the quality of the MODIS product ofatmospheric temperature and watervapor profiles. The 5-km LST productwill be released in early 2001.

Vegetation Indices (A. Huete, KamelDidan)

The MODIS Vegetation Index (VI)Algorithm has undergone four updatesas a result of improved understanding ofthe instrument data quality and the

performance of upstream products. TheMODIS VI products have benefitedgreatly from improvements in perfor-mance of both the surface reflectance and

the cloud detection algorithms. More-over, the effort by LDOPE to identifyquality and performance issues witheach of the MODLAND products has

helped with further adjustments of theVI algorithm, leading to more completespatial coverage with better handling ofnon-ideal data (e.g., cloudy viewing

angles). The Enhanced Vegetation Index(EVI) is performing well, improving thedetection of vegetation changes wherethe normalized difference vegetation

index (NDVI) has little sensitivity. We arecontinuing analysis of possible problemswith NDVI saturation, which does notappear in the EVI product. Although

some problems were found with the EVIover snow and cloud, the newestalgorithm update (end of 2000) shouldcorrect this anomalous behavior and

further improve the quality of theMODIS VI products. Comparisons withthe AVHRR NDVI show a strongersensitivity of MODIS to vegetation, with

the MODIS NDVI signal being 1.3 timesthat from the AVHRR. A preliminaryglobal quality analysis of the MODIS VIproduct indicates that the compositing

algorithm has greatly minimized view-angle-related problems (encounteredwith MVC) resulting in a more consis-tent, potentially long-term data set. Our

current analysis of the 250-m MODISVI’s shows significantly higher spatialvariations in vegetation amount andstructure. All indications lead us tobelieve that the MODIS VI product will

be robust and reliable, with resultingmajor improvements to all VI-basedapplications and products.

BRDF/Albedo (A. Strahler, C.Schaaf, J.P. Muller)

Albedos and Nadir BRDF-AdjustedReflectances (NBAR) produced by theMODIS BRDF/Albedo Product Algo-rithm (MOD43B) from July 2000 onward

have now been released to the public.BRDF parameters will be released inearly 2001. The algorithm makes use of akernel-based BRDF model and multi-

date, cloud-cleared, atmospherically-corrected MODIS surface reflectances toprovide gridded global spectral andbroadband products every 16 days at 1

km, based on two 250-m and five 500-mbands covering the whole shortwaveregion. While the initial products werederived from surface reflectances that

had not been accurately corrected foraerosols and thus could not be evaluatedquantitatively, qualitative examination ofthe spatial patterns and temporal trends

indicate that the algorithm is behavingvery well and producing consistent andreliable albedo and NBAR quantities.Although some contamination by clouds

also affected the products, the algorithmwas able to consistently flag these areasas poor quality results. Now that theaerosol correction has been implemented

operationally and the cloud mask hasbeen updated, quantitative assessmentsare possible. Improvements were alsomade to QA flags. Validation field data

were collected during the 2000 growingseason at agricultural sites in Beltsville,MD; Luancheng, China; and BartonBendish, UK. Data were also collected at

the SAFARI-2000 site at Mongu, Zambia.Although most of the extensive compari-sons between field data and MODISresults will have to await the reprocess-ing of the initial products, some valida-

tion is currently underway with fielddata that were acquired in October fromBeltsville, MD, and Mongu. Validationactivities include use of AirMISR,

IKONOS stereo, POLDER, and AVHRRdata.

LAI/FPAR (R. Myneni, J.Knjazihhin, Y. Zhang)

The MODIS green Leaf Area Index (LAI)

and Fraction of PhotosyntheticallyActive Radiation absorbed by vegetation(FPAR) are 8-day composite values ofdaily retrievals at 1-km resolution. The

product, technically called MOD15A2,has been released to the public sinceearly August 2000 from the EDC DAAC.Tile and global level quality assurance

(QA) on both the daily and 8-dayproducts has revealed some occasionalradiometric artifact (striping) inheritedfrom the upstream reflectance inputs.


15

Recent temporal analysis of some dailyand 8-day FPAR, LAI images also revealssome moderate radiometricdiscontinuities across period boundaries;

these do not yet appear to reflectsystematic biases, and are now beinginvestigated at the SCF. In general, theFPAR, LAI product appears to be

producing very reasonable biophysicalsignals, when stratified across biometypes and examined across a number ofsuccessive periods. Users are strongly

encouraged to consult the associatedQuality Assessment (QA) files in view ofthe evolving maturity of the product.Validation of MOD15A2 with data from

grasslands and savannas (southernAfrica), broad leaf forests (Massachu-setts), and needle-leaf forest (Finland)are encouraging. Further details on

MOD15A2 algorithm physics,prototyping, data access, user guides,and validation activities can be found onthe LAI Web site provided below.

Snow and Ice (D. Hall, G. Riggs, V.Salomonson)

The MODIS snow algorithms haveundergone several refinements and thecode has been revised several times inresponse to quality assessment of the

products. Changing of MODIS calibra-tion settings affecting the responses ofdetectors has caused problems in thealgorithm, but those problems were fixed

with changes in the algorithm and in theMODIS radiance data product. Therecent move to the B-side electronicsappears to have improved the MODIS

data. Preliminary analysis of MODISLevel 2 snow products (MOD10_L2)generated with B-side data has revealedan improvement in a qualitative sense.

Confusion of snow and cloud persists inthe algorithm and limits the accuracy ofsnow maps in some situations. Investiga-tion of ways to alleviate the snow and

cloud confusion continues. The globalsnow climate modeling grid product,MOD10C1, and algorithm have beenrevised in response to validation studies.

A daily global view of snow cover extentis shown with the MOD10C1 product.Release of the MOD10C1 product isexpected in early 2001. Preliminary

comparisons of the MODIS snowproduct with operational snow productsfrom NOAA have shown good corre-spondence. However, some differences

among the maps have been observed. Itis not known which of the maps is themost accurate. A field and aircraftcampaign was conducted last March in

New Hampshire. Results, which areunder analysis, show good correspon-dence with two NOAA operational snowmaps for that period, except in the

Catskill Mountains in New York. Anupcoming experiment in January 2001will focus on the area where the snowmaps disagree. A fractional snow coverenhancement to the snow maps, and a

daily snow albedo map should beavailable in the spring and summer,respectively.

Sea Ice Products (D. Hall and G.Riggs, V. Salomonson)

Sea ice products contain two arrays ofdata. One is the extent of sea ice asdetermined by reflectance characteristics.The other is sea ice surface temperature

(IST) determined by a split-windowtechnique. Dr. Jeff Key, University ofWisconsin, has provided the coefficientsbased on geographic location and

temperature for the IST algorithm. Amajor revision of the MOD29 algorithmand code is underway to incorporatethose IST coefficients and to incorporate

many changes similar to those made inthe snow algorithm for processingMODIS radiance data. Uncertainknowledge of detector responses in

thermal bands 31 and 32 was used in theIST calculation, and optical crosstalkfactor in the calculation of IST. Recentchanges to B-side electronics and

improved understanding and handlingof those effects in the L1B processinghave reportedly diminished those effectsin the data. An investigation is under-

way. The revised code will be deliveredearly in 2001 and release of the data willfollow shortly thereafter. Preliminaryvalidation work by Shusun Li, from

March 2000 in the Southern Ocean,shows good qualitative correspondencebetween the MODIS sea-ice maps andship observations. This data set is

currently under analysis by Li. Futureefforts will focus on comparison of theMODIS sea-ice maps with NOAANESDIS passive microwave sea-ice

maps. A daily sea ice albedo map shouldbe available by the summer of 2001.

Fire / Thermal Anomalies (C.Justice, L. Giglio, D. Roy, Y.Kaufman)

The MODIS Fire Algorithm is beingrefined and the MOD09/14 code has

undergone a number of patches. Wehave made several refinements to the firealgorithm over the last four monthsremoving or reducing identified prob-

lems associated with sunglint and falsedetections from hot and highly reflectivesurfaces. A new version of the fire codecan be expected in early 2001. We have

encountered several problems associatedwith dead and noisy detectors in band 22and 21, related to electronic crosstalk andincomplete sensor calibration. The recent

move to the MODIS B-side electronicshas improved the quality of the 1B datareducing the number of out-of-familydetectors and we are evaluating the

impact on the fire product. Initialevaluation shows an overall improve-ment in product quality, particularly

THE EARTH OBSERVER

16

with respect to the detection of smallerfires. Initial comparisons betweenMODIS, TRMM, and AVHRR fireproducts are very favorable, however

there are clear differences in fire detec-tion as a result of overpass time. In someof the cases the earlier MODIS overpasstime has shown reduced cloud cover and

clearer viewing conditions. Productvalidation is underway and emphasis isbeing placed on comparison withground and airborne data collected for

fires in the Western U.S. and the SA-FARI-2000 campaign. The latter willinclude the use of ASTER, Landsat,MAS, and field data. An experimental

MODIS burned-area algorithm is beingtested and evaluated associated with thislatter intensive campaign.

PSN, NPP Level 4 Algorithm Status(S. Running, J. Glassy)

The PSN, NPP Level 4 land algorithm isa model-oriented, rather than a primarilyinstrument-driven algorithm. As such, it

is currently the only Level 4 algorithm inthe MODIS land product suite per se,and one of the few requiring time-continuous ancillary climatology data

(from the Goddard DAO). Since it isdriven from the FPAR/LAI product(requiring 1-km aggregated MODISsurface reflectances), it is the final step of

the entire vegetation processing string.In short, any and all upstream gaps inthe production system end up dispropor-tionately affecting the PSN, NPP

products. An implication of this is thatthe PSN, NPP production rules are morecomplicated, since we must track eachpixel’s history throughout the model

year. In particular, the time-continuity ona tile basis is especially important, due tothe way we must continually accumulatethe GPP and respiration terms over the

course of a model year to computeannual NPP for each tile. For this first

data-year, the EOSDIS production chainas a whole has been quite challenged toprovide the level of tile-level timecontinuity required to make reasonably

accurate 8-day PSN products. Thus far,the MOD17A1, A2 string (PGE36, PGE37) chain has only been run successfully(globally) at the University of Montana

SCF. Our most recent global runs usingMODIS 8-day FPAR, LAI, and DAOinputs represents the period June 29 toJuly 6 and September 7-14 2000 and may

been seen from our Web page (below).Several recent changes have dominatedthe current MOD17 activities. These are:a) changes in the MOD15A1 (daily

FPAR) and MOD15A2 (8-day FPAR)pixel level QA organization and interpre-tation to better track non-vegetationpixels in the final product, b) the recent

discovery and correction of a units-labelerror (by the DAO) in the specifichumidity variable (was labeled as g/kg,should have been labeled kg/kg) thatcaused seriously depressed GPP values,

and c) the addition of the 5-km coarselogic to match the daily and 8-day FPAR,LAI products.

MODIS Land Cover (A. Strahler, M.Friedl)

The MODIS land-cover algorithm iscurrently generating early results from

the available MODIS data. These resultsare currently being evaluated at theScience Computing Facility in Boston.The algorithm employs a supervised

classification methodology, and a globaldatabase (n > 900) of site data compiledfrom Landsat Thematic Mapper, whichare used in the final stages of quality

assurance. Initial global runs are beinggenerated using three dates of nadir andBRDF adjusted reflectances for sevenMODIS bands. The results from these

runs are very encouraging and attest tothe spectral and radiometric quality of

MODIS data for land-cover applications.As more dates are added to the classifi-cation runs, we expect that the quality ofthe land-cover maps will improve

accordingly. The land-cover product istargeted for release in April 2001.

Vegetative Cover ConversionProduct (J. Townshend, X. Zhan, R.Sohlberg)

The MODIS 250-m Vegetative CoverConversion (VCC) (MOD44A) product is

currently in the initial data collection andprocessing stage. Generation of theproduct requires composited SurfaceReflectance data for the two 250-m bands

for two 32-day periods which are at leastthree months apart. For the reasonsstated above, the 250-m Surface Reflec-tance data processed to Level 2G have

been available only for the conterminousUnited States for limited 16-day periods.The first 32-day composite will beavailable during December. At present

we are running the VCC code withlimited inputs to generate some earlyresults. We are expecting that the Look-Up Tables (LUTs) derived originally with

AVHRR data and used by the currentVCC code, will need to be updated usingMODIS retrievals. The procedure forupdating the LUTs is being conducted

with the available June and SeptemberMODIS data. Validation of the VCCproduct will be carried out with LandsatETM+ images, IKONOS 1-m and 4-m

data, and ground observations whereavailable. An automated change-detection procedure is being preparedfor using the ETM+ data to validate the

land cover changes indicated by the VCCproduct. The planned date of publicrelease for the first MODIS VCC productis March, 2001.


17

Vegetation Continuous FieldsProduct (J. Townshend, R. DeFries,M. Hansen)

The MODIS Vegetation Continuous

Fields (VCF) product is a post-launchproduct. The generation of the productrequires monthly composites of 500-msurface reflectance data for a full year.

The VCF product generation code hasbeen prepared with algorithms used forthe prototype VCF product generatedwith AVHRR data. The procedure is

being updated with MODIS datacurrently being collected. A validationdata set for Kalahari woodlands wascollected as part of SAFARI 2000

activities. This will be complemented bya Miombo data set to be collected during2001. The validation data combinesIKONOS fine resolution data with

ground observations collected using alaser canopy closure instrumentation. Ina separate validation activity, canopydensity data are being collected for a

transect across the Appalachian Moun-tains from Washington, DC, to Colum-bus, Ohio. This activity utilizes a four-band airborne radiometer with a spatial

resolution of 1 m. Leaf-on data werecollected during July 2000. The leaf-offflight is scheduled for late March 2001.The planned date for releasing the first

MODIS VCF product is July 2001.

Overall Goals for MODLAND

The overarching goals for MODLAND

include: developing the next generationof new and improved global land dataproducts of known accuracy (i.e.,validated products); establishing the

relationship and continuity with existingland data sets; providing the proof-of-concept for land products to be used bythe next generation of operational

sensors, i.e., the NPOESS Visible InfraredImaging Radiometer Suite (VIIRS); and,most importantly, developing a consis-

tent science-quality climate data recordof high-priority land products for thescience community. This latter task willnecessitate reprocessing of the MODIS

data record. The first major reprocessing,to be undertaken in 2001 will provide thetime-series of ‘science-quality products.’The reprocessing will fill in the large

gaps that occurred in the global dataduring the first six months and willapply revised versions of the code,modified to account for instrument

performance and initial validationresults. There are major concerns thatreprocessing at 0.75X of the 96 baselinelevel (i.e., less than one day in a day) will

be inadequate to respond to the sciencecommunity immediate need for a timelydelivery of one year of science qualitytime-series products. Clearly, rapid

development of new and innovativeapproaches to large-volume dataprocessing, reprocessing, and archivingneeds to be given a high priority if weare to realize the potentially enormous

science pay-off from the MODIS data. AMODIS reprocessing ‘pathfinding’activity will be started in January toestablish procedures and logistics for

large-volume reprocessing with thecurrently available hardware. Issues ofdata retrieval and staging will needcareful attention. Candidate periods for

the reprocessing trial include July, toprovide data sets for SAFARI 2000validation, and April, coincident with anoceans validation initiative.

WWW Sites Associated withMODIS Land Products

(To Obtain MODIS Land Data)

EOSDIS Data Gatewayhttp://eosdatainfo.gsfc.nasa.gov/eosdata/terra/data_access.html

Level 2/3 Land Data / EDC DAAChttp://edcdaac.usgs.gov/

MODIS L2/3 Snow and IceProducts/ NSIDC DAAChttp://nsidc.org/NASA/MODIS/

Level 1b, Geolocation, Cloud Mask/GSFC DAAChttp://daac.gsfc.nasa.gov/CAMPAIGN_DOCS/MODIS/index.shtml

MODIS 250-m Datahttp://modis-250 m.nascom.nasa.gov/

250 m V.I.http://glcf.umiacs.umd.edu/MODIS2/

main.html

MODIS Sample Imageryhttp://modland.nascom.nasa.gov/gallery/

(MODIS Overview Information)

MODLAND Sitehttp://modis-land.gsfc.nasa.gov/

MODIS Sciencehttp://tpwww.gsfc.nasa.gov/MODIS/

MODIS.html

Land Quality Assurancehttp://modland.nascom.nasa.gov/QA_WWW/qahome.html

MODIS Land Validationhttp://modis.gsfc.nasa.gov/MODIS/LAND/VAL/MCST

MODIS Instrument Performancehttp://mcstweb.gsfc.nasa.gov/

MODIS Science Support Teamhttp://ltpwww.gsfc.nasa.gov/MODIS/SDST/

Direct Broadcast Sitehttp://ltpwww.gsfc.nasa.gov/MODIS/MODIS.html

THE EARTH OBSERVER

18

(MODIS Land Product Web Sites)

Surface Reflectance/AtmosphericCorrection Algorithm Productshttp://modis-land.gsfc.nasa.gov/mod09

Land Surface Temperature (LST) andEmissivityhttp://www.icess.ucsb.edu/modis/modis-lst.html

Leaf Area Index (LAI) andFractional PhotosyntheticallyActive Radiation (FPAR)http://cybele.bu.edu/

Vegetation Indiceshttp://gaea.fcr.arizona.edu/projects/modis/index.html

BRDF and Albedohttp://geography.bu.edu/brdf/index.html

Snow and Ice Coverhttp://snowmelt.gsfc.nasa.gov/MODIS_Snow/modis.html

Fire and Thermal Anomalieshttp://modis-fire.gsfc.nasa.gov/

Vegetation Production, Net PrimaryProductivity (NPP)http://www.forestry.umt.edu/ntsg/RemoteSensing/modis

At-Launch Land Coverhttp://www.geog.umd.edu/landcover/

modis.html

Land Coverhttp://geography.bu.edu/landcover/index.html

Vegetation Continuous Fields andVegetative Cover Conversionhttp://www.geog.umd.edu/landcover/modis.html

It is generally recognized that develop-ing a data system to support processing,archiving, and distribution of NASA’s

Earth Science data is an extraordinarilydifficult endeavor. Part of the difficultylies in the technical problems associatedwith the very high throughput required

of the system, as well as the highprocessing power, the large storagevolume, and the complexity of makingthe data accessible to a wide variety of

data users. Another part of the difficultylies in the fact that this system mustengage a large number of differentcommunities that are unaccustomed to

exchanging data with each other. Thedevelopment process has been difficult,and the initial operations have been alearning experience for all parties

involved. It is a tribute to the hard workof many members of these communitiesthat the Earth Observing System Dataand Information System (EOSDIS) is

beginning to produce and distributedata. The data volumes currently beinggenerated are unprecedented for theEarth Science Community.

A few weeks before the end of 1999, theEOS Terra satellite was launched andEOSDIS began to work with the data

from its instruments. These instrumentsbegan observing the Earth near the endof February and have continued their

observations since. At the March 16-17meeting of the Science Working Groupfor the AM Platform (SWAMP), SWAMPmembers noted that there needs to be

on-going evaluation of how well thecurrent EOSDIS can support the Terra,Aqua, and Aura missions for which ithas been designed. While there has been

considerable discussion concerning thetype of Data and Information System(DIS) required beyond the currentEOSDIS implementation, there has notbeen a systematic evaluation of how well

the current system is working. ChrisJustice (of the MODIS Land Team)accepted a motion that he shouldconvene a workshop of data representa-

tives from the Terra instrument teams tocarry out this evaluation and report tothe SWAMP regarding both theirconcerns and possible solutions.

On June 1-2, the workshop was held atNASA Goddard Space Flight Center(GSFC). It constituted the inauguralmeeting of a Science Working Group on

Data (SWGD). Participants included (1)Terra instrument team representativesfrom CERES, MISR, MODIS, MOPITT,(2) EOS Project and Earth Science Data

and Information System (ESDIS) Projectrepresentatives, (3) EOS Project ScienceOffice representatives, including theTerra, Aqua, and ESDIS Project Scien-

Formation of the Science WorkingGroup on Data (SWGD)— Bruce R. Barkstrom ([email protected]), Atmospheric Sciences, NASA

Langley Research Center— Chris Justice ([email protected]), University of Virginia


19

tists, and (4) Distributed Active ArchiveCenter (DAAC) and Science Investigator-led Processing System (SIPS) representa-tives. Prior to the workshop, the partici-

pants set the following specific objec-tives:

1) Assess the current status of the datasystem for Terra, and, if possible,

assess the planned capacity for Aqua;

2) determine how requirements havechanged with respect to the currentdesign and system capacity;

3) identify mismatches between current

plans and needed capabilities; and

4) recommend solutions to meetevolving needs.

While there were production problemsduring the middle of the summer, the

immediate EOSDIS situation appears tohave improved as early EOS Data andOperations System (EDOS) problems arebeing solved. However, EDOS remainsbacklogged*. Beyond these immediate

problems, the EOS data system remainsunstable and appears to be systemati-cally undersized. There is concern thatthe lack of hardware will delay the

scientific validation effort by 2-3 years.The preliminary SWGD hardwareappraisal needed to deal with validationand reprocessing shows a need of about

$15 million in computer hardware ineach of FY01 and FY02 to solve theproblems, with small additions in lateryears (~$2M). The problem goes back to

unrealistic downsizing of hardwarewhen budget problems hit a few yearsago.

During the workshop, the participantsdiscussed the current operating status ofEOSDIS, and, in particular, the lower-than-expected throughput and how it

should be addressed. They noted that theFebruary 1996 baseline sizing used toimplement EOSDIS is not adequate tosupport the science data needs. Because

that baseline was established before thealgorithms were developed and could berun in the production environment, itdid not have a clear empirical basis. In

addition, the 1996 baseline** does notappear to have been based on previousNASA experience in validating andproducing Earth science data. Terra

instrument team representativespresented revised system sizing esti-mates based on current experience andimproved understanding of the EOS

production environment. The groupnoted that the current performance ofthe system has yet to meet an opera-tional level of production equivalent tothe Option A+ first year capacity

of 1x of Level 1 products and 0.5X ofLevel 2 and higher products as volumesspecified in the 1996 baseline.

The primary finding from the workinggroup was the need for a marked andrapid increase in the system capacity togenerate data products. The current

budget situation is difficult on all sides.It is clear that there will need to be frankand open discussions between all of theparties involved in EOSDIS regarding

possible options. One area of concern isthe impact of the rapid evolution ofinformation technology on the obsoles-cence of the current system. This impact

suggests that we may need to moverapidly from the current system to the

more distributed system being envi-sioned for NASA’s New Data andInformation Systems and Services(NewDISS).

In addition to this finding, the workshopcovered a greater range of challengesand opportunities for improvements to

the service EOSDIS can provide to itsuser community. Some specific chal-lenges for the coming months and yearsthat were identified at the workshop

include:

• reprocessing;

• perturbations to operations, particu-larly those involving the Aqua turn-on and the associated increase involumes and loads;

• data distribution;

• dealing with the diversity of theproducer and user communities;

• obsolescence; and

• transition to a non-ECS environment.

This workshop provided a useful forumfor sharing views and experiencesamong the DAAC’s, ESDIS, and the

instrument teams. It succeeded inidentifying more accurate and meaning-ful system-capacity requirements thanare currently implied by the existing

baseline, which dates from February1996. The SWGD expects to helpmaintain these estimates in the future inconjunction with ESDIS, the DAAC’s,

and the SIPS. There is a need to extendthe SWGD-type of activities to includethe Aqua and Aura teams. Neither ofthese teams were able to participate in

the workshop.*The EDOS backlog at the time of the meeting(June 2000) was more than 1000 hours andgrowing. At the end of December 2000, it was 700hours, being worked off at the rate of 70-100hours/week.

**At the end of December 2000, the productionwas exceeding the Option A+ baselinerequirements in both processing rates and volumesof data archived.

THE EARTH OBSERVER

20

The semi-annual science working group

meeting for Terra was hosted by theMOPITT Instrument Team at theUniversity of Toronto, September 7-8,2000. Many thanks to PI Jim Drummond

and the MOPITT team for selecting abeautiful time and venue for thissuccessful meeting.

Yoram Kaufman started off the meetingwith the reminder that he was steppingdown as Project Scientist and that JonRanson, his Deputy for the past four

years, would succeed him. He alsoannounced that Si-Chee Tsay has beenselected as the new Deputy ProjectScientist. Yoram will continue his

activities as a MODIS team member andplans to stay involved in Terra educationand outreach activities.

Kaufman reiterated the need for aspacecraft maneuver to provide a viewof deep space for calibration of MODISand CERES. He mentioned that Terra

Level 1 data were mostly released(ASTER forthcoming) and Level 2 and 3data products were being made avail-able. Kaufman also stated his concern

that the EOS Data Information System(EOSDIS) was underscoped with respectto data processing margin and reprocess-ing capacity, especially with the Aqua

launch and the expected increase in dataprocessing next year.

Science Working Group on Data

In response to Kaufman’s request at thelast SWAMP meeting, Chris Justice wasasked to hold a workshop to examinecurrent capabilities and expected needs

of Terra data processing. Subsequently, aScience Working Group on Data (SWGD)was formed among representatives ofthe Instrument Teams, DAAC’s, and

ESDIS. A report was prepared by B.Barkstrom (CERES), F. Bordi (EOSProject), G. Bothwell (MISR), C. Justice(MODIS), E. Masuoka (MODIS), R. Wolfe

(MODIS), and D. Ziskin (MOPITT) andpresented to the SWAMP. A summary ofthe report is offered on page 18 of in thisissue.

It has since been decided to give theSWGD a prominent role dealing withEOS data issues. Jon Ranson and Skip

Reber, EOSDIS Project Scientist, took theaction to include Aqua Project Scientist,Claire Parkinson, and develop a charterfor this group. The charter is now in the

final stages of approval.

ASTER

Yasushi Yamuguchi from Nagoya

University presented the status ofASTER on behalf of Japanese PI, HirojiTsu. As of September 1, ASTER hadacquired over 50,000 scenes. These data

are expected to be released in the UnitedStates through the EROS Data Centerduring November. Issues discussed atthe SWAMP included changes in

calibration lamps resulting from thelamps starting brighter than expected,some delay in implementing the Version1.0 of the processing software and the

desire by the ASTER team to acquire 16minutes of data over two orbits ratherthan the 15 minutes that are currentlyallotted. Yamaguchi showed some high

quality examples of ASTER data frompresentations made earlier at IGARSS’00. Anne Kahle followed with moreASTER images including a very interest-

ing composite of Shuttle Radar Topo-graphic Mapper data and ASTER of theMiyake-Jima, Japan, volcano.

CERES

Bruce Barkstrom of NASA Langleydiscussed the status and issues regardingthe CERES instruments on Terra. Allplanned activation events have been

accomplished except the deep-spacecalibration maneuver. The Terra CERESinstruments are performing very well.Barkstrom mentioned that problems

with the EOS Data Operation Systemresulted in missing data for CERES overthe summer. For the most part themissing data can be recovered or

compensated for. He suggested thatsuccess criteria be established formonthly data ingest to gauge theeffectiveness of the data system for

science activities. Given the scope of thecurrent data processing and reprocessingcapabilities, Barkstrom suggested thefollowing science impacts: greater

uncertainty in products, 2-4 year delay inproducing validated data, reduced usercommunity, limited commercial activitiesand limited school involvement.

Terra Science Working Group(SWAMP) Meeting

— Jon Ranson ([email protected]), Terra Project Scientist, NASA Goddard SpaceFlight Center


21

MISR

David Diner, MISR PI, agreed thatreprocessing of data was an issue thatneeds to be addressed. He mentioned

there has been public release of MISRLevel 1 data since June; Level 2 will beavailable for preview in the near futureand routinely available later in the fall.

He presented some interesting imagesshowing MISR derived cloud heightsand wind vectors. He also showedexamples of bidirectional reflectance

properties of smoke. Diner stated thatcurrent MISR data reprocessing capabil-ity is inadequate.

MODIS

Vince Salomonson, MODIS PI, statedthat MODIS is working very well andthat there are data products available for

atmosphere and land disciplines. Oceanproducts are in progress but more time isrequired before they can be released tothe public. Instrument issues concerningthe A/D converter, striping, and bi-

directional properties of the mirror arebeing resolved. Salomonson stated thatMODIS direct broadcast is working andcited the recent use of data by the forest

service to help locate wildfires inMontana. At least 20 requests for directbroadcast software have been filled.Recent examples of MODIS images for

arctic ice, Montana fires, atmosphericmoisture, subvisible cirrus clouddetection, and African dust werepresented.

MOPITT

Jim Drummond described the status ofMOPITT as good with a few issues being

worked regarding possible minorleakage of length modulation cell anddecreasing current in a Pressure modula-tion cell. John Gille stated that MOPITT

level 1 data were to be released inOctober and Level-2 scheduled forrelease April 15, 2001. Level 3 beta dataare expected by October 2001.

Terra Spacecraft

Paul Ondrus reported on the status ofthe Terra Spacecraft and indicated that

Terra is in working order. Main issuesconfronting the operations team are anupcoming deep space calibrationmaneuver, and the need to develop a

plan for Terra’s equatorial crossing time.Regarding the first issue, a study is beingconducted to examine possible deepspace calibration maneuver scenarios

that could result in loss of control of thespacecraft and assumption of a sun-pointing safe hold. These events could bedetrimental to ASTER since the sun

could track across the instrumentapertures. Currently, likely scenarioshave been simulated and found not topresent danger to ASTER. Furtherscenarios were simulated to understand

the effects of loss of one of the reactionwheels on the control of the spacecraftduring the maneuver and the possiblerecovery strategies. A final report and

recommendation to proceed was to bemade in November with the actualmaneuver to be performed in February2001. (Currently, the review process is

on-going and approval of the maneuveris still pending). The crossing time is anissue since currently Terra is at a crossingtime around 10:45 am and is on a

trajectory to get to a 10:30 am crossingtime in about two and one-half years. Adecision must be made as to when toinitiate the series of burns to keep the

crossing time from becoming progres-sively earlier. This issue will be workedover the next few months after theproject has provided the SWAMP a

report on various options.

ESDIS response to the SWGD

Dolly Perkins presented a summary ofthe ESDIS situation and positionregarding the recommendations of the

SWGD. She reviewed the process thatled to the current set of ESDIS require-ments, the now famous A+ Option, thatwere developed from a February 1996

baseline. She then outlined the ESDISactivities directed at the problemshighlighted by the SWGD report, andindicated future solutions will require

efficiencies both within EOSDIS and theinstrument teams. She stated thatwithout a budget augmentation therecommendations of the SWGD could

not be accommodated in the near term.

Data Production Status

Vanessa Griffin updated the SWAMP on

the EOSDIS status and plans. The datasystem is capturing over 350 Gb/dayincluding nearly 200 Gb/day for Terra.After the Aqua launch next year this datarate will increase to 480 Gb/day. Archived

data for Terra has reached 850 Gb/day.She summarized the problems experi-enced by the EOSDIS Data OperationsSystem that impacted Terra data

production. These included a series ofproblems with tape drives, ftp, backupand operations that conspired to create alarge data processing backlog during

May through July. Efforts by ESDIS toresolve these problems were alsodiscussed. The data system has im-proved greatly with smoother operations

since late August. The backlog of datacontinues to be worked off.

Terra Beta Data Products

Terra Beta Data Products are beingreleased by the instrument teams. Beta


THE EARTH OBSERVER

22

2nd Science Team Meeting for theSolar Radiation and ClimateExperiment — SORCE— Gary Rottman ([email protected]), University of Colorado.— Robert Cahalan ([email protected]), NASA Goddard Space Flight Center.

Introduction

The SORCE Science Team Meeting washeld September 13-15, 2000, inSnowmass, CO. This is now a criticaltime with the four instruments (TIM,

SIM, SOLSTICE, XPS – see the 1999 EOSReference Handbook) designed andbeing built, but still awaiting fullcharacterization, calibration, and testing.

The meeting provided a forum toacquaint the solar and climate communi-

SORCE Science Meeting 2000, September 13, 14, 15, 2000 Snowmass, Colorado Attendees

Gail Anderson, NOAA, Boulder, CO; Jim Brault, LASP, Univ. of Colo.; Rock Bush, Stanford SolarObservatory, Stanford Univ. Stanford, CA; Robert Cahalan, NASA Goddard Space Flight Center,Greenbelt, MD; Gary Chapman, CSUN, San Fernando Observatory, Northridge, CA; DominiqueCrommelynck, Royal Meterological Institute, retired, Brussels, Belgium; Matt Deland, RaytheonTechnical Services Co., Lanham, MD; Giuliana DeToma, LASP, Univ. of Colo., Boulder, CO; GingerDrake, LASP, Univ. of Colo., Boulder, CO; Peter Fox, HAO, NCAR, Boulder, CO; Claus Frohlich,PMOD, World Radiation Center, Davos Dorf, Switzerland; Jerry Harder, LASP, SORCE SIM InstrumentScientist, Univ. of Colo., Boulder CO; Greg Kopp, LASP, Univ. of Colo., Boulder, CO; Robert Kurucz,SAO CfA, Harvard, Cambridge, Mass; George Lawrence, LASP, SORCE TIM Instrument Scientist,Univ. of Colo., Boulder, CO; Judith Lean, Naval Research Lab, Washington, D.C; John Leibacher,National Solar Observatory, Tuscon, Arizona; Sasha Madronich, Mesa Lab, UCAR, Boulder, CO; BillMcClintock, LASP, SORCE SOLSTICE Instrument Scientist, Univ. of Colo., Boulder, CO; Jim McGuire,NASA Goddard Space Flight Center, Greenbelt, MD.; Bill Ochs, NASA Goddard Space Flight Center,Greenbelt, MD, SORCE Mission Manager; Chris Pankratz, LASP, Univ. of Colo., Boulder, CO; PeterPilewskie, NASA Ames Research Center, Moffett Field, CA; George Reid, NOAA, Boulder, CO; DavidRind, NASA Goddard Space Flight Center, Greenbelt, MD; Ray Roble, HAO, NCAR, Boulder CO; GaryRottman, LASP, SORCE Principal Investigator, Univ. of Colo.; Rodney Viereck, NOAA, SpaceEnvironment Center, Boulder, CO; Dick White, HAO, NCAR, Boulder, CO; Dick Willson, Center forClimate Systems Research, Columbia Univ., Coronado, CA; Martin Woodard, Big Bear SolarObservatory, New Jersey Institute of Research, Big Bear, CA; Tom Woods, LASP, SORCE XPSInstrument Scientist, Univ. of Colo., Boulder, CO.

ties with the instruments’ capabilities

and measurement program. It is also acritical time in the development ofinstrument observing sequences andobservation plans, and the meeting

provided an ideal opportunity to reviewmeasurement requirements. Thesediscussions insured that the individualsresponsible for the instruments were

fully aware of the recent developmentsin solar physics as well as the presentneeds of the atmospheric and climate

communities. A unique aspect of thisSORCE Science Team was, therefore, theparticipation of specialists in the varietyof solar and terrestrial physical processes

that influence Earth’s climate. In fact, asignificant aspect of the first two SORCEteam meetings in Taos and Snowmasshas been the communication across these

two disciplines. Consequently, terrestrialscientists on the team have learned muchabout the Sun, and likewise solarscientists have learned about the Earth.

Collaboration over a wide spectrum ofdisciplines is essential to achieving thegoals of SORCE.

Instrument Status and Overview

The presentations at the Science TeamMeeting provided an overview of thefour instruments’ capabilities, since more

detailed information has been reviewedand critiqued in a number of projectreviews. See the SORCE Web site (http://lasp.colorado.edu/sorce/) for moredetail.

George Lawrence and Greg Koppdescribed the Total Irradiance Monitor(TIM). This is a four-channel active

cavity radiometer adhering to the basicprinciples of electrical substitutionradiometry (ESR) — that is, a balancedradiometer that substitutes a precise

amount of electrical power or Jouleheating to replace solar radiation thathas been intermittently removed by ashutter. This is the same observing

technique used by the ACRIM instru-ment series, the VIRGO instruments onSOHO, and other devices measuringTotal Solar Irradiance (TSI). TIM intro-

duces several improvements through theuse of modern materials and enhancedelectronics. However, its major advancewill be the use of phase lock detection at

the shutter frequency resulting in adecrease in the measurement uncertainty


23

of at least a factor of 10 — determinationof the aperture area will remain thelargest uncertainty factor (≈ 70 ppm).

Jerry Harder provided an overview ofthe Spectral Irradiance Monitor (SIM).This spectrometer covers the spectralrange 200 nm to 2 nm, and is intended to

provide the first, reliable measurementof the visible and near infrared irradi-ance of the Sun. The challenge is toachieve precision and accuracy suitable

for establishing solar variations that arelikely smaller than 0.1%. The newinstrument design is a small prismspectrometer using only a single optical

element and a miniaturized version ofthe ESR’s developed for TIM. The SIM isfully redundant and self-calibrating.

Bill McClintock reviewed the SolarStellar Irradiance Comparison Experi-ment (SOLSTICE). This is a secondgeneration version of the highly success-ful UARS SOLSTICE which has been in

continuous operation since September1991. This spectrometer will cover thespectral range 112 to 320 nm, but inaddition to measuring solar irradiance it

has the unique capability of observingbright, blue stars with the very sameoptics and detectors. These stars becomethe “standard candles” against which the

Sun is compared, and they serve as amethod of establishing changes in thesensitivity of SOLSTICE throughout theSORCE mission. However, the solar/

stellar comparison technique serves anadditional purpose and allows a directcomparison of the SORCE solar observa-tions to those made by UARS, or by any

other future SOLSTICE-type observation.

Tom Woods provided information on theXUV Photometer System (XPS). This

instrument has 12 photodiodes with ninedifferent filters to isolate spectral bandsin the XUV from 1 to 35 nm, and at

Lyman-a. It has heritage from otherLASP programs including SNOE andTIMED SEE.

All data processing for SORCE will occurat the Science Operations Center (SOC)at LASP. Chris Pankratz reported that allSORCE data will be processed within 24

hours of reception. Provisional productswill then be distributed via the SORCEWeb site within 48 hours, and fully-calibrated science products will be

available approximately three monthsafter reception. These standard dataproducts, as well as the raw data andassociated metadata, will be archived at

the GSFC DAAC.

Present TSI Observations

There are presently five different

instruments returning TSI data: twoACRIM instruments, one on UARS andone on ACRIMSAT, two instruments onthe ESA/NASA SOHO spacecraft, andthe ERBE on ERBS. It is not inconceiv-

able that several, if not all, of theseinstruments will still be returning data atthe time of the SORCE mission. Suchdata overlap will provide a tremendous

enhancement to the value and reliabilityof the SORCE data but, moreover,sharing the experience and findings ofthese programs provide valuable insight

to the present SORCE development.Dick Willson of Columbia Universitydescribed recent achievements of theACRIM program which today includes

two instruments, one on UARS operat-ing for the past nine years and the otheron the recent ACRIMSAT, launchedDecember 20, 1999. The latter ACRIM III

has gathered more than 60 minutes ofhigh quality TSI data per orbit sinceApril 5, 2000, and these data have beenarchived at NASA’s Langley Research

Center’s (LaRC) Distributed ActiveArchive Center (DAAC) since June 2000.

The initial calibration phase for theACRIM sensors was completed October2000, and archiving of fully calibratedresults has begun on a regular basis.

Willson presented the initial ACRIM IIIresults compared with results from hisother on-going TSI monitor, the UARS/ACRIM II experiment. The close agree-

ment is extremely encouraging anddemonstrates the value of having at leasttwo TSI devices making simultaneousobservations.

Claus Fröhlich of PMOD and DominiqueCrommelynck of the Royal Meteorologi-cal Institute of Belgium presented the

latest results from their respectiveinstruments on SOHO launched late in1995. Claus is the Principal Investigatorfor the VIRGO instrument and discussed

TSI observations from the PMO6Voperational radiometers. The secondtype of radiometer is the DIARADprovided by Dominique and his col-leagues (analyses by S. Dewitte). Both

VIRGO instruments provide excellentTSI data that enhances and extends thetime series of the ACRIMs and ERBE.Moreover, by operating from the same

space platform, the two VIRGO instru-ments provide simultaneous datarecords for direct intercomparison. Bothinstrument teams have completed an

initial age-dependent correction to theirdata, and the subsequent comparisonshows differences now to be corrected ina second phase. An initial interpretation

indicates that a simple exposure depen-dent degradation will not suffice, andthat unknown and more complicatedprocesses likely play a role.

Dominique has a second instrument,SOLCON, that has flown twice as aShuttle Hitchhiker, with another flight

scheduled within the upcoming year(STS 107). The SORCE program isbuilding two TIM instruments in

THE EARTH OBSERVER

24

addition to the one to be flown onSORCE. One of these extra units will bepreserved at LASP as a “witness unit,”and the second, referred to as the

Traveler, is intended to move outside thelaboratory and participate in variousintercomparisons. The SORCE programhas received NASA Headquarter’s

approval to fly this “traveling” unit as aHitchhiker payload with SOLCON,perhaps as early as the fall of 2002.

Understanding Solar Variations

Most of one full day was devoted todiscussions of our present understand-ing of the Sun’s radiant output, and to

those variations that should be detect-able by SORCE. SORCE TSI data, andperhaps to a greater extent the spectraldata, will include fluctuations of a solar

origin, but intermixed with inherentnoise and artifacts of the instrumentand/or operations. Experts in solarvariability provided important tutorialson solar physics and observations.

Claus Fröhlich provided additionalanalysis of the SOHO data, consideringreconstructions of both the TSI data and

the three spectral channels (402, 500, and862nm). The short-term (~ 27-day)variations seen in the SOHO time seriesare adequately modeled using limited

knowledge of solar surface magnetism(including sunspots and faculae). RockBush of the SSSC at Stanford Universitydescribed the observations of the

Michelson-Doppler Imager (MDI) onSOHO. This instrument images the Sunwith 4” spatial resolution, throughincreasingly narrow spectral filters. In

general, MDI sends down completeintensity and velocity (Doppler) imagesevery minute

John Leibacher of Kitt Peak thenprovided an informative overview of the

characteristics of pressure (p-mode) andinternal gravity (g-mode) oscillations inthe Sun. Present understanding of thesephenomena are provided by the tempo-

ral and spatial properties of solarobservations — observations now beingprovided by the MDI on SOHO as wellas from ground based observations of

the Global Oscillations Network Group(GONG). Decomposing the velocity(Doppler) images into spatial sphericalharmonics provides power spectra and

identifies the oscillation modes. SORCEwill not provide data relevant tohelioseismology (for example, TIMprovides neither spatial nor velocity

information), but understanding andaccounting for these solar oscillationswill insure that our SORCE samplingscheme will not alias these fluctuations.

There were a number of talks describingcontributions from today’s ground-basedobservations. These observations includeprecision photometry of the full solar

disk and image data as well. GaryChapman described activities at the SanFernando Observatory using twotelescopes. Because of large atmospheric

corrections, ground-based observationsuse relative photometry, and then studythe magnetic features on the Sun interms of their contrast against a constant

quiet Sun. Chapman et al. constructsimple photometric models and comparethem to TSI data (e.g., NIMBUS-7 ERBdata) finding correlations > 0.95. These

ground-based programs allow “gap-filling” in the space record of TSI and theidentification of instrumental jumps andoffsets in the record (e.g., SFO found two

jumps of approximately 250ppm in theNIMBUS-7 data). Martin Woodarddescribed observations made since 1993by the SolarDisk Photometer at Big Bear

Observatory. These high-precisionmeasurements are being analyzed tounderstand the physical origin of the

Sun’s luminosity variations.

Observations of Solar SpectralIrradiance

One of the major new contributions of

SORCE will be the first accurate mea-surement of solar spectral irradiance inthe visible and near infrared. Noprevious observations, from the ground

or from space, have had adequateprecision and accuracy to establish solarvariability — a variability that isexpected to be no more than 0.1%. At

shorter wavelengths in the ultravioletthe variation is larger, and previousprograms (e.g., UARS) have reportedsolar variations over time scales from

minutes, to weeks, to the much longer11-year solar cycle. Several talks summa-rized our present understanding of andthe importance of the new SIM observa-

tions.

Jim Brault described the currentlyexisting Kitt Peak high-spectral resolu-tion data. He included other observa-

tions as well, and considered how theycompared with the Kitt Peak data. Thebest available models of solar spectralvariability, and also models of atmo-

spheric transmission, require high-resolution data. SIM has been designedwith attention to its precision andhighest possible photometric accuracy,

but with its moderate spectral resolutionit will not provide information on a line-by-line basis. It will be essential tointerpret the SIM spectra in conjunction

with the higher resolution groundobservations, an analysis that benefitsfrom the continuous and “seamless”nature of the SIM observations.

Little is known about solar variability inthe visible and near infrared, due to thelarge corrections that must be made for

atmospheric absorption and due to the


25

fact that the solar variations are small(≈ 0.1%). From space measurements ofthe past 30 years we have informationthat the ultraviolet (λ < 300 nm) varies

by far more, increasing to 10’s of percentbelow 200 nm and factors of 2 to 10below 100 nm. Tom Woods provided areview of the ultraviolet observations

made from space. In particular, theSOLSTICE and SUSIM instruments onUARS, the SBUV instruments, and theGOME on ERS-II now give a good

empirical description of the wavelengthvariation of the Sun’s radiative outputover the last two solar cycles. Bycombining these data sets, we now have

a record of UV variability from 1978 tothe present.

When full spectral observations are not

being made, or when the quality of theobservations is in question, a “proxy”may be used to estimate irradiance. Themost widely used proxy or index for theUV irradiance is the Mg II index that

uses the core-to-wing ratio of the Mg IIabsorption feature at 280 nm. Instrumentresponse changes should be cancelledout by taking an irradiance ratio using

equally spaced wings. Matt DeLandpresented long-term data sets producedsince November 1978 by SBUV andSBUV/2 instruments onboard Nimbus-7,

NOAA-9, NOAA-11, the SUSIM andSOLSTICE instruments on UARS, andthe GOME instrument on ERS-2. Threenew instruments will be launched

during the next two years to continuethis valuable data record.

Solar flares are a well known solar

phenomenon that might appear in thetime records of solar irradiance variabil-ity. Such transient and highly energeticevents occur in the solar corona, but

provide only small disturbance in thephotosphere. Consequently, althoughTSI data records have been closely

examined, flares have not been seen atthe level of about 50 ppm. In theultraviolet and extreme ultraviolet thesituation is of course quite different, and

Gary Rottman described flares that areobvious in the ultraviolet time series(SME and UARS). Strong flares can morethan double the irradiance, but only at

wavelengths short of 160 nm and onlyfor short time periods of five to tenminutes.

NOAA presently monitors the solarultraviolet and X-rays from a family ofpolar and geostationary platforms, andin 2002 they will add EUV capability.

Rod Vereick described these on-goingNOAA programs and introduced thefuture NPOESS spacecraft as well. Theinstruments planned for NPOESS

include a TSI device, and are alsoexpected to continue spectral measure-ments similar to those of SIM.

Physics of Solar Variability

Atomic and Molecular Data Bases

One session of the meeting was devotedto the physics of solar variability —namely, the evolution of the active

regions on the Sun and the radiativetransfer out of the solar atmosphere. Atthe heart of this problem is a complex setof atomic and molecular databases. The

NSF SunRise program, for example, hasdeveloped a physical model of solarvariability using these databasestogether with ground-based solar images

to distinguish magnetic features on thesolar surface.

Robert Kurucz of CfA reviewed his work

spanning the past 35 years computingatomic line lists for all elements (includ-ing ions) up through zinc. He describedthe monumental task of compiling

atomic line data on over 100 million

lines. He now computes the solarirradiance spectrum using these data,but, even after removing the effects ofatmospheric transmission, he finds

deficiencies — perhaps due to missinglines and perhaps due to inaccurateatomic data. Bob continues to refine theline-lists, now including some diatomic

molecules. He is also working on animproved analysis of the Kitt Peak SolarAtlas that was obtained using the NSOFTS developed by Jim Brault.

The SunRise Spectral SynthesisProgram

Peter Fox of NCAR described the presentapproach that he, Dick White, and Juan

Fontenla are using to model irradiancevariations using seven semi-empiricalmodels— taking into account sunspots,plages, enhanced and quiet network, and

the quiet Sun. They compute theintensity as a function of wavelength andposition on the solar disk for each of thestructure types corresponding to theirmodels. These calculations include three

different approximations for the linesource function; one suited for the verystrong resonance lines where partialredistribution (PRD) is important,

another for the most important non-resonance lines, and another approxima-tion for the many narrow lines that areprovided in Kurucz’s listings. The image

analysis and calculations of the irradi-ance variation as a function of time weredescribed. This work provides anunderstanding of the sources of variabil-

ity arising from structures that areapplicable to the SORCE/SIM wave-length range.

The Sun’s Impact on Climate

Bob Cahalan provided an overview ofthe impact of the Sun on the Earth’satmosphere and climate system, and

THE EARTH OBSERVER

26

chaired a session which focused onheating of Earth’s atmosphere andoceans. A common theme emphasized inthis session was the importance of the

spectral measurements of SIM andSOLSTICE —we must know how solarvariations are distributed over ultravio-let, visible, and infrared wavelengths,

since these have distinct influences onEarth’s ozone layer, clouds, and upperlayers of the oceans. Emphasis was alsogiven to understanding both direct and

indirect influences of the Sun on theEarth, which involve feedbacks betweenEarth’s stratosphere, troposphere, andoceans. Cahalan emphasized the role of

all three phases of water on Earth —water vapor being the primary green-house gas in the atmosphere, theimportance of trace gases such as CO2

arising from their absorption in the“water vapor window” at 800 - 1250 cm-1

(12.5 to 8 micrometers). Melting of polarice is one major response to the post-industrial global warming, enhanced

due to “ice-albedo” feedback. Finally,water in liquid form has a majorinfluence due to cloud albedo and oceanabsorption, particularly at visible

wavelengths. A large fraction of solarenergy absorbed by the ocean goes intothe latent heat of vaporization. Thus thesolar heating of the atmosphere-ocean

system is strongly coupled through thewater cycle of evaporation, cloudformation, precipitation, surface runoffand ice formation. Each different climate

component responds to variations indifferent solar spectral bands — atultraviolet, visible and infrared wave-lengths.

Gail Anderson of the Air Force PhilipsLaboratory reported on recent status ofthe Moderate Resolution Transmittance

(MODTRAN) Code that calculatesatmospheric transmittance and radiancefor frequencies from 0 to 50,000 cm-1 (200

nm through the infrared). Except for itsmolecular band model parameterization,MODTRAN adopts all the LOWTRAN 7capabilities, including spherical refrac-

tive geometry, solar and lunar sourcefunctions, and scattering (Rayleigh, Mie,single, and multiple), and defaultprofiles (gases, aerosols, clouds, fogs,

and rain). MODTRAN’s newest 2cm-1

band model employs the spectral lineparameters of the HITRAN96 databasefor each of 13 molecular species, plus

cross sections for the heavy molecules(e.g., CFCs). A major upgrade of thecurrent version is the capability of usersto easily define cloud and rain descrip-

tions. A second major upgrade is theinclusion of molecular band modelparameters based on the HITRAN96spectroscopic database.

The impact and importance of atmo-spheric forcing were discussed, begin-ning at high altitudes with stratosphericprocesses. David Rind of Goddard

Institute for Space Studies (GISS)reported on recent calculations using theGISS Global Climate Middle AtmosphereModel (GCMAM). This research has

studied the impact of UV variations onthe stratosphere and resulting affects onthe troposphere via wave-mean flowinteractions. On long-time scales,

variations in solar activity in combina-tion with altered stratospheric ozonemay provide for decadal andmultidecadal changes in planetary wave

patterns, and hence regional climates.Placing the potential solar influencemechanism in the broader context ofwave/mean flow interactions may help

explain some of the empirically derivedsolar-climate relationships.

Moving to lower levels of our atmo-

sphere, we considered the solar influenceon the troposphere. Peter Pilewskie ofNASA/Ames discussed recent measure-

ments of the downwelling solar irradi-ance at the ground. Spectra obtainedunder cloud-free conditions werecompared with calculations using a

coarse-resolution radiative transfermodel to examine the bias betweenmodel and measurement due to watervapor. Indeed there is a bias strongly

correlated with the amount of watervapor, and the source of the discrepancyremains unknown. In light of thesedifferences and the inherent uncertain-

ties in spectral solar source functionsused by current radiative transfermodels, SIM will provide a significantimprovement. The inherent accuracy of

both TIM and SIM may potentially proveto be valuable reference standards forsurface and airborne sensors. During thecalibration, validation, and comparison

phases of the SORCE program (e.g., atthe JPL Table Mountain Solar Labora-tory) we will conduct instrumentcomparison between TIM and otherTSIM sensors (ACRIM). Similar compari-

sons between SIM and, for example, theNASA Solar Spectral Flux Radiometer(SSFR) and other airborne and surfaceradiometers will also be pursued.

There was discussion of atmosphericphotochemistry and Sasha Madronich ofNCAR considered the influence of solar

ultraviolet (UV) radiation on atmo-spheric chemistry by photo-dissociatingrelatively stable molecules into highlyreactive molecular and atomic frag-

ments. The added reactivity leads to theformation of key atmospheric oxidantssuch as ozone (O3) and hydroxyl radicals(OH), which in turn remove many other

species including methane and non-methane hydrocarbons, hydrochloro-fluorocarbons (HCFCs), and variousoxides of sulfur and nitrogen. Ultimately,

these chemicals affect a number ofenvironmental concerns, e.g., formationof urban photochemical smog, acid rain,


27

and regional gaseous oxidants, thelifetimes of greenhouse gases, and thedepletion of stratospheric ozone. Modelsof atmospheric chemistry require

accurate calculations of the rate coeffi-cients for photolysis reactions, which inturn require accurate knowledge of thesolar spectral irradiance incident at the

top of the atmosphere.

The solar forcing of the ocean systemwas also discussed and George Reid of

NOAA considered global sea-surfacetemperature (SST) data that are nowavailable for the period beginning about1856. The decadal-scale variation of SST

in all three ocean basins, with a range ofabout 0.5°C, shows a remarkablecorrelation with the decadal-scaleenvelope of the solar-activity cycle. The

similarity of the behavior in all of theocean basins suggests an externalforcing, since random stochastic forcingor changes in ocean circulation would beexpected to produce different responses

in different regions. Studies of theresponse of globally averaged SSTs to avarying irradiance have been carried outwith a one-dimensional ocean thermal

model. The analysis uses a reconstruc-tion of irradiance variations based on alow-pass filtered version of the sunspotcycle, and a variation in greenhouse-gas

forcing based on the known variation incarbon dioxide and other greenhousegases during the century. The resultsindicate that about 30-40% of the

cumulative change in global SST sincethe beginning of the 20th century couldhave been caused by solar-irradiancevariability.

There was discussion of how the SORCEobservations will be adopted by theclimate modeling activity. Ray Roble of

NCAR discussed progress in developinga global atmospheric model extendingfrom the ground to the exosphere.

Planetary wave and gravity wavebreaking in the mesosphere has beenshown to affect the stratosphere through“downward control.” There is an

important need for the development of amodel of the entire atmosphere that canbe used to investigate couplings betweenatmosphere regions and solar terrestrial

interactions. Such a model is now underdevelopment coupling the NCARCommunity Climate Model (CCM3) andthe Thermosphere-Ionosphere-Mesos-

phere-Electrodynamics General Circula-tion Model (TIME-GCM).

The final discussions centered on the

longer term influence of the Sun onclimate. Judith Lean of NRL discussed anarray of climate parameters that exhibit“cycles” appearing to match solar

activity records. This empirical evidenceexists on time scales of decades tocenturies. Climate change simulationswith the GISS GCM in response toreconstructed total solar irradiance

during the past four centuries suggestthat global surface temperature varia-tions on the order of 0.4°C occur inresponse to solar forcing on centennial

time scales. Some 65% of the responsearises from the feedback of water vapor,clouds and snow/sea-ice cover. Theclimate change simulations do not,

however, reproduce the apparentstrength of sun-climate relations ondecadal time scales that the empiricalevidence suggests. One possibility is that

climate models lack processes by whichsolar cycle irradiance forcing maybecome phase-locked with decadaloscillations occurring within the climate

system. Another possibility is that themodel simulations thus far have notaccounted for indirect climate forcing bysolar UV-induced changes in ozone.

Dr. Gerald (Jerry) Soffen, former EOSSenior Project Scientist, founder of

The Earth Observer newsletter, andleader of the Viking science teamthat performed the first experimentson the surface of the planet Mars,

passed away November 22 at GeorgeWashington University Hospital inWashington, DC. He was 74.

Soffen joined NASA’s GoddardSpace Flight Center in 1983 where hehelped establish the Mission toPlanet Earth program and served as

the EOS Senior Project Scientist fromJanuary 1988 until September 1990 atwhich time he formed the UniversityPrograms Office at Goddard. For the

past 10 years, he directed activitiesand programs designed to maintainand broaden the Center’s interactionwith the university community.

Dr. Soffen will be sorely missed byhis colleagues and friends through-out NASA and the educationalcommunities.

In Memorium

Attention EOS Researchers: Please sendnotices of recent media coverage inwhich you have been involved to:

Rob Gutro, EOS Project Science OfficeCode 900, NASA/GSFCGreenbelt, MD 20771Tel. 301-286-4044

e-mail: [email protected]

(Continued from page 2)

EOS Scientists In News

THE EARTH OBSERVER

28

From October 24 to 26, 2000, the manag-ers of the Earth Science Enterprise (ESE)Earth Science Data and InformationSystems (ESDIS) Project, and the

Distributed Active Archive Centers(DAACs) met at Goddard Space FlightCenter (GSFC). The DAAC Managersmeet quarterly, alternating between

being hosted by the ESDIS/GSFC DAACand another DAAC (i.e., every othermeeting is hosted by a different DAAC).Meeting at GSFC twice a year affords allDAAC Managers the opportunity to

interface with a larger number ofEnterprise and Project personnel.

The DAAC Managers are faced with

many far reaching technical and pro-grammatic issues and recognize, basedon individual independent experiences,that many of these issues can be jointly

solved. The implicit, yet very real,benefit of the DAAC Manager’s discus-sions is the potential extrapolation ofresolved issues being applied to support-

ing the formulation of future ESE dataand information management systems.

The following topics were addressed

during the three-day meeting:

Informational Sessions:

• Computer security and training

• Budgets/Work Plans

Distributed Active Archive Center(DAAC) Managers Quarterly Meeting— Steve Kempler ([email protected]), Distributed Active Archive Center,

NASA Goddard Space Flight Center

• Guest Speaker - Mike Thomas, CodeY0, NASA HQ

Issues:

• Long Term Archive (LTA)

• Billing and Accounting• EOSDIS Data Gateway (EDG)• DAAC Alliance Issues• EOSDIS Core System (ECS) Issues

Reports:

• Data Center Benchmarking Study• Readiness for Aqua Reviews• User Services Working Group

(USWG)

• Federation• NewDISS

Informational Sessions:

Security — Clayton Sigman, ESDIS

Project, presented security guidelinesand an interpretation to facilitate theDAACs efforts to be compliant withNASA security requirements. Sigman

spoke extensively on various relevantissues, answering questions, and leavingthe DAAC Managers with the following:

• DAACs must comply with NPG2810.1 security requirements. (whichwould ensure compliance with OMBA-130).

• Security documentation: ESDIS has toconcur on all security plans. If aDAAC complies with its NASA localsecurity requirements, it also has to

submit the plans to ESDIS forcertification. The Authorization toProcess form has to be completed bya civil servant related to each DAAC.

• National Agency Checks (NACs) arerequired for all personnel with rootaccess.

It was pointed out that some of theserequirements “go against the grain” ofDAACs located at universities. Universi-

ties often invite and enroll promisingforeign national students to enhance theuniversity’s level of research and studies.Security requirements may inhibit these

opportunities. Obviously, discussionswill continue.

Budget/Work Plans - Vanessa Griffinmentioned that the ESDIS-provided

DAAC budget will remain as planned.However, the extent of the DAACsystem-enhancements budget, which hasbeen a resource for allowing DAACs to

develop and implement EOSDISfunctionality removed from ECSimplementation, is questionable.

Guest Speaker — Mike Thomas,Director of the Applications Commer-cialization and Education (ACE)Division of the Earth Science Enterprise

(Code YO), spoke to the DAAC Manag-ers about new and existing ESE Applica-tions Program initiatives. Thomas’ ideaswere insightful, and he showed much

willingness to learn more about theDAAC’s collective and individualapplications and outreach programs, andhow these programs can better contrib-

ute to the ESE.


29

Issues:

Long Term Archive (LTA) - Implementa-tion scenarios are still being discussed, asthe time to make hard budget decisions

draws nearer. Recognizing that the long-term responsibility for the data falls toNOAA, implementation alternativesinclude: 1) archive the data in place at

the DAACs; 2) transfer data to thechosen NOAA data center sites; and 3)outsource the archives to universities,museums, or commercial sites. The

science community (through DAACUser Working Groups) needs to maketheir feelings known on this issue. It isexpected that some NASA/NOAA LTA

decisions will be made by the nextDAAC quarterly meeting.

Billing and Accounting (B&A) - Bill

Potter, ESDIS Project, is leading thestudy to deploy the B&A system at theDAACs. Due to increasing volumes andlimited resources for data distribution,this may be necessary. The DAACs will

have to charge for data to offset thereduced funding being provided for datadistribution. DAACs will be definingrequirements and contributing opera-

tional concepts to Potter’s workinggroup, as well as developing a B&ABusiness Plan for handling funds. TheB&A working group will work on a

policy definition for review by theDAAC managers.

It was noted that all components of Earth

science will have to recover the costs fordisseminating data, including RESACsand ESIPs. NASA-funded researchers(including ESIPS) will not be charged for

data. If costs for distribution of data toNASA-funded researchers exceed thebudget, Griffin will ask for additionalfunds.

EOSDIS Data Gateway (EDG) - Robin

Pfister and Richard Ullman of the ESDISProject, answered several questionsregarding the use of EDG. There wasdiscussion that, although EDG is being

improved continuously, currently newusers have a hard time learning to useEDG. Tutorials at JPL and George MasonUniversity are planned in the future. The

new EOS Clearing HOuse (ECHO)system will allow alternative userinterfaces to be developed. Ullmanpresented slides on the new guide

system.

DAAC Alliance Issues - The DAACAlliance, composed of the eight DAACs

and the Global Hydrology ResourceCenter (GHRC), have come together toform an Alliance to address interests andissues of common concern. These

include:

• Alliance System Engineering Team(ASET) - in its formulation stage, thisteam of DAAC engineers can address

open source software implementation,B&A implementations, security, etc.

• Alliance Charter - with additional

small corrections, the Alliancecharter, outlining the joint DAACeffort to be premier data managementcenters, is ready to be signed by its

members.

• Alliance Web Page - discussionscontinue on how to develop an

Alliance Web page economically.

ECS Issues - various ECS issues wererecorded. Typical problems at variousDAACs include:

• Troubles with production planning —operations generally run smooth withsteady data flow; however, smallperturbations cause many system

problems.

• Gaps in the data sets require a lot ofmanual cleaning.

• Hardware to accommodate I/O

needs to be enhanced.

Reports:

Data Center Benchmarking Study - Bud

Booth and Ron Holland, from ESDIS,along with others on Vanessa Griffin’sstaff, have toured many data centersnationally and internationally,

benchmarking their best practices andlessons learned. The objective of thisstudy is improve EOSDIS performanceand reduce operation costs. A formal

report is forthcoming.

Readiness for Aqua Reviews - reports onScience Operations Readiness by JohnMoses, ESDIS Project and each AquaDAAC showed that, although we are at

least six months before the Aqua launch,the systems at the DAACs are comingtogether, and issues are being exposedand addressed. The Aqua Operations

Readiness Review will be ready twomonths prior to launch.

Mike Moore, ESDIS Project, providedinsight into additional system capabili-ties needed for operations readiness.

These include:

• a system upgrade at LaRC in order todo science software integration forMOPITT; and

• the Simple Scalable Script-basedScience Processor (S4P) to providemore efficient processing.

User Services Working Group (USWG) -Diana Starr, NSIDC DAAC, new chairperson of the Alliance USWG, raisedseveral topics for discussion:

• the next DAAC yearbook, to be

THE EARTH OBSERVER

30

reviewed by the DAAC Managers, isexpected to be ready by December2000;

• Diana Starr will provide DAACManagers with proposed guidelinesfor the activity report contents; and

• the USWG was asked to providemore guidance on kudos andcomplaints. The group decided toleave it up to each DAAC to provide

its own standard.

Federation - The next Federationmeeting will be January 8-11. It will

feature a science showcase showingwhat the ESIPs have accomplished.Discussion on the DAAC’s viewsregarding increased Federation member-

ship ensued. This will be furtherdiscussed by the Federation’s ExecutiveCommittee.

NewDISS - New Data and Information

Systems and Services activities arestarting to move out, and it is essentialthat the DAACs be involved. It isperceived that DAAC experience is very

valuable. Program personnel discussionsto develop prototypes are underway.

Please contact the DAAC Manager of

your choice or the ESDIS ScienceOperations Manager if you would likefurther information.


The West Antarctic Ice Sheet’s contribu-tion to global sea-level rise may be much

slower today than it was in the past.New evidence indicates that the size ofthe ice sheet thousands of years ago hasbeen overestimated and the ice sheet

may not have been as big or as steady asource of sea-level rise as scientiststhought.

Glaciologist Robert Bindschadler from

NASA’s Goddard Space Flight Center(Greenbelt, Md.) discussed the latestresearch results and changing views ofthe West Antarctic Ice Sheet’s history at

the American Geophysical Union’sannual meeting in San Francisco onDecember 16.

“Our previous best estimates that the icesheet is adding 1 millimeter per year tothe global sea level are almost certainlytoo high,” says Bindschadler. This

revised assessment is based on asynthesis of new data including past sea-level rise estimates presented at aworkshop this fall on the West Antarctic

Ice Sheet organized by Bindschadler.

Calculations of how much and how fastthe ice sheet has thinned and retreated

since the peak of the Earth’s last majorIce Age 20,000 years ago are based inlarge part on a recent reconstruction of

how big the ice sheet was during thatlast glacial maximum. That reconstruc-

tion included a West Antarctic Ice Sheetthree times as large as it is now. Cur-rently, the ice sheet averages 2000 metersthick, covers an area the size of Mexico,

and contains enough water to raiseglobal sea level 5 meters.

But analysis of a 30-by-50-mile rise in theice sheet near the Ross Ice Shelf called

Siple Dome suggests that this featurewas not overrun by a massive ice sheetin the past, which is what the reconstruc-tion suggests. A team of glaciologists

from the University of Washington ledby Charles Raymond used an ice-penetrating radar to study the subsur-face layering of Siple Dome.

Another line of evidence that throws theice sheet’s ancient bulk into question isthe discovery that the ice sheet was still

growing as recently as 8,000 years ago.The reconstruction assumed that the icesheet reached its maximum growth20,000 years ago and has only been in

retreat since then.

According to a new reconstruction ofhistoric sea-level around the world by W.

R. Peltier of the University of Toronto, amajor jump in sea level occurred before

West Antarctic Ice Sheet May Be aSmaller Source of Current Sea-LevelRise— Rob Gutro ([email protected]), Science Systems & Applications Inc.— Harvey Leifert, AGU Press Room— Lynn Chandler ([email protected]), NASA Goddard Space Flight

Center


31

Introduction

The First National Space DevelopmentAgency (NASDA) Workshop on EarthObserving Satellites for Earth ResourcesMonitoring was held in Bangkok on

September 11-12 in conjunction with aCommittee on Earth ObservationSatellites (CEOS) Working Group onInformation Systems and Services

(WGISS) Subgroup Meeting (Sept. 11-13).The meetings were hosted jointly by theNASDA and the National ResearchCouncil of Thailand (NRCT). Over 170

scientists and information technologistsattended the workshops. At these back-to-back meetings, the first early results ofthe partnership that has formed between

the WGISS and an international scienceinitiative, the Global Observation ofForest Cover (GOFC), were demon-strated. The partnership was formed

because of the importance the interna-tional science community placed onexploring new information technologiesthat can improve data services. The

system demonstrations shown at the

workshop were a step towards definingand establishing the role of the WGISSTest Facility (WTF)-GOFC which wasformally endorsed at the subgroupmeeting.

Background

CEOS was created as a forum for allnations that operate Earth observing

satellites to coordinate their activities.WGISS (wgiss.ceos.org), a workinggroup of CEOS, concentrates on thedevelopment of guidelines and ap-

proaches that enable the data processing,archival, and distribution systems of themembers to interoperate and easilyexchange data and information. The

WTF will apply infrastructure-relatedservices and tools developed by theSpace Agencies to projects of relevanceto CEOS. An initial focus for the WTF is

to meet the information and data needsof the GOFC project, which is part of theGlobal Terrestrial Observing System.GOFC (www.gofc.org) is a coordinated

international effort with the goal of

developing and implementing a suite ofsecurely funded, on-going observationalprograms which will provide timely andconsistent information about forests

worldwide, using an appropriatecombination of spaceborne and in situ

data. GOFC is designed to help provideinformation needed to address questions

of global change research, associated, forexample, with the carbon cycle, as wellas information for improved naturalresource management.

NASA’s Terra instruments and Landsat 7provide important sources of data forGOFC. The data management systems,

services, and tools that NASA hasdeveloped through the Earth ScienceInformation Partners (ESIPs), EOSDIS,the Science Investigator-led Processing

(SIPs), and Distributed Active ArchiveCenters provide important contributionsto the WGISS Test Environment.

Demonstration Summaries

The September workshop targetedregional applications of Earth observa-tion satellites in Southeast Asia fornatural resource monitoring with an

emphasis on information technology.One of the main topics of the workshopwas GOFC and the development ofoperational monitoring systems for

forests, land-cover change, and fire usingsatellite data. In addition, the associatedinformation management needs werediscussed. Many of the systems demon-

strated Open GIS Consortium (OGC)-developed capabilities. Open GISConsortium is a standards organizationfor geospatial information systems

composed of government agencies andcommercial vendors from around theworld working together to createstandards which are only approved if

there will be commercial productsimplementing them. A number of new

The Emerging Committee on EarthObservation Satellites (CEOS)Working Group on InformationSystems and Services (WGISS) TestFacility (WTF) — Global Observationof Forest Cover (GOFC)— Yonsook Enloe ([email protected]), SGT Inc.— Chris Justice ([email protected]), University of Virginia— Ken McDonald ([email protected]), NASA Goddard Space Flight

Center— Matt Schwaller (NASA Goddard Space Flight Center)— David Skole (Michigan State University)

THE EARTH OBSERVER

32

data and information systems capabili-ties relevant to GOFC were demon-strated at the workshop.

Tropical Rain Forest InformationCenter (TRFIC)

The Michigan State’s Tropical RainForest Information Center (NASA/ESIP)

provides scientific information on theworld’s tropical forests using Landsatand other high-resolution satellite data.The TRFIC system uses visualization

techniques similar to the OGC Web-based visualization to show deforesta-tion and land-cover changes in theforests of Southeast Asia. The tremen-

dous visualization capabilities shownincluded the capabilities to overlaymultiple data images with maps, tomosaic image tiles, to turn on or off data

footprints, and to zoom and pan. TheTRFIC system has data servers inmultiple distributed locations around theworld. Information about deforestation

and land-cover changes in areas aroundthe world can be accessed through theTRFIC system, www.bsrsi.msu.edu/trfic.

A Distributed Data System forGlobal and Regional Fire Analysis:MOCHA

Francis Lindsay demonstrated how theLos Alamos fires could be analyzed,

using the MOCHA (Middleware basedOn a Code sHipping Architecture)system (NASA/ESIP), developed at theUniversity of Maryland Institute for

Advanced Computer Studies, and multi-resolution data including the highresolution IKONOS and Landsat TM,high resolution 250-m MODIS, and

coarse resolution AVHRR and DMSPdata. Data from these diverse anddistributed sources were integrated anddynamic image processing tools were

utilized for the visualization of burnedareas, enabling post-burn assessments.

MOCHA provides an extensible frame-work to access multiple sites withdiffering DBMS and data formats,performs basic image processing and

data manipulation functions at the sitewhere the data reside, and then bringsonly the desired data to the user. TheJava code to operate these functions is

kept in a code repository and automati-cally shipped to the remote data sites toperform the data manipulation andanalysis requested by users. MOCHA’s

extensible framework enables additionaldata sources and custom data applica-tions to be added to the system quicklyand without extensive code rewriting.

MOCHA development and research isbeing led by Nick Roussopoulos of theUniversity of Maryland, mocha.umiacs.umd.edu.

World Fire Web – EuropeanCommission Joint Research Center(JRC)

The European Commission JointResearch Center’s World Fire Webconsists of AVHRR active fire data

collected by a global network of environ-mental monitoring centers, equippedwith receiving and processing facilities.Each receiving station operates a data

processing chain for detecting fires in thesatellite imagery. Daily, global fire mapsare built up from this regional fireinformation by automatically sharing

data over the Internet. OGC Web-basedvisualization techniques were imple-mented in the World Fire Web. Clive Bestdemonstrated JRC’s Java client that

handled requests to distributed serversthat included servers in Bangkok,Australia, Brazil, and Canada andprovided interactive mapping and

visualization. This system demonstrateda quasi-real-time AVHRR fire map thatcan be accessed over the internet,providing powerful visualization

capabilities showing fires around theworld. By adding new data sources inthe next 12 months, JRC expects to havecomplete global land coverage with the

World Fire Web, opengis.jrc.it

North American and Global FireMonitoring Program – CanadianCenter for Remote Sensing (CCRS)

Brian McLeod, CCRS, demonstrated two

GOFC-focused systems prototyped byusing existing WGISS technologiesprofiled with a specific GOFC theme.The first demonstration presented a

GOFC “view” into the CEOS Interna-tional Directory Network for metadatadiscovery. After consulting with thescience community, pre-defined IDN

searches based on IDN keywords werescripted in the CCRS GOFC portal. Thisallows GOFC users to search and accessmetadata with a specific GOFC focus

without having to learn how to use thesystem or to know which keywords yieldthe desired data. The second demonstra-tion combined existing WGISS inventoryand browse service with emerging OGC

Web-based visualization using Landsat 7data to support validation of burnedareas. Because both prototypes leveragedexisting WGISS technologies, the two

systems were started and completed inthree months, ceodev.ccrs. nrcan.gc.ca/prototypes/gofc/GofcPortal.html.

ESA Experimental Web Mapping

Ivan Petiteville of the European SpaceAgency demonstrated two prototypes. Inthe first demonstration, the original

ATSR Fire data (available ationia.esrin.esa.it) were ingested intoESA’s existing Multi-mission UserInterface System (MUIS) Catalog system.

EOLI, ESA’s newest Java-based userinterface to MUIS, was used to visualizethe Along Track Scanning Radiometer


33

(ATSR) Fire metadata and the ATSRproducts with GIS thematic maps usingOGC Web-based visualization. In thesecond prototype, ATSR fire data were

ingested into a vendor product thatimplements OGC Web-based visualiza-tion, which allowed over 288,000 fires tobe detected. A user could select a

detected fire and access both metadataand data about that fire. In both proto-types, the commercial vendor productswere integrated with existing data and

systems to achieve new capabilities,URL: odisseo.esrin.esa.it/eoli/eoli-gis/html; www.ionicsoft.com/esa_firei/.

GOFC International DirectoryNetwork

Lola Olsen (NASA) demonstrated aGOFC customization of the InternationalDirectory Network (IDN), also known as

the Global Change Master Directory. TheCEOS IDN group has been experiment-ing with providing the capability forspecial interest groups to serve theirfocused sets of metadata. Either IDN

partners or interested focused groupscan customize interfaces, specificallytargeting the nature of the selectedsubsets of the entire content. Descrip-

tions of only those datasets pertinent toforest coverage, forest characteristics,and forest fires were included in theGOFC customization of the IDN. The

IDN team will be working with theGOFC scientists to identify keywordsappropriate for GOFC data collections.

Other WTF-GOFC demonstrations at theworkshop included:

• MODIS Active Fire Products –

University of Virginia / GSFC

• Earth Explorer - U.S. GeologicalSurvey / EROS Data Center

• Space Time Toolkit - University ofAlabama / Huntsville

• Data Information and Access Link

(DIAL) OpenGIS Consortium WWWMapping Demonstration – Raytheon,ITSS

• SPOT Vegetation Data with VEGA2000 Initiative - SPOT Image, CNES

The systems demonstrated showed a

variety of new capabilities that can beused by global change researchers andforestry agencies around the world.Several systems showed excellent data

search and access capabilities. A numberof the prototypes implemented the OGCWWW mapping technologies thatprovided capabilities to overlay image

data from multiple sources with multiplemaps. A variety of other data servicessuch as data subsetting, data mosaicing,data modeling, and data combinationwere also shown. Such distributed,

heterogeneous multi-mission services aresimilar to those being advocated by theNASA NewDISS Strategy for evolvingNASA’s Data and Information Systems

and Services activities.

With the information technologists fromWGISS and scientists associated with

GOFC, a WTF-GOFC integrated teamwill define profiles of WGISS services tosupport thematic/project views, and williteratively prototype new services and

extend current capabilities.


West Antarctic Ice SheetMay Be a Smaller Source ofCurrent Sea-Level Rise

the West Antarctic Ice Sheet began itscurrent retreat, but there is no sign of asubsequent rise large enough toaccount for melting of so much West

Antarctic ice.

The question of how fast the ice sheetretreated still challenges scientists.

Recent work, however, leadsBindschadler to conclude that the icesheet experienced a rapid retreat phasesome 7,000 years ago that was pre-

ceded and followed by a slower retreatthat continues today. Bindschadlerpoints to the geologic record of datedstages in the retreat of the ice sheet’s

continental base as evidence that it hasshrunk in fits and starts. Such episodicretreats may be controlled more by thevarying depth of the underlying

surface and water than by the chang-ing climate.

“The portion of the West Antarctic Ice

Sheet we have focused on for the pastten years appears to be in a stage ofnear-zero retreat now,” saysBindschadler, “but what it will do in

the future is still uncertain.”

“If you extend the new evidence andthe new line of reasoning into the

future, the behavior of the ice sheet ismore difficult to predict. It suggests,however, that if the ice sheet loses itshold on the present shallow bed it is

resting on, the final retreat could bevery rapid.”

THE EARTH OBSERVER

34

Adding iron to the diet of marine plantlife has been shown in shipboardexperiments to boost the amount of

carbon-absorbing phytoplankton incertain parts of the world’s oceans.

A new study promises to give scientists

their first global picture of the extent ofthese unique “iron-limited” oceanregions, an important step in under-standing how the ocean’s biology

controls the flow of carbon between theatmosphere and the ocean.

The study by researchers at NASA’s

Goddard Space Flight Center (Greenbelt,Md.) and the Department of Energy’sOak Ridge National Laboratory, waspresented at the American Geophysical

Union’s annual meeting in San Franciscoon December 15.

Oceanic phytoplankton remove nearly as

much carbon from the atmosphere eachyear as all land-based plants. Identifyingthe location and size of nutrient-limitedareas in the open ocean has challenged

oceanographers for nearly a century.

“We know where the major iron-limitedareas are, but we can’t draw a line

around the precise geographic extent ofthese areas,” says Goddard’s Michael J.

Behrenfeld, a co-author of the new study.“This new result may help us do that.”

The study pinpointed iron-limitedregions by seeing which phytoplankton-rich areas of the world’s oceans werealso areas that received iron from wind-

blown dust. Iron is one of the essentialnutrients needed for microscopic marineplant life to flourish, along with nitro-gen, phosphorus, and silicate. Where

dust from arid regions around the worldfalls into the ocean depends on thelocation of the dust source and therapidly shifting patterns of wind and

weather.

The biologically productive oceanregions were identified by images from

the SeaWiFS (Sea-viewing Wide Field-of-view Sensor) instrument on theOrbView-2 satellite, which maps globalocean biological activity. Because no

similar satellite observations existed fordust-borne iron falling into the ocean,the researchers estimated the location ofoceanic dust deposition with a newly

improved global dust model. The modelcalculates where dust travels and falls byidentifying the location of the majorsources of wind-blown dust around the

world and simulating atmosphericcirculation patterns.

“The estimates of global dust depositionproduced by this model are moreaccurate than previous ones,” saysGoddard scientist Paul Ginoux, “because

we use a new assessment of dust sourcesbased on TOMS (Total Ozone MappingSpectrometer) satellite observations. Themodel also incorporates actual data on

global weather patterns, rather thansimulated circulation, produced byGoddard’s Data Assimilation Office fromsatellite weather observations.”

An annual cycle of average monthly dustdeposition maps was produced usingthree years of simulations up to 1998.

These maps were analyzed for a relation-ship with SeaWiFS 1998 monthly mapsof ocean phytoplankton productivity.The areas where the correlation between

iron deposition and ocean color werefound to be high may indicate iron-limited regions.

To double check their results, the

researchers compared their dust mapswith ocean color maps from the JapaneseADEOS satellite and found that theregions of high correlation between dust

deposition and ocean productivity wereessentially the same.

“Global, satellite-based analyses such as

this gives us insight into where irondeposition may be limiting oceanbiological activity,” says lead authorDavid Erickson of Oak Ridge National

Laboratory’s Computer Science andMathematics Division. “With thisinformation we will be able to infer howthe ocean productivity/iron deposition

relationship might shift in response toclimate change.”

Behrenfeld is leading a series of NASA-

supported research cruises to study in

Global Diagnosis Completed ofOcean Regions Most Sensitive To anIron-Rich Diet— Rob Gutro ([email protected]), Science Systems & Applications Inc.— Harvey Leifert, AGU Press Room— Lynn Chandler ([email protected]), NASA Goddard Space Flight

Center, Greenbelt, Md.— Ron Walli, Oak Ridge National Laboratory, Oak Ridge, Tenn.

(Continued on Page 43)


35

The fourth annual HDF-EOS Workshopwas held September 19 - 21 at

Raytheon’s ECS facility in Landover,MD. Approximately 130 people at-tended. They represented data produc-ers, data users, and tools vendors and

developers, as well as the HDF-EOS andHDF development teams and ESDISproject members.

The purpose of HDF-EOS is to facilitatescientific data storage, exchange, access,analysis and discovery by making EOSstandard data products more easily

accessible. It is both the recommendedstandard format for EOS peer-reviewedscience data products, and a library thatimplements the standard by building on

the HDF library. HDF is a self describing,cross-platform portable data format andfile access library developed by NationalCenter for Suppercomputing Applica-

tions (at the University of Illinois)(NCSA). An HDF-EOS product includesstorage of EOSDIS metadata within thesame HDF file that stores the data. HDF-

EOS also expicitly defines types thataccommodate the most commonorganizations of EOS data: point, swathand grid.

The annual HDF-EOS workshop hasbecome, after four years, the single mostimportant event for reviewing progress

and setting priorities in the developmentof HDF-EOS. The comments and

criticisms received from conferenceparticipants, both speakers and those inthe general audience, inform participants

about the continuing work of developingand maintaining HDF-EOS and HDFlibraries, as well as documentation anduser support activities.

This year’s attendance was the largestever, and conference participation wasvery active and constructive. This year

marked the first workshop after publicrelease of Terra data. All EOS peer-reviewed standard products from Terrause the HDF-EOS format, and several

presentations were devoted to lessonslearned by Terra instrument teams inHDF-EOS use in product definition. But,the most significant and most actively

discussed subject was the transition toHDF5 and especially how the HE5library (HDF-EOS based on HDF5) willbest take advantage of new functionality

in HDF5 while accommodating theexisting use of the HE4 library (HDF-EOS version 2.x based on HDF4). Manyfelt that HE5 has potential to add very

important functionality, but others feltthat HE4 was established as a standardand that it will be difficult to switch or,more likely, to support two different

formats.

Presentations

The HDF-EOS workshop has become an

annual showcase for the status ofinnovations in HDF-EOS, in particular,and HDF, in general. More than 30presentations were made, far too many

to summarize here; the complete list is atthe workshop Web page: hdfeos.gsfc.nasa.gov.

Transition to HDF5

The most important developmentreported in the HDF-EOS library was thenew HE5 library. Larry Klein and David

Wynne presented the implementationtimetable and preliminary design.Discussion of the design was dynamicand valuable. An ad-hoc session was

convened over lunch and many partici-pants expressed satisfaction that thedesign was largely responsive to theneeds of the community. Larry Klein

pledged to shortly make HE5 Beta-2available for trial and feedback. Theofficial HE5 release will be forthcomingin late January. The design of HE5 wasthe central discussion over the course of

the workshop.

Mike Folk, Elaine Pourmal, and RobertE. McGrath of NCSA each presented

aspects of HDF library and toolsdevelopment over the past year. Thesecentered on the HDF4 to HDF5 transi-tion. Elaine Pourmal presented results of

performance studies that show that mostimplementations using HDF5 could be atleast as efficient as their correspondingHDF4 implementation. A major lesson

was that developers must understandaccess patterns to yield the highestefficiency.

Robert McGrath discussed the difficulttechnical issues that surround thetransition from HDF4 to HDF5. Theconclusion was that no general-purpose

conversion will provide good results.Each data producer must consider the

HDF-EOS Workshop IVMeeting Summary— Richard Ullman ([email protected]), ESDIS

Information Architect, NASA Goddard Space FlightCenter

hdfeos.gsfc.nasa.gov

THE EARTH OBSERVER

36

best use of HDF5 for his/her situation.As a corollary, because EOS standardproducts typically contain not only HDF-EOS standard objects (point, swath, or

grid) but also other HDF objects, eachproduct will require unique analysisbefore conversion. This finding furtherset the tone for discussion of HE5.

Ray Milburn reported on some toolsdeveloped by ECS for working withHDF-EOS files. They include“hdfeos4to5,” a command-line conver-

sion utility with capability to convertpoint, swath, and grid; H5EOSView; anHE5 version of EOSView; and JEB is aJava EOS Browse tool.

HDF-EOS Application

The large number and wide variety oftools that use or support the EOS data

format standard is a measure of theincreasing acceptance of the standard. Italso speaks to the importance of EOSdata that conform to the standard.

Many developers, especially thosedeveloping tools for general EOS use,noted that the standard was a usefulstarting place. However, ambiguities in

the standard and enhancements appliedby data producers through addition of“extra” HDF objects made the creation oftools more difficult.

Speakers reported substantial progress intools that use or support the standard.These included: View HDF, the HEW

subsetting appliance, HDF Explorer,MATLAB, commercial HDF toolsthrough RSI, the HDF-EOS DataBlade,WebWinds, the DODS server, The Polar

HDF-EOS Data Imaging and Subsetting(PHDIS) tool, a Validation Toolkit forAtmospheric Profile Data Stored in theAura HDF-EOS Swath Format, the

MODIS Map Re-Projection tool, theDistributed Interactive Computing

Environment (DICE), Data ProcessingError Analysis System (DPEAS), and theDistributed Image SpreadSheet (DISS).Further details are on the workshop Web

site.

Also presented were efforts to extend thestandard. R. Suresh outlined a proposal

for an EOS vector data standard. Theformat, targeted for the GIS communityand requested by some Earth ScienceInformation Partners (ESIPs), would be

an independent extension to HDF-EOS.As such, it would be an “add-in” to theHDF and HDF-EOS libraries. The librarydesign uses HDF4. Liping Di reported

progress in the use of HDF-EOS in theOpen GIS Consortium Web mapping testbed.

Reports from Data Providers

EOS data providers spoke about theirexperiences in developing productformats. The experiences relatedindicated that use of HDF-EOS required

careful analysis and engineering of thedata product requirements beforeimplementation.

Concerns included how to maintainaccess to heritage data, data productsize, layout of data to facilitate efficientaccess, access to meaningful metadata,

and standardization across data productsby a given producer and among EOSdata producers. A serious concerncontinued to be support for the standard

in visualization and analysis tools.

The consensus among data producerswas that adoption of HDF-EOS was not

simple. However, each presenter notedbenefits of the standard. Most impor-tantly, each noted the potential forsimpler user interfaces and application

of more standard tools. Significantly,several presenters indicated that the

standard accelerated data verificationand analysis activities.

Michael Reid reported on the use of HDF

within the Landsat 7 Processing System(LPS). LPS products are not hybrid: theyare either pure HDF or HDF-EOS. TheHDF-EOS products contain geolocated

swaths and ancillary data and use theLandsat-specific Index Map feature ofHDF-EOS. Data are distributed to usersin native HDF because more users are

familiar with that format.

Mike Bull and Kathleen Crean reportedon MISR Project use of HDF and HDF-

EOS. HDF-EOS swath and grid formatsare the basis of the products, withadditional native HDF structuresincorporated where needed.

Peter Spence gave a concise presentationon experiences in developing CERESdata products. He warned that HDF andHDF-EOS implementation must be

linked closely to development of productgeneration software in order to facilitatethe best design. The use of the standardwas critical in expediting data verification.

Robert Wolfe’s presentation was aboutexperiences the MODIS Land Group hashad with HDF-EOS. He asserted that

HDF-EOS has made collaborationseasier, that the self-documenting natureof the format allows easier exchange ofinformation, and that an essential feature

is the standard method for representinggeolocation. Subsetting is especiallyimportant because of product size. Hediscussed several shortcomings but he

concluded that more standards likeHDF-EOS would be helpful.

Richard Hucek told how the MODIS

Atmosphere Group also generatesproducts in HDF-EOS format. TheAtmosphere software opens and creates


37

the MODIS product files using the HDF-EOS interface routines in the ESC SDPToolkit. The HDF-EOS interface allowssubsetting of planes within 3-D data

arrays, such as height above the Earth’ssurface.

Cheryl Craig detailed HDF-EOS Aura-

File-Format Proposed Guidelines. Shenoted that, while HDF-EOS constrainsHDF with point, swath, and gridimplementations, it is still possible to

create two HDF-EOS files that arecompletely different and would requiredramatically different readers. To easecross-platform use of Aura data sets, the

Aura teams have agreed to make theirfiles match more closely. Areas ofagreement include:

• organization of data fields andattributes;

• dimension names;

• geolocation names and dimensionordering;

• data field names and dimension

ordering;

• units for data fields; and

• attribute names, values, and units.

The audience enthusiastically supportedthe ideas presented. Several participants

suggested that the standardization effortfor Aura should extend to all new dataproducts. However, after discussion, theconsensus was that it is more practical

and achievable for this detailed level ofstandardization to apply to a singlemission rather than broadly to all HDF-EOS products.

Doug Moore and Huan Meng discussedtheir experiences producing operational

real-time products in HDF-EOS in theMicrowave Surface and PrecipitationProducts (MSPPS) from AMSU-A andAMSU-B data. The team had trouble

early in implementation, and still hassome problems, but the advantage ofHDF-EOS was wider portability and theself-describing format. Tools, such as

IDL, view_hdf, and WebWinds thatrecognize the HDF-EOS standardnatively read the products. HDF-EOS toBUFR and GRIB to HDF-EOS conver-

sions are implemented in the MSPSSsystem.

Experts’ Panel Questions andAnswers

The workshop design encouragesparticipation between speakers present-ing new progress and the generalaudience. Time was also set aside to

explicitly facilitate exchange in the formof the “Ask the Experts Panel.” Theformat encouraged discussion amongthe audience and the “experts panel.”Some of the notable discussion topics

follows:

Q: In application, EOS Standardproducts are “hybrids” in the sense

that, while all contain HDF-EOSobjects, most (some participantsasserted ALL) contain other addi-tional HDF objects. Has the HDF-EOS

standard proved to be useful, orgiven its variation in application, isthe more general HDF standard moreappropriate?

A: After discussion, it was generally agreed

that NASA’s use of HDF and HDF-EOS

has been advantageous, although not

perfect. The response from the experts

included that HDF-EOS was developed

to standardize geolocation data, not all

data. Some non-conformance issues are

addressed in HE5, but a certain amount

of variation will have to persist due to the

wide range of sciences.

Q: Will a single HDF-EOS library be

produced that handles both HE4 andHE5?

A: It is feasible but, because such a library

will require use of both libraries, design

can’t be started until after the release of

the HE5 library. This was an extremely

important discussion, which followed

from the previous day’s discussion of the

design of HE5.

Q: Will Terra data be ported to HE5?

A: There is no EOSDIS policy that requires

that data products defined in HDF-EOS

be ported to the new HE5 standard.

ESDIS’ commitment to the HDF-EOS

standard includes continuing support for

HE4. Each data producer and science

team will need to consider transition to

the new standard on a case by case basis.

There will need to be a compelling reason,

and timing would likely be coincident

with scientific reprocessing.

Q: There is a perceived need to include aversioning number in HDF-EOS filesthat would enable users or applica-tions to identify: (1) the HDF-EOS

version used to create the file, and (2)whether the file is “pure HDF-EOS”or a hybrid.

A: HDF-EOS developers will consider

adding this functionality.

Q: Some discussion was centered on a

hypothetical case of a grad student ata small university. How could such aperson sort out all of the differentHDF’s and figure out how to deal

with them?

A: There was no conclusive answer. It was

THE EARTH OBSERVER

38

agreed that documentation, especially the

HDF-EOS primer, needs to be updated.

Other ideas given were: concentrate

more on HDF-EOS; make a list of all of

the standard products that are coming,

and list all of the available tools that will

help people get started; most don’t really

need to know about the HDF-EOS/HDF,

just look at the data. Ask the DAAC

first, then the science team and HDF-

EOS support team; provide a user’s guide

that helps people first determine at what

level they need to come in; and provide a

tutorial.

Q: Will HDF-EOS be available forLinux?

A: Linux ports are planned. Expect them

before spring.

Q: Does HDF5 support parallel I/O?

A: HDF5 supports MPI-IO. It does not

support applications running on separate

machines and accessing the same file at

the same time. But MPI-connect might

make this possible.

Demonstrations and Tutorial

The HDF-EOS Workshop provides anopportunity for vendors and developersto lead hands-on demonstrations and

even tutorials on the use of their tools.The third day of the workshop wasdevoted to this topic and the interest wasextremely strong. All of the demonstra-

tion sessions were filled to capacity. Afterthe sessions became over-filled on thefirst day, the schedule was expanded toadd new tutoring sessions and those

spaces were also filled. Each of thefollowing tutorials and demonstrationswas well received.

Barbara Jones, Elena Pourmal, Mike FolkNCSA: HDF5 Hands-on Tutorial: Anintroductory tutorial to HDF5, coveringhow to use the HDF5 application

programming interface (API) to read andwrite simple HDF5 files, and NCSA-supported tools and utilities for workingwith HDF5 files, was taught. Two

sessions were filled.

Robert P. Comer, The MathWorks, Inc.:The MATLAB hands-on demonstration

and tutorial covered relevant capabilitiesin the MATLAB/Image ProcessingToolbox used example data from Terra.

Bill Okubo. Research Systems, Inc.: IDLtools hands-on demonstration andtutorial.

Richard Chinman, UCAR-IITA: DODSservers and clients operating in anumber of data systems were presentedin a real-time network demonstration.

Pedro Vicente, Space Research Software:A hands-on session using HDF ExplorerVer 1.1, a Windows only HDF reader andvisualizer, was given. It was the only tool

demonstrated to support both HDF5 andHDF-EOS.

Linda Hunt and Kam-Pui Lee, NASA

Langley Atmospheric Sciences DataCenter: View_hdf, developed for theCERES project, was shown as generallyuseful for analysis and plotting of HDF

and HDF-EOS formatted data.

R. Suresh, SSAI: The DistributedInteractive Access Link (DIAL), a WWW-

based data distribution system demon-stration showed how data providers caneasily serve their Earth science datadirectly to their users.

Mark Nestler, SSAI: A short tutorial onthe use of the EOS Data Gateway (EDG),the common WWW interface to EOSDIS

data located at all DAACs, was given.

Postscript

The Aura Data Systems Working Groupratified HDF5 and HE5 as the standard

for the Aura mission on November 1,2000.

The HE5 Beta 2 library was released byECS for public trial and comment onNovember 30, 2000.


Terra Science WorkingGroup (SWAMP) Meeting

products are early release data productsthat can be used to gain familiarity with

data formats and parameters, and areintended as a test bed to discover andcorrect errors in the data products. Thedata product has been minimally

validated and may still contain signifi-cant errors. The data product will onlyexist temporarily in the archive, and isnot appropriate for scientific publication.

Beta versions of Terra data products arenow available for order from the ASTER,CERES, MISR, and MODIS sensors.

MOPITT data will be made availablesoon. Calibration and validation

activities are still underway, but the dataquality from all of Terra’s sensorsappears to be exceptional. Subsequentreprocessing of the data will yield

enhanced scientific utility. Beta datashould only be used to become familiarwith the data products and their formats,and to prepare for the fully-assessed and

operational Terra data products. CERESis now releasing Science Validated DataProducts. See eosdatainfo.gsfc.nasa.gov/ for current information on Terra

data products

The next meeting will be a mini-SWAMP

that will be held during the IWGmeeting in Ft. Lauderdale, FL, January29, 2001.


39

NASA, FEMA Partner to Use Scienceand Space Technology for DisasterPrevention— David E. Steitz ([email protected]), NASA Headquarters, Washington, DC— Michael Widomski, Federal Emergency Management Agency, Washington, DC.

(Phone: 202/646-4337)

NASA and the Federal EmergencyManagement Agency (FEMA) has signeda Memorandum of Understanding

(MOU) and joined in a partnership on amajor natural-disaster initiative.

It is affiliated with Project Impact:

Building Disaster Resistant Communi-ties. The cooperative agreement willresult in updated and more accuratemaps of flood plains, a better under-

standing of wildfires, and maps toimprove disaster recovery and mitiga-tion by state and local communitiesthroughout the United States.

Under the new partnership arrangementsigned by NASA Administrator Daniel S.Goldin and FEMA Director James Lee

Witt, NASA and FEMA will applyscience, technology and remote-sensingresearch images of the Earth taken bysatellites to emergency management

issues on the ground, such as mappingof flood plains and earthquake fault linesand observation of wildfires and othernatural hazards.

“This new partnership between NASAand FEMA demonstrates the diverse andwide-ranging applications of NASA’s

Earth science research and technologyand its benefit to the American people,”said Dr. Ghassem Asrar, Associate

Administrator for Earth Science, NASAHeadquarters, Washington, DC. “TheOffice of Earth Science is eager to form

new partnerships with other governmentagencies, such as FEMA, as well as withindustry and public groups to expandAmerica’s use of our Earth-science data.”

“I am extremely happy to have NASA asa Project Impact partner,” said FEMADirector James L. Witt. “Using the

technologies by NASA for disasterprevention will help in saving lives andmake communities all across Americadisaster resistant.”

The agreement outlines a first coopera-tive effort to map flood plains inCalifornia’s Los Angeles basin; around

Sacramento, CA; Virginia Beach, VA; theRed River along the North Dakota andMinnesota borders; and San Francisco,CA. Using laser-imaging and radar-

mapping data, NASA and FEMA areevaluating technology for creating moreaccurate maps of these areas that willhelp state and local officials model and

understand drainage and run-off, whichare vital to their disaster preparedness.Local communities will benefit fromthese precise maps by better understand-

ing the physical characteristics of theircommunities.

At the same time, NASA Earth scientistswill gain valuable data for technologydevelopment, validation and calibrationof satellites, and the understanding of

land use, land cover and flood hazards.America’s flood-insurance industry alsowill benefit from the accuracy of thesenew maps, which will provide more

precise views of flood-threatened areas.

As the agreement is implemented, NASAresearchers and their FEMA colleagues

will use a variety of public and privatesatellites and aircraft-mounted Earth-observing instruments. These efforts willhelp in understanding issues such as soil

permeability and saturation, which affecthow much water during a flood wouldlikely be absorbed, as opposed toremaining above the ground and

possibly causing damage to crops,houses and communities.

Satellite imagery also can provide stateand local officials with maps of vegeta-

tion in areas prone to wildfires. Thisinformation can be used by firefightersto determine which types of plants aremore likely to fuel wildfires and better

predict what paths such fires may take.

Using airplanes and spacecraft thatobserve characteristics of the Earth

invisible to the naked eye, researcherscan better see characteristics of theEarth’s surface that are changing and canindicate where earthquake fault lines or

volcanoes may be expanding, vital datafor understanding and preparing forthese dangerous phenomena.

The partnership between the spaceprogram and FEMA is part of NASA’sEarth Science Enterprise, a coordinatedresearch program that studies the Earth’s

land, oceans, ice, atmosphere and life as


THE EARTH OBSERVER

40

The User Working Group (UWG) for the

NSIDC DAAC (PoDAG) met at theNational Ice Center in Suitland, Mary-land, in late September. This brief reportcovers the highlights of that meeting.

For more information and more detailplease consult the PoDAG Web pages atthe NSIDC website (nsidc.org/NASA/PODAG/MEETING17/)

In addition to the NSIDC DAAC-specific

topics reported below, the UWG wastreated to an overview of the NationalIce Center history, products and opera-tions, a tour of the NIC facility, and

overview of the science program andlong range plans. NIC staff membersalso presented information on theircurrent work and plans for sea-ice data

assimilation, seasonal sea-ice forecasting,assessment of the Polar Ice PredictionSystem, and the Environmental WorkingGroup Arctic Atlases for oceanography,

meteorology, and sea ice.

Data Charging

The NSIDC DAAC manager briefed theUWG on the current NASA ESE plans

for charging for data and, more precisely,the recovery of costs for distribution ofdata to the media. Data charging wasviewed as uneconomic and undesirable.

It was noted that NASA would cover the

National Snow and Ice Data Center(NSIDC) User Working Group MeetingSeptember 25-27, 2000— Ron Weaver ([email protected]), NSIDC DAAC Manager—David Bromwich ([email protected] ), PoDAG Chairman

costs for NASA projects. Greg HunoltAlso noted that the charges would beless than comparable charges from, for

example, the National Climatic DataCenter.

Science Working Group for theAM-1 (now Terra) Platform(SWAMP) Working Group on Data

A key item of discussion was theproposal by the SWAMP Working Groupon Data for a huge increase in resources(~$15M) for EOS science data processing

in an environment of very tight funding.The danger is that these costs, if enacted,might starve other existing activities.

Allocation of EOS versus non-EOSNSIDC DAAC resources

The UWG had asked NSIDC to present ahigh-level budget assessment, withspecific information on the balance of

resources expended for the various sub-components of the DAAC (EOS Satel-lites, Version 0/Heritage Data, Infra-structure etc.) A key concern was that

30-40% of staff effort would now bedevoted to EOS data support, potentiallyjeopardizing essential data sets fromother sources such as heritage data. The

question was raised as to whether theECS, which comprises some 60% of the

EOS data support, could be scaled backto allow more resources to be allocatedto existing product support.

Data Set Status Report

NSIDC Operations Manager, MarkParsons, provided information onrecently completed and released data

sets, an assessment of upcoming data setreleases, and longer term priorities forthe DAAC. (See presentation on Website for details.) PoDAG agreed to

review the NSIDC data set calendar andcomment on the priorities assigned tothe current and planned DAAC data sets

User Services Status Report

NSIDC User Services Manager, MichelleHolm, presented a summary of useraccess to NSIDC DAAC data. Points

emphasized were: NSIDC focus is onscience users; primary mode of distribu-tion is CD-ROM, but the expectation isthat FTP and tape will play an increasingrole with the advent of EOS-MODIS

products; continued strong interest in“Heritage” data is expected; percentageof DAAC requests in relation to NSIDCas a whole has remained relatively

constant; and DAAC data supportrequests represent 67%-75% of all datasupport requests handled by the NSIDCUser Services staff.

PoDAG in the EOS Era RoundtableDiscussion

The UWG and NSIDC invited MichaelKing and Skip Reber for a discussion on

how they thought the UWG should fitinto the EOS program. The followingsummarizes this discussion:

There was general agreement on theneed to revise the EOS Science Linkageschart prepared by Vanessa Griffin,


41

particularly as it does not show the roleof the User Working Groups (see ActionItems). DAAC UWGs come under thedata side of EOS and since King is on the

science side of EOS, he is not directlylinked to the UWGs or completely awareof their activities and membership.UWGs most appropriately link to HQ

Program Scientists, as these managershave purview over both EOS and non-EOS data activities. UWG recommenda-tions and observations should pass

through the DAAC manager via theProgram Scientist at NASA Headquar-ters, to ESDIS, and through Skip Reber tothe EOS Project Science Office under

Michael King.

Heritage data in the DAACs may betargeted if budget cuts are enacted. EOS

data activities are mandated, andincreases in resources at the DAACs aregoing into EOS data handling. SkipReber needs to have articulated the valueof ancillary data sets for EOS data

analysis.

Current management is not interested inlong-term data management, and there

are no funds to support it. Long termmeans launch plus five years, approxi-mately.

The question was raised as to whatauthority does PoDAG have for rejectingan EOS data set from the NSIDC DAAC.The data set generation has been

reviewed (e.g., the ATBDs), but theactual data have not been independentlyvalidated. Any problems that the UWGfinds with products should be communi-

cated to Skip Reber.

PARCA Data Archiving

The Program in Arctic Regional Climate

Assessment (PARCA) is winding downand the disposition of data collected by

the PIs was discussed. Higher level datasets will be available on PI Web sites byMarch 1, 2001. A data catalog will beavailable on the NSIDC DAAC Web site.

Lower level and higher volume data aremore of a problem. Requests for lowerlevel data will determine what actionsare taken for archiving of these data.

NSIDC will provide an estimate for thearchiving of the higher level data sets.This estimate will be completed by theend of 2001. The HQ Program Manager

for PARCA will provide guidelines to PIsfor depositing data at the NSIDC DAAC.

Earth Science Enterprise (ESE)Outreach

Ming-Ying Wei (NASA HQ Code YO)presented an overview of the ESEeducation and outreach program. Theprogram is divided into education

(formal and informal), outreach, andapplications programs. The OutreachDivision is proposed to reach a target ofabout 3% of the total ESE budget within

a few years.

CryoSat Program

Keith Raney, Applied Physics Laboratoryat Johns Hopkins University, presentedan overview of the proposed CryoSat

effort. CryoSat is a European near-polarorbiting altimeter mission scheduled forthe 2003-06 timeframe. It appears tohave great utility for the sea-ice and ice-

sheet communities. The European SpaceAgency is planning three Announce-ments of Opportunity; calibration/validation, processing, and data utiliza-

tion in the coming three years.

U.S. scientists may propose under no-exchange-of-funds agreements to theseAOs. Such proposals gain access to the

data, but financial support must comefrom U.S. sources.

MODIS Snow and Ice Products

Dorothy Hall, PoDAG member andMODIS snow and ice product leadscientist, presented early results of the

MODIS snow and ice data products fromTerra. Following are the main pointscovered, and associated observations ofthe PoDAG members.

The MODIS cloud-cover algorithm nowdoes a very good job separating cloudfrom snow/sea ice in many of the areas

that have been studied. Initial compari-sons of MODIS snow cover with otherdata sources for New Hampshire duringMarch 2000 shows both agreement and

disagreement, with the latter depending,in part, on the spatial resolution contrastof the data sets. Because none of theoperational products are fully validated,

it is difficult to know which snow map isthe most accurate. MODIS snow coverenhancements to come are fractionalsnow cover and daily snow albedo.

Hall reported that one early user saidthat there is a steep learning curverequired for use of the data. Data requestprocedures need to be enhanced.

The Sea-ice product is determined bytwo different methods, one based onreflectance, and the other based on ice

surface temperature. Cloud coverage isnoted and the land is masked out. Analgorithm for snow over sea ice does notexist. This might be feasible to do, but

will not be done under the existingMODIS contract at Goddard. MODISsea ice was compared to ship observa-tions from the Southern Ocean in March

2000 by Shusun Li of the University ofAlaska, and good agreement wasobtained. The MODIS sea-ice edgecoincided with the 1.7C isotherm

observed by the ship-based observers.

THE EARTH OBSERVER

42

MODIS does have a surface temperaturealgorithm specifically for ice sheets, butZengming Wan of UCSB has produced asurface-temperature product that

includes the ice sheets. A climatemodelers’ snow cover grid will beproduced. Snow cover is not detected atnight, but sea ice is (from the surface

temperature). Snow cover compositeshave been, and will continue to be,produced to try to limit the cloudobscuration. AMSR and MODIS prod-

ucts will be combined to produce a snowcover extent and depth analysis that isnot obscured by persistent cloud cover.

For snow and ice product examples,consult the NSIDC MODIS Web page atnsidc.org/NASA/MODIS/ and followthe links to browse images or example

products.

New PoDAG Members

With regret, PoDAG and NSIDCaccepted the resignations of three

members. Greg Flato, Drew Rothrock,and Mark Drinkwater have steppeddown. They were all thanked for theirefforts on behalf of the NSIDC DAAC, in

particular Rothrock and Flato who arelong-time members. PoDAG discussedreplacement members, but no decisionwas made at the meeting.

Summary of Recommendationsand Actions Taken

Recommendation:

PoDAG recommends to Skip Reber,

ESDIS Project Scientist, that he organizea coordination meeting with DAACmanagers and chairs of User WorkingGroups. The agenda should include data

set charging, heritage data in the EOSera, the SWAMP Working Group on DataFunding Proposal, and long-term dataarchiving.

Action Items:

A full list of action items may be found atthe PoDAG Web site mentioned above.Listed below are action items of interest

to the wider EOS community:

1. Dave Bromwich and Ron Weaver willredraw the EOS Science Linkages

chart prepared by Vanessa Griffin,and will circulate the revised chart toGriffin and the other User WorkingGroups.

2. Dorothy Hall should prepare a morecomprehensive overview of MODISsnow and ice products for the next

PoDAG meeting. This should includea survey of users. 1-2 users should beinvited to the next PoDAG meeting.PoDAG members should use the

MODIS data before the next meetingso that they can provide meaningfuldiscussion.

3. Mark Parsons will evaluate the NISE

sea-ice product in relation to higherquality sea-ice analyses and report toPoDAG at the next meeting.

4. PoDAG members will review theDAAC data set priority list on thePoDAG website by December 4, 2000and provide feedback to NSIDC

DAAC.

5. Dave Bromwich will write a letter toSkip Reber stressing the value of

heritage data to provide a context forEOS data sets.

6. Dave Bromwich and Ron Weaver will

explore the need for a coordinationmeeting between DAAC managersand the chairs of User WorkingGroups.

7. NSIDC will estimate the resources

required to hold all of the PARCAhigher level data that are listed onWeb pages maintained by PrincipalInvestigators (after March 1, 2001).

8. The Program Scientist will sendinstructions to Principal Investigatorsfor depositing data at NSIDC DAAC.

9. Ron Weaver, Roger Barry, and DaveBromwich will recruit potential newPoDAG members discussed at the

meeting.

The next PoDAG meeting will be atNSIDC in Boulder, probably on April 25-

27, 2001. Agenda items proposedinclude:

• MODIS Snow and Ice Products: user

survey, input from users, exampleproducts to be presented. DorothyHall to take the lead here.

• Further discussion of the EOS Science

Linkages chart prepared by VanessaGriffin.

• Cost assessment from NSIDC for

holding higher-level data productsfrom PARCA project.

• Data set priorities.

• Report of SWAMP Working Group onData.

• Data charging.

• New member introductions.

• Selection of a co-chair for PoDAG toeventually succeed D. Bromwich aschair.


43

a total system. This initiative is part of anaggressive new strategy devoted tosignificantly increasing the application ofNASA remote sensing data, information,

science and technologies to societalneeds, ensuring maximum return ontaxpayer investments.

NASA, FEMA Partner toUse Science And SpaceTechnology for DisasterPrevention


February 8Aqua Science Working Group Meeting, NASAGoddard Space Flight Center, Greenbelt, MD,building 33, room H114. Contact: ClaireParkinson, tel. (301) 614-5715; e-mail:[email protected].

February 14-15ICESat/GLAS Science Team Meeting, UCSD,La Jolla, CA. Contact Bob Schutz, tel. (512)471-4267; e-mail: [email protected].

February 15-16SEDAC User Working Group, Arlington, VA.Contact: Bob Chen, tel. (845) 365-8952;e-mail: [email protected].

February 27-March 2AVIRIS Earth Science and ApplicationsWorkshop, Jet Propulsion Laboratory.Contact Robert O. Green, e-mail:[email protected];URL: makalu.jpl.nasa.gov.

April 25-27User Working Group (UWG) for the NSIDCDAAC (PoDAG) Meeting, Boulder, CO.Contact Ron Weaver, e-mail:[email protected].

April 2-4International Conference on InformationTechnology: Coding and Computing(ITCC2001), Las Vegas, Nevada. Call forPapers. Contact Pradip K Srimani, e-mail:[email protected]; URL:www.cs.clemson.edu/~srimani/itcc2001/cfp.html.

April 3-5Oceanology International 2001, Miami BeachConvention Center, FL. Contact for USA:(703) 312-91214; for Worldwide: +44 208949 9822.

April 8-11GWXII: The XIIth Global WarmingInternational Conference & Expo, 2001Annual Conference: KYOTO ComplianceReview, Cambridge University, UK. Call forPapers. For abstract submission see URL:www.GlobalWarming.Net; tel. (630) 910-1551; Fax: (630) 910-1561; e-mail:[email protected].

April 23-27ASPRS: The Imaging and GeospatialInformation Society, St. Louis. See URL:www.asprs.org; e-mail: [email protected]; tel.(410) 208-4855; Fax: (410) 641-8341.

May 14-18Environmental Risks & the GlobalCommunity, Argonne, IL. Contact JoanBrunsvold, tel. (630) 252-5585;e-mail [email protected].

July 9-13International Geoscience and RemoteSensing Symposium, Sydney, Australia. Callfor Papers. For abstract submission seeURL: www.IGARSS2001.org;tel. 61.2.6257.3299; Fax: 61.2.6257.3256;e-mail: [email protected].

July 10-13A Global Change Open Science Conference,“Challenges of a Changing Earth,”Amsterdam. Contact Conference secretariat,tel. +31 20 5040200; e-mail:[email protected]; URL:www.sciconf.igbp.kva.se.

July 10-18International Association of Meteorology andAtmospheric Sciences, Innsbruck, Austria.Contact Conference secretariat, IAMAS 2001,tel. +43 512 575600; Fax: +43 512 575607;e-mail: [email protected];URL: meteo.uibk.ac.at/IAMAS2001/.

September 17-20Fifth International Airborne Remote SensingConference & Exhibition, San Francisco.Contact Marilyn Dehring, tel. (734) 994-1200, ext. 3350; e-mail: [email protected]; URL: www.erim-int.com/CONF/IARSC.html.

EOS Science Calendar

Global Science Calendar


Global DiagnosisCompleted of OceanRegions Mosts SensitiveTo an Iron-Rich Diet

more detail iron-limited areas of the

ocean. The first cruise was completedthis September in the equatorial PacificOcean near Hawaii.

Images to support this release can befound at: ftp://www.gsfc.nasa.gov/earthpix/agu/.

October 7-102001 International Conference on ImageProcessing, Thessaloniki, Greece. Call forPapers. Contact Diastasi, tel. +30 31 938203; Fax +30 31 909 269; [email protected].

Code 900National Aeronautics andSpace Administration

Goddard Space Flight CenterGreenbelt, Maryland 20771

Official BusinessPenalty For Private Use, $300.00

Printed on Recycled Paper

Bulk Rate MailPostage and Fees PaidNational Aeronautics andSpace AdministrationPermit G27

The Earth Observer

The Earth Observer is published by the EOS Project Science Office, Code 900, NASA Goddard Space Flight Center,Greenbelt, Maryland 20771, telephone (301) 614-5559, FAX (301) 614-6530, and is available on the World Wide Web ateospso.gsfc.nasa.gov/or by writing to the above address. Articles, contributions to the meeting calendar, and sugges-tions are welcomed. Contributions to the Global Change meeting calendar should contain location, person to contact,telephone number, and e-mail address. To subscribe to The Earth Observer, or to change your mailing address, pleasecall Dave Olsen at (240) 232-1097, send message to [email protected], or write to the address above.

The Earth Observer Staff:

Executive Editor: Charlotte Griner ([email protected])Technical Editors: Bill Bandeen ([email protected])

Jim Closs ([email protected])Robin Williams ([email protected])

Design and Production: Winnie Humberson ([email protected])Distribution: Hannelore Parrish ([email protected])

o b s e r vings a r t e m e the earth observer

Documents