unclassified ad number · ed befense ... parabola designed for optimum use with the perkin-elmer...

UNCLASSIFIED

AD NUMBER

AD240188

NEW LIMITATION CHANGE

TOApproved for public release, distributionunlimited

FROMDistribution authorized to U. S. Gov't.agencies and their contractors; SpecificAuthority; 8 Jun 60. Other requests shallbe referred to Naval Research Lab.,Washington, DC.

AUTHORITY

Naval Research Laboratory, TechnicalLibrary, Research Reports Section Notice,dtd October 24, 2000.

THIS PAGE IS UNCLASSIFIED

UNCLASS lE]! ED

BEFENSE DOCUMENTATION CENTERFOR

SCIENTIFIC AND TECHNICAL INFORMATION

.•MEBNIN STATION ALEXANDRIA. VIRGINIA

UNCIA ASSIZFIED

j / 1S/ /

K A.ftL Report 5453

aAINFRARED TIRANSMISSION OF THE ATMOSPHERE

H. W. Yates and J. H. Taylor

Radiometry U BranchOptics Division

June 8, 1960

ASI IA•JUL 2f' 196(Y

N'W TIPOR S

U. S. NAVAL RESEARCH LABORATORYWashington, D.C.

NOTE

Enlarged cropies of the c u r v e s shown in Figs. 5-11

will be furnished to qualified persons upon request.

Address inquiries to Director, U.S. Naval Resea.rch

Laboratory (Code 7360), Washington 25, D.C.

CONTENTS

Abstract it

Problem Status it

Authorization t

INTRODUCTION 1

INSTRUMENTATION I

EXPERIMENTAL PROCEDURE 6

RESULTS 6

ATTENUATION BY SCATTERING 7

Introduction 7

Experimental Results 8

SELECTIVE ABSORPTION 35

Introduction 35

Atmospheric Windows 36

Experimental Results 38

ACKNOWLEDGMENTS 44

REFERENCES 45

ABSTRACT

The transmission of the atmosphere in the spec-tral range from 0.5 to 15 p has been studied over threesea-level paths 0.3, 5.5, and 16.25 km in length, withan average spectral resolution, A/AX, of 300. The con-ditions encountered covered a wide range, from 0.11cm precipitable water in the 0.3-km path to a maxt-

I G.M -- -mumtL n-. of 34cmn p-r eciptLab)l e water if Liu the& L.•.)-NJ INL pCaLh,and values of the visual scattering coefficient weremeasured from 0.048 km- 1 to 9.24 km- 1 . Data havebeet, evaluated on selective transmission vs water-vapor concentration for the infrared "window" regions,and the behavior of the scattering coefficient withwavelength has been plotted for representative atmos-pheres. Similar data were plotted for several meas-uremeni ,-., - over a range of 27.7 km at an altitudeof 10,000 F.

PROBLEM STATUS

This is a final report on Problem N03-05, whichwas terminated Sept. 30, 1959.

AUTHORIZATION

NRL Problem N03-05Project NO 501/825/51032/02371

Manuscript submitted January 12, 1960.

ii

INFRARED TRANSMISSION OF THE ATMOSPHERE

INTRODUCTION

The present study was undertaken to satisfy a basic requirement in the field ofmilitary infrared technology - the measured transmission of long paths in the lower

atmosphere. There have been a number of excellent direct studies of this sort in thepast (1-11), but their assembled results leave a few important factors ill-defined. Forone thing, tne paths investigated were mostly only a few thousand yards long, the longestbeing three miles. Modern practice requires a knowledge of path lengths many timesgreater than this, and extrapolation from data on quite short paths is highly unreliable.In addition, most of the earlier experimental studies have dealt individually with eitherone or the other of the two major processes involved in atmospheric attenuation -sclcctivc --.oecular absorption hv the atmospheric gases and dissipation by scatteringfrom fog, haze, or smoke.

Ln order to expand the limited knowledge of atmospheric transmission, it was feltadvisable to be able to measure and account for both absorption and scattering in onedirect observation and to have these data apply without extrapolation to atmosphericpaths of significant length. Therefore absorption spectra covering the entire practicalinfrared have been obtained over paths up to nearly 30 km. The instrumentation employedcould not be carried to desired elevations for measurements in the upper atmosphere -certainly it was not capable of being airborne - but one path lying between two mountainsabout 10,000 ft above sea level was investigat-d in order to supply some of the informa-tion needed. The present report is the final summary of this effort. Two earlier reportshave described the problem and the instrumentation and have given some of the firstresults (12,13). In an effort to be concise, much detail and some of the results containedin these earlier reports have not been reproduced here. It is necessary, therefore, toconsider these three reports together as the final product.

INSTRUMENTATION

The three sea-level horizontal paths used in this work were of lengths 0.3, 5.5, and16.25 kin, at the Chesapeake Bay Annex (CBA) of the U.S. Naval Research Laboratory.The Annex is situated on the west shore of Chesapeake Bay near North Beach, Md. Thetwo long paths are shown in Fig. 1. The 0.3-km path lay along the shore within thefenced area of the Annex. The measuring instruments were housed in the Optics Buildingat the Annex, and sources were placed at the ends of the three paths. A fourth path,27.7 km long, lay between the mountains Mauna Loa and 1tmina Kea on the Island ofHawaii. In this installation the source was located at an elevation of 9300 ft at the endof the road up Mauna Kea, while the recording instruments were housed in the U.S.Weather Bureau Station at the 11,100-ft level of Mauna Loa. Both installations were onthUe opposing faces of the two mountains, which are 13,784 and 13,680 ft high, respectively.Figure 2 shows the geometry of the Hawaiian path.

Note - H. W. Yates is now at Barnes Engineering Co., Stamford, Conn. J. H. Tayloris now at Southwestern Univerhitv, Memphis, Tenn.

1

2 NAVAL RESEARCH LABORATORY

T3NOR-- BEAC- ,C; " i TILGHMAN

S .j5.5 Km

Fig. I - Long transmission paths in theChesapeake Bay area

IAUNA LOA SOURCEIELEV '3,F•0) AT 9300" ELEV

AECEIVER / MAUNA KEAAT 1.100 E EV / (ELEV 3.74')

S277 Km*P

Fig. 2 - Optical-path geometry for theHawaiian measurements

NAVAL RESEARCH LABORATORY

The instrumentation consisted of an unmodulated source of radiation at one end of

the atmospheric path, and a monochromator with collecting mirror at the other end of

the path. An auxiliary instrument was used for measuring and continuou3ly recording

the transmission of the atmosphere in the visible region of the spectrum. The sources

were 60-in.-diameter carbon arc searchlights -ith the glass windows removed to permit

operation in the infrared. With the windows removed the arc is extremely sensitive to

wind, so the searchlights were placed behind baffles that acted as windshields (Fig. 3).

The successful use of the carbon axc in this program required considerable care, and

the interested reader is referred to the first report on this program for a detailed dis-

cussion of its operation (12). For the 0.3-kim and the 5.5-km paths, one light was usedas the source, for the 27.7-km path two lights were used, and for the 16.25-km path

three lights were used. Multiple sources were employed both to overcome the inverse-square loss of intensity with distance and to minimize atmospheric fluctuation, or"twinkle," which usually was the linmiting factor in any given measurement. By increas-

ing the angle subtended at the collector by the source, the refractive noise, or twinkle,in the intervening atmosphere can be reduced. The rms ac noise due to twinkle has beenshown to be proportional to me square rout of the angle subLte•nded by the source at thereceiver (14).

Fig. 3 - Rear view of 6 0-in. searchlightfacing into a house-type windshield

The recording instruments used hi this program were Perkin-Elmer monochromators.The installation at CBA included a Model 83 Monochromator, with LiF prism, coveringthe region 0.5 to 5.5 W,and a Perkin-Elmer Model 12-C Spectrometer, with NaCl prism,covering the region 5 to 16 x(Fig. 4). These two instruments were used simultaneouslyto reduce the time required for taking data and to allow the use of prisms appropriate tothose two spectral regions without frequent prism changes. Frequent exposures of anNaCi prism to the air during prism changes, particularly in the atmospheres of the testareas, would result in rapid fogging and unknown changes in the spectral transmissionof the instrument. Flux collection for each monoehromator was accomplished by a 16-

in. focal length, f/3.8 parabolic mirror arranged so that by reflection from two secondarymirrors, the images of several sources placed side by side would be rotated through90 degrees and appear vertically along the entrance slit of the instrument (12). For theinstallation on M1auna Loa, one instrument was used, the Medel 12-C Spectrometer, andprisms were changed to cover the full spectrum.


Fig. 4 - Receiver installation for the Chesapeake Bay measurements.One c oll ec to r rmrror is shown in the right foreground, The twomonochromators with auxiliary Newtonian flats are shown center andleft. One recording rack is visible in the background.

The collector mirror for the Hawaiian installation was a 24-in.-diameter, f/3.6parabola designed for optimum use with the Perkin-Elmer instrument and manufacturedby the Optical Shop, U.S. Naval Weapons Plant. The shadow of the Newtonian diagonalmirror was matched to the head of the thermocouple detector in the instrument, and theshadow of the Newtonian support rod was arranged to fall on the t~hermocouple supportrod so that there was no loss of effective collector area. This increase In effectivecollector area and optical quality relative tu the 16-in. collecturs caused the signal levelfrom two lights over the 27.7-km path in Hawaii to be only moderately less than thesignal from three lights over the 16.25-km path at OBA. The use of only two lights inthe Hawaiian installation was the result of logistic limitations. The installation andmaintenance of such large equipments in difficult areas are formidable problems.

The radiation detectors were Perkin-Elmer thermocouples. The outputs from thethermocouples were amplified by Perkin-Elmer 13-cps breaker amplifiers and recordedon Speedomax Type G recorders. Wavelength calibration was carried out on the spec-trometers at all installations by using the emission of a General Electric H-4 mercurylamp and the absorption bands of HO, CO, COC, CH4 , 11C1, HBr, and CH3 OH. Wave-lengths for this calibration were taken from Ref. 15.

In Hawaii, the transmission in the visible region of the spectrum was determinedby observing a calibrated lamp with a visual telephotometer. The telephotometer is ofthe low-brightness Rochester type and must be used at night (16). Therefore, operationat the Hawaii site was limited to night, when measurements of the visual trarnsmissioncould be made before and after a transmission spectrum was taken. At CBA, tworecording transmissometers were installed, and they continuously recorded the visualtransmission over the 5.5- and 16.25-km paths by day and night so that infrared trans-mission spectra could be measured at any time. These transmissometers are linearrecording instruments that require nighttime calibration by the low-brightnesstelephotometer (16).

NAVAL RESEARCH LABORATORY 5

Conventional use of the low-brightness telephotometer requires the use of a calibratedlamp and absolute calibration of the telephotUwneter. The lamp is placed at one end of thepath to be measured, and the illumination produced at the other end of the path is meas-ured with the telephotometer. Any reduction below that calculated from inverse-squareloss is attributed to the losses in the atmosphere. These quantities are related by thefollowing equation:

R K K , (1)

where

R = telephotometer reading (arbitrary divisions)

K = calibration constant of photometer (ft-candles/divisIon)

I = intensity of lamp (candlepower)

D path length (f)'

T = transmission of the atmosphere over path length D.

The technique used in the present program takes advantage of the fact that transmis-sion is a dimensionless number requiring absolute calibration of neither the lamp nor thetelephotometer. A calibration of the teiephotorneter relative to a given large lamp iscarried out by using a small intermediary lamp to simulate at short distance the effectof the large lamp at great distances. This small lamp (about I candlepower) is placedsfriit 15a ft rrrnm te tenhn~rMtnn !o-nr 4 -1-te dta-ca -p I fl - ---

by whirling sector discs in front of the lamp. The telephotometer readings are then givenby

R = K - t, (2)

where

I = intensity of the small lamp (candlepower,

d = effective distance (increased by sector discs) betweentelephotometer and lamp (ft)

t - transmissiun of 150-ft path used.

If R e i4justed to be the same for both the large distant lamp and the small lamp, &p..(1) and (2) can be equated.

-T =.it, (3)DO d'

with K dropping out. Over the short 150-ft path, t can he considered to be 100 percent,so that

T uoF)(P) (4)

It in therefore necessary only to know the intensity ratio (i/1) for the two iamps and thetwo distances d and D. The large lamps used were 1000-waft airway beacon lamps of


about 1500 candlepower, and the intensity ratios between them and the small 1-candlepowerlamps were obtained on a photometry bench, where one leg was about 130 it.

Measuring d for each individual transmission measurement was not practical in thefield, and so the telephotometer was calibrated in terms of 1/d' for one particular smalllamp. Thus, when a measurement was made over a particular path of known length D,the telephotometer was used to measure a value of 1/0I for the large remote lamp, andtransmission was calculated from D, l/d 2 , and the previously measured Intensity ratio(0/I).

EXPERIMENTAL PROCEDURE

The first step In the measurement of the infrared transmission of the atmospherewas to measure the radiation from the source over the 0.3-km path. It was felt that asatisfactory vacuum envelope could be drawn from such a spectrum. The LIF instrumentscanned the spectral region from 0.5 to 5.2 gi, and the NaCI instrument scanned the regionfrom 2,72 to 15 -. In recording the spectirum uver the 0.3-km path, maximum resolutionwas not used. Instead, a slit schedule was used which would give sufficient signal whenused with the longest paths. The choice of a 0.3-km reference path, rather than a muchshorter one, was made primarily to make the geometry approach that which existed inthe case of the distant lights. In all cases, it was the searchlight mirror which was imagedon Lie entrance slits. The vacuum envelope used in reducing the data was obtained fromseveral runs made on the 0.3-km path. All of these runs were reproducible to about twopercent.

The second step in the procedure was to record the infrared spectrum over one ofthe distant paths with the same slit program used in the 0.3-km path. By taking the ratioof the deflection at a given wavelength obtained over the distant path to the deflectionobtained over the 0.3-km path, a number proportional to transmission was obtained.Since absolute transmission values were required, it was necessary to determine a factorwhich, when multiplied by the above-mentioned ratio, would yield the absolute transmissionat the given wavelength. This factor was determined at 0.55 pL for each path by means ofthe visual telephotometer and auxiliary light source.

When the factor at 0.55 g which -would yield absolute transmission values was deter-mined, the spectrum obtained with the LIF instrument was then reduced point by point toabsolute transmission values. A plot was then made of transmission versus wavelengthfrom 0.55 to 5.2 u. Since the region from 2.72 to 4.2 p was scanned by both the LiF andNaCl instruments, it was used as a tie-down region for the normalization of the NaClmeasurements. For this region, the area under the curve obtained with the LiF instrumentwas planimetered, and an average absolute transmission was calculated. The same pro-cedure was repeated for this window using the spectrum obtained with the NaCI instrument.The value obtained was proportional to the average absolute transmission in the window.it was then possible to calculate the factor which, when multiplied by the average valueobtained with the NaCI instrument, would yield the average absolute transmission asobtained with the LIF Instrument. Knowing this factor, it was possible to reduce point bypoint the spectrum obta-ined viRth the NaCl instrument in the long-wavelength region toabsoiute trAnsmission values.

RESULTS

The basic data on which thu results of this report are based consist of ... ±, a:pletespectra covering the region 0.5 to 15 ai with an average resolution, XsAX, of about 300.Of the 51 complete spectra, 15 were reduced to absolute values of transmission vs wave-length over the full spectral range, five more were reduced only to values of scattering


coefficient vs wavelength, and five others were reduced only to values of selective windowtransmission vs water-vapor content. The remaining 26 were found, upon examination,to duplicate the conditions and results of spectra already reduced. Except for a cursorycheck sufficient to establish the degree of agreement (45 percent for selective windowtransmission and ±10 percent for scattering coefficient is the average agreement foundbetween any two runs covering near-identical conditions), nothing further has been doneto these spectra. They are all on file at the U.S. Naval Research Laboratory in the formof unreduced Speedomax chart records.

Nine of the complete spectra, selected to be representative samples of the totalnumber, are reproduced in Figs. 5 through 10. One additional spectrum (Fig. 11) wasrecorded for a short path under maximum resolution with the prism spectrometer byusing LiF, CaF, NaCl, and KBr prisms to cover the interval from 0.5 to 26 j' with aresolution ranging from 1 to 2 wave numbers in the region beyond 2 p. Another set ofthree spectra, one each over the 0.3-, 5.5-, and 16.25-km paths, was published in anearlier report (12) and is not included here. A wide range of precipitable water is repre-sented from 0.11 cm in the 0.3-km path to a maximum of 38 cm in the 16.25-km path -and values of the visual scattering coefficient range from 0.048 km-1 to 0.24 km- 1 .

ATTENUATION BY SCATTERING

Introduction

Scattering acco4nts for a significTnt loss of flux from any collimated or narrow lightbeam passing through the atmosphere. The molecules of air scatter light, but their effectis negligible in the infrared, which is the region of the spectrum considereld here. Althoughsome solid matter :is suspended in the air and contributes to the scattering, it is priinkrily-droplets. of liquid water, deposited on condensation nuclei, that account for most of thescattering in thib atmosphere. Arid regions are always known for atmospheric clarity andhumid regions for haze and fog. A short atmospheric path on the east coast contains suf-ficient haze to give a milky appearance to the terrain beyond, while the Rocky Mountainsare usually visible for a hundred miles or more.

The scattering properties of the atmosphere are most frequently described by thescatterinig coefficient a, which gives the flux scattered into 4v steradianp from a unitvolume of air irradiated by parallel light producing unit-irradiance at the unit volume.,The transmission t of a purely scattering path of length x for parallel radiation is relatedto the scattering cpefficient by the expression

axt=e

where the assumption is that a does not vary along the path.

- Most generally, the quantity a is a function of wavelength and isCproperly definedonly by reference to the spectral distribution of the light. In this re~ort it is defined formonochromatic light and should be called the nionochromatic !scattering coefficient. Forconvenience the shorter term is used. -

The variation of a with wavelength was one of the areas of interest in the presentwork, and it is appropriate to review briefly the parameterfs which -deermiine-the 'variation.The theory of light scattering by dielectric particles predicts that a is defined by the fol-lowing expression:

a - ' 2 rK r2N (5)

N


where N is the concentration of particles with radius r, K is a function of the quantityr/X and the refractive index of the particle, and x is the wavelength. The quantity K (r/x).can be computed theoretically.

The key parameter is the quantity r/X, the dimension of the particle relative to thewavelength. For r/X < 1, a varie-- as 1A 4 (Rayleigh scattering). For r/ >> 1, a isindependent of X, and the scattering is nonselective. The intermediate case r o¢ A iscalled Mie scattering and Is theoretically the most complex case. In any real atmospherethere is a wide distribution of particle sites, and the scattering coefficient at a particularwavelength is determined >.y .itýrnrnig over the whole size range.

The present data were used to estimate typical values of a as a function of wavelengthfor real atmospheres. There are very few published data for this type of function throughthe near inirared, yet the information is quite necessary to estimate transmission overlong paths.

The present work deals with real atmospheres, and in any work of this type it isconvenient to characterize an atrmosphere by indicating its transmittance for visiblelight. Frequently this is expressed as a "visibility," or the distrnce at which a largeblack object can just be detected against the horizon. This concept, however, involvesthe variable contrast threshold of the. human eye and is not a satisfactory one. A sort ofidealized "visibility" is preferable, and this has been termed Meteorological Range (MR).This is the length of air path in which the scattqred light produces a luminance equal to98 percent of the luminance of an infinitely long airpath of the same characteristics. Thecontrast of a dark object against the horizon is thus reduced to 2 percent if the object isat the MR. Although the concept involves white: light and the standird curve of humanvision, it can also be applied to monochromatid.: light at. 5500 A, :and is so applied here.The MR is related to the scattering coefficient .ar by the relation

MR 3.94a

.Figtire 12 shoWs a plot of MR vs hcattering coefficient. Also shown in Fig. 16 as aseries of solid points are values of visual range vs scattering coefficient, as derived fromthe International Visibility Tables. The two sets of values differ slightly because thecontrast threshold of the human eye, which enters into the visibility criteria, differs some-what from the 2-percentthrgshold assumed for the-MIR.

Experimental Results ...

":Settking coefficients, crAm, were determined at 12 wavelengths from 14 of themost reliable, speutra. The determination was based on the assumption that absorptionis -negliglb!e at ce-tain spectral points away from intense absorption bands, and that inthese "wiitdows" the measured attenuation could be completely attributed to scattering.This assumption appears quite valid for the wavelengths 0.55, 0.60, 0.70;. 080, 0.90, 1.06,:1.26, 1.67, 2.17, 3.6, and 4.0.g, and a was computed for these points. Values of a werealso computed for 4.70 A&, even though the assumption is nit vklid at this wavelength"becauseof'the preSence of weak absorption bauids throughout -the "window." The assump-tion is also genei-lly invalid at longer wavelnths. It -pt-ibl&-hat-with-V-ery-tigh-

.... resolution smaz.'regions free of absorption could be found at longer wavelengths, but-.... none were discernible In this work.


I 8o/

1 I00 I T '

260

04 20

0.1- -1 L -YI A _J

05 06 07 08 09 10 1.1 12 1-3 I4 .5 [6 .7 8 19WAVELENGTH (MICRONS)

10010 1 1 1 , -J = i I -I-]'

10

60-

64

Vz 20-

018 19 20 2-1 2 2 2.3 24 2-5 2 6 2.7 2-8WAVELENGTH (MICRONS)

100E

080-

a; 60 uz

L .40 n

Inz 20

2229 3.0 3&1 32 33 3A4 3,5 3.6 3.7 38 3.9 4.0 4.1 42 4.3WAVELENGTH (M;CRONS)

Fig. 5 - Atmospheric transm-ission over 5.5 km and 16.25 kit; in theChesapeake Bay area. Data from runs 58 and 60 over 5.5 km andfrom runs 59 and 61 over 16.25 km.

5.5 kr - April 19, 1956, 38'F, 66 percent R.H.,2.2 cm H2 0 in path, 40 percent transmission at 0.55 .L

16.25 km - April 19, 1956, 530 F, 41 percent R.H.,6.5-6.9 cm H.0O in path, 29 percent transmission at 0.55 At


400

U4O 1. lrk ..- -

<z2;2JAALM

rr--

q546.046 47 48 49 5-0 51 52 53 54 55 56 51

WAVELENGTH (MICRONS)

5100rn- Aril19 196 38F 6pretr.,

I6- m- Ap~ 9 96,50,4 eretI..

z.ah8.0 H8lt

cr

o AA)\

Cý56.0O 7.0 8.0 -90- loii.0 1'. 30 14.0 .15WAVELENGTH (MICRONS)

Fig. 5 (Continued) - Atmospheric transmission over 5.5 km and 16.25 kmin the Chesapeake Bay area. Data from runs 58 and 60 over 5.5 km andfrom runs 59 and 61 over 16.25 km,

5.5 km - April 19, 1956, 38 0 F, 66 percent R.H.,2.2 cm H-1O in path, 40 percent transmission at 0.55 sA16.25 km - AprJ 19, 1956, 530F, 41 percent R.N.,6.5-6.9 cm 1420 in path, 29 percent transmission at 0.55A


40

•zOfi I I I I , r I L T


10 T F T I I "I 1 -I ""

WI o

Cr

_ IO&i4

I-2 lipI I I •

. 0. 7 8 II .2 l 2. 1 2 23 14 15 16 1.7 28 21 27 1 .


1001 6 - -i I aI i oI I I d 1 5h

Wa 6

Chesapeake Bay area. Data from runs 70 and 71 over 5.5 km andfrom runs 68 and 69 over 16.25 kmn.

5.5 kin- June 19, 1956, 9 p.m., 64 0 F, 51 percent R.H.,4.18 cm EIsO in path., 70 percent tra.nsmission at 0.55 i

16.25 km - JIune 19, 1956, 4 pan., 68.70 F, 53 percent R..Ht.,15.1 cm IfO in path, 43 percent transmnission at 0.55 27


w, 80

60

m40

-I

%3 4.4 4 5 46 47 48 4.9 50 5- 02 53 5.4 55 d56 -57


1000

0 -

5-5 6.0 7.0 80 9.0 10.0 ;,3 12,0 1ý.v 14-0 15.0


Fig. 6 (Continued) - Atmospheric transmission over 5.5 km and 16.25 kmin the Chesapeake Bay area. Data from runs 70 and 71 over 5.5 km andfromn runs 68 and 69 over 16.25 kin.

5.5 km - June 19, 1956, 9 p.m., 640F, 51 percent R.H.,4.18 cm H2O in path, 70 percent transmission at 0.55 A

16.25 km - June 19, 1956, 4 p.m., 68.70F, 53 percent R.H.,15.1 cm H210 in path, 43 percent transmission at 0.55A


100 T . ] A -I F T- 1 , ...

40L-

0.5 0.6 0. 0-8 09 t.o 1 L2 [3 14 15 L6 17 L8 [9WAVELENGTH (MICRONS)

WLrI I

08 19 2D 2.1 22 23 24 25 tS6? 2.8ZWAVELENGTH (MICRONS)

!00- ----i , ,FT--T --- F r F l

a;60-

20 2.9 30 3.1 32 33 34 35 3.6 37 38 39 40 4] 42 43


Fig 7 - Atmospheric transmission over 5.5 km and 16.25 km in theChesapeake Bay area. Data from runs 78 and 81 over 5.5-kmn path,and from runs 77 and 80 over 16.25 km.

5.5 kmn - Aug. 27, 1956, 6 p.m., 78 0 F, 73 percent RH.,9.4 cm H2fo in path, 30 percent transmission at 0.55 L

16.25 km - Aug. 27, 1956, 11 a.m., 740F, 82 percent R.H.,27.7 cm H20 in path, 10 percent transmission at 0.55 IL


1001 1 r T -" r-- ----- -• • -• 1 T • [1

800

I60-&240-

%344 45 46 47 41, 49 S 0 .51 ? .%.A 54 55% 56 97

Uj

6001 1 , ,6,

z

zI 2

5.5 6.0 7.0 8.0 9.0 I0.0 110 120 13.0 14.0 150WAVELENGTH (MICRONS)

Fig. 7 (Continued) - Atmospheric transmission over 5.5 km and 16.25 kmin the Chesapeake Bay area. Data from runs 78 and 81 over 5.5-kmn path,and from runs 77 and 80 over lb.25 km.

5.5 km - Aug. 27, 1956, 6 p.m., 78(F, 73 percent R.H.,'9.4 cm H 20 in path, 30 percent transmission at 0.55 p.

16.25 km - Aug. 27, 1956, 11 a.m., 740F, 82 percent R.H.,27.7 cm H 2 0 in path, 10 percent transmission at 0.55 A.


100 1 I I I 11 -

Lo t0fit III]

60[4 IZr /

Z20

I- I


1001 1-- I- ]- I T [ml--

H1 • I

[8 I.S 20 2.1 2.2 2.5 2.4 2.5 2.6 2.7 28


I0 i i i i i i i i i [ i [

I,-

zL

"'880

z

0

zZ 20'

.8 1- 20 21.13 2.2 23 2.2.5 2.6 2.7 2-3. 40 41 4_ {'WAVELENGTH (MICI:ONtS).

1 I00 ; , i

w 80 -

'600 I

•_ 40 -

0 20}-

0 I I -I J

42-8 4_9 3.0 31 3.2 3-3 3-4 3.5 36 3.7 36- 351 40 4.1 42 4.3WAVELENGTH CMICRONS)

z

;0-

0.

9

z'<20-

4.3 4 - 4.65 4 4.8 4.9?0 ji !2 5. 4 t. ý 5.7WAVELENGTH (MICRONS)

Fig. 8 - Atmospheric transmission over 16.Z5-km path in theChesapeake Bay area. Data from run 83.

16.25 km - June 16, 1957, 7 p.m., 87.5 0 F, 59 percent R.H.,36-38 cm H 2 0 in path, Z percent transmission at 0.55 ii


I00

so0z

40

20 NS . I I 1 ,,II I

°O5 0.6 0.7 0.8 0.9 .O 1.1 L2 1.3 L4 1.5 1.6 17 LB 19WAVELENGTH (MICRONS)

1 00080a I-

-60-zL(n 40- tiN .

ýj20-1" I , I I I I•F'

018 1.9 2.0 2.1 2-2 2.3 2A 2-5 2.6 2.7 28


100 1 1 i I I

zw 80oU

GO60

Cn40

1 •2. 2.9 3.0 3.1 32 33 3A 35 3-6 3-7 3.8 33 4.0 41 4.2 4.3


Fig. 9 - Atmospheric transmission over 27.7 km in theHawaiian Islands. Data from run ML-3.

z7.7 km - 10,000-ft altitude, Sept. 1, 1957,11:30a.m., 430F, 100 percent E•M., 20 cmH 2 0 in path, 26.5 percent transmission at0.55 KL


lOG i I I TI F F I I I I

~60-

20

20I.,,I II I I . I

043 4.4 4.5 4.6 47 48 4-9 5-0 51 5.2 53 5.4 5.5 56 .7WAVELENGTH (MICRONS)

100

60-'Ln

Z 2In

5 6.0 7.0 8.0 9-0 I 1.0 12.0 13-0 14.0 t5.0WAVELENGTH (MICRONS)

Fig. 9 (Continued) - Atmospheric transmission over 27.7 km in theHJawaiian Islands. Data from run ML-3.

27.7 km - 10,000-ft altitude, Sept. 1, 1957,11:30 a.m., 431F, 100 percent R.H., 20 cmH 2 0 in path, 26.5 percent transmission At0.55 p


z s-

Z

E- 60

040 /20I '.5 -L.I -- t-

0.6 0-7 08 0.9 i. U 12 1.3 IA 1.5 L6 1.7 L8 1.9WAVELENGTH (MICRONS)

G°o-00 I._-_L I I I I• ! I j

i-n 40F -

201 1i-8 19 2.0 2,1 22 2.3 .4 2.5 2.6 2-7 2.8

WAVELENGTH KM!CRONS)

100 , , 1

so-

20MF--

2.8 2.9 3-0 3.1 32 3.3 3-4 35 3.6 37 3.8 3.9 4D 4.1 42 43WAVELENGTH (MICRONS)

Fig. 10 - Atmospheric transmission over Z7.7 kmn in theHawaiian Islands. Data from run ML-6.

27.7 km - 10,000-ft altitude, Sept. 6-7, 1957,11:00 p.m., 8-13 cu, NI2O in path, 41 percenttransmission at 0.55 1


00 I I I I

M 880a:hU

0-60

S40

04'.3-44- 4.5 46 4.7 48, 4-9 5.0 5. 52 5.3 5.4- 5.5 5.6 5.7 5-8WAVELENGTH (MIgRONS)

IUO F r• -, , -

o 6

0Cn 40.in_

I-

055 6.0 7.0 SO 90 10.0 I110 120 13.0 140 15.0WAVELENGTH (MICRONS)

Fig. 10 (Continued) - Atmospheric transmission over 27.7 km in theHawaiian Islands. Data from run ML-6.

27,7 km - 10,000-ft altitude, Sept. 6-7, 1957,11:00 p.m., 8-13 cm H 20 in path, 41 percenttransmission at 0.55 I'

90- 02 V 2 1

w 60

z

,020N

100

I.IoL, n i i j n l ii i n I .i i i.hllI I. [.I n I: nli m In ii I n i n | .I II U

0.5 0-6 0.' 0_8 0.9 .0 1. .2 1.3 .4 .5 1.6

WAVELENGTH (MICRONS]

[00

90 I 5.7MM PRECIPITABLE 1H

2 0 AND Co \I2 1 79* F TEMP

N 2 Li F PRISM80 V~

;7 701z

60o

o 50 -H 2 0 AND CO 2

ch 40

30

20-002

I0 . .. . .. I .. . . . . ,-J J- I • ' u I I I I I I n- -- L n .L J .J Jn n , I I n . I I I I I I nI UL~ n I I ; I ;n L- i i i i t i

1.7 L.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8


Fig. 11 - Atmospheric transmission over a 0.3-km path in the Chesapeake Bay area.Maximum available resolution used.

I

21(PAGI 2, SLANK)

V (4

^ I

0 .7 o .8 0 .9 1. 0 I 1. 2 - 1 _3 1.4 1 _5 1[ 6 L

WAVELENGTH (MOCRONS)

A 1 5.7MM PRECIPITABLE WATER79° r, _EM

,AND CO 2 Li F PRISM

H20 AND C02

19 20 22 2.3 24 2.5 2 .6 2 -7 2,8 2.9 30


Fig. 11 Atmospheric transmission over a 0.3-km path in the Chesapeake Bay area. IMaximum available resolution used.

C)'

so !70

5 ,An'p 2 2.2CM-I50

20.

i" .. . . . I j l 1- •1 1 1 [i L 1 . , 11 . l iL l ll l : 11 i l l I i ii iI' ~ L ll I -I i- i | l • i l f r [ .1 i i Ii i i Ii i I I I i]1 r I 11 1• i 4

30 3. 3.2 3. 3 A 3.5 1 31 ' 3.7 3.8 40 41


90-0

80_ 57 MM PRECIPITABLE WATER

79* F TEMP

7 0 - J I

F , ,S

i __ A~l X :O.O030$A XT O.0035p X-.031

,1. 1v=Ll CM-I

< 30--l

I - :I ll i i II iii i ' i : . . i l III I11 II1ni l ii l ii I1, 1 il, , I i Inn r ~ r L t i. ii Ii - .L

42 4.3 44 4.5 4.6 4.7 4.8 4.9 5.0 5.1 5.2 5.3- •WAVELENGTH (MICRONS)

Fig. 11 (Continued) - Atmospheric transmission over a 0.3-kmn path in the Chesapeake Bay area.Maximum available resolution used.

'i~j

23

li1t tII • •

LjjI ... .. i - i L III ii 1 11 -J 11 1 I I jIIii~I Iu l ,- i ...... I I I I I II 1,11I 1I I I I I I I I I I

3.2 3.3 3.4 3.5 3-6 37 38 3.9 40 4.1 42


II 5.7 MM PRECIPITABLE WATER

79;ii FTEMPLi PRISM

Lw. 1CM-'at-.1L CM1

H20

4.4 4.5 4.6 4_ 4.8 4-9 5.0 5_1 5.? 5.3 5.4 55

WAV-LENGTH (MICRONS)

Fig. 11 (Continued) - Atmospheric transmission over a 0.3-km path in the Chesapeake Bay area.Maximum available resolution used,

100

90

80

70-) 57 MM PRECIPITABLE WATERw 790 F TEMP

Uw60a;Z 50 H20

H22

2 400,z

X 30-

20

10

05. ' , .56.0 65


0 Ca F2 PRISM -4 Na

9o0 A XrO0096

801.8 CM- '

Z70I5.7 MM PRECIPITABLE WATER

S_. 79' F. TEMP. ii0. OU•

z0 50 -

n 40 -

1- 30-

I10

I08TO0 7.5 8,0

WAVE LENGTH (MICRONS)

Fig. 11 (Continued) - Atmospheric transmission over a 0.3-kmi path in thc Chesapeake Bay areaMaximum available resolution used.

25

5.7 MM PRECIPITABLE WATER79° F TEMP.

H20

40.0 6-5 7.0WAVELENGTH (MICRONS)

Ca F, PRISM - NoCI PRISM --

AOX=0.0O96

A, 1.8 C M-1 A!&O

5RECI PITABLE WATER IP9* F. TPMP. I

SN I I I I I I I7-5 8.0 8.5

WAVE LENGTH (MICRONS)

Fig. 11 (Continued)- A-tmospheric transmission over a 0.3-km path in the Chesapeake Bay area.

A2"

100 nfl.-_____90

soI70

57 MM PRECIPITABLE WATERI-. JV77 F. TEMP.Mb No CL PRISM

w, ,")' 2.1 (•M"1

10

cj 5 IIII,

.5 io 9.5 0.0 10.5 I!_0

WAVELENGTHi (MICRONS)

100

90I

80

70---

60 A1.0.017 A)6*0.02 ILcc Av-12 CM " AV- w CM "

-50 --

l40 - u

• 305.7 MM PRECIPITABLE WATER

77* F. TEMP. C020- No CL PRISM

10 -

0115 1.0 12.5 13.0 13.5 14.0


Fig. 11 (Continued) -Atmospheric transmission over a 0.3-kmn path in the Chesapeake Bay area.

Maximum available resolution used.

27(PAGI 2,3 Bh.;

5.7 MM PRECIPITABLE WATER77° F. TEMP.No CL PRISM

AX-O.Offu

I II ,-9.5 I0.0 10.5 1!.0 1.5

WAVELENGTH NMIGRONS)

V

.A -12 CM-I A . 0 CMq

5.7 MM PRECIPITABLE WATER77* F. TEMP. 1-0N• CL PRISMI

.I LI

120 12.5 13.0 13.5 14.0


Fig. 11 (Continued) - Atmospheric transmission over a 0.3-km path in the Chesapeake Bay area.Maximum available resolution used.

C9/

t

90(PAUF Y; BLAqk)

100

9s-

80 -H- F-II-Av=I.8 CM4j,•,'. 3 CM-1

70-H20

060-

50-

0

0 L0 " "I i I I I I I I I I II I I. -- L_

15 !6 17 3 19WAVELENGTH (MICRONS)

100

&X-O.O631. A.,O.05Tu 2.2 MM PRECIPITABLE WATER

80* 1 F 56-F TEMPiv 1 [6 CM-' & 12 CM-

1 K8, PRISM

70

Lz 6O - 20

(40

<430 IIIi I

201

10

Maximum available resolution ut~ed.

29(PAGE j1) B-ANt)

AA ,- )O041jA I-

Av-L3 CM'I

H20

Vr

I I II L I I II I6 7 18 19


A),=O063j. A). .57, 2 2 MM PRECIPITABLE WATER

-41--IF56*F TEMP1 16 CM

1 , 1.2 CM- I KBr PRISM

H 20

20 21 222 2 3 24 25WAVELENGTH (MICRONS) 2

Fig. i1 (Continued) - Atnospheric transmission over a 0.3-km path in the Chesapeake Bay area.Maximum available resolution used.


2 2"24,

-__1____________

IA-

• 1.2•2

L __ _. • • _. I I I I ,L JI

1 2 3 4 5 G a7 9 V 0 30 40 50 9 80 ?00RANGE (KM)

Fig. 1Z - Visible scattering coefficient a per kin, as afunction of range. The range is meteorological range(MR) for the solid curve, which assumes a 2-percentthreshold limit. The solid points indicate visual ranges

taken from the International Visibility Code.

The experimental scattlerig coefficients are shown in Figs. 13 through 16. InFig. 13, four runs over the 16.25-km path are shown, ranging from a visible-regionscattering coefficient of 0.24 to 0.073 km-', Figure 14 shows four additional curvesfor the same path and over the range 0.185 to 0.51 km-1 in visible-region scatteringcoefficient. Figure 15 shows four runs over the 5.5-km path, where a in the vitsiblespectrum ranges from 0.58 down to 0.065 km-*

Also plotted on Figs. 13 and 15 are typical atmospheric scattering coefficients inthe 0.3 to 0.6 gi region as measured in NRL work reported by Dunkelman (17). They areincluded to indicate the behavior of the atmosphere in the visible and ultraviolet.

In Fig. 16 are shown the points from two runs over the 27.7-km path between themountains in Hawaii. These two sets of points lie very close together, and one curve isdrawn through them. The scattering coefficient at this relatively high altitude (one endof the path at 11, 100 ft, the other at 9300 it) is surprisingly high and, except for theoccasional intervention of dense clouds, also surprisingly constant. There appeared to bean appreciable amount of scattering material lying below a fairly well defined ceiling ataround 9500 to 10,000 ft. The searchlight beam, originating at 9300 ft and rising to11,100 it, could be seen from the side to start out bright and rather abruptly decrease inbrightness. From the Weather Bureau Station at 11, 100 it, the 10,000-ft peak Haleakela onMaui (80 mi away), the mountainous Molokai (135 mi away), and the mountains on Oahuwere always clearly visible when there were no intervening clouds. This would not be pos-sible .f the scattering coefficient over these high paths were as large as that measuredbetween Mauna Loa and Mauna Kea. From Fig. 16, it will be seen that at Mauna Loa themeasured value of a at 0.55 gt was 0.05/km which, from Fig. 12, corresponds to an MRof only about 40 km. It is believed that the measured scattering coefficients would havebeen appreciably smaller if the source on Mauna Kea could have been moved another twothousand feet up the mountain, but the logistical problems involved in such a move wereinsurmountable in the present program.


TYPICAL[SUMMER DAY

Fig. 13 Scattering coefficient a per kin, forbeJeC.edo avcluagth po~ints waj from absorp-tion bands, d at a from measurements ove r

00 16.Z5 km at sea level. The dashed lines are0.01 1111111 from Table 5 of Ref. 17. The solid straight

0.1 1.0 io line is the Rayleigh curve for sea-level.WAVELENTH (MICRONS)

RUN 83, 38 CM .20

RUN 39, 6 CM H20

- NUN b4, ZZ.f CSM H20

a RUN 61, 6.9 CM H;,O

TYPICAL

0.1 ~ OJWNTER DAY

000 a

Fig. 14 - Scattering coefficient a po!r kmn, forselected wavelength points away from absorp-tion bands, d a t a from measurements o v e r16.25 km at sea level. The solid straight line 0o, I I i,is the Rayleigh curve for sea-level air. o-i i. 10


X RUN 35, 15 CM H20

* RUN 77z2 CN H20

o RUN 5= 5.2 CM IH0

A RUN 68, 15 CM H20


I. I-

Fig. 15 -Scattering coefficient u per k, forselectedwavelengthpoints away fromabsorp-

tion ba-nds, data from measurements over 5.5km at qea levPl. The dashed lines are fromTable 5 of Ref. 17. The solid straight line isthe Rayleigh curv. for sea-level air. 0o 1.0 I_


X RUN 40, 2.20 CM H20

RUN 19, 3.75 CM H,0RUN 60, 2.15 CM H20

A RUN 70. 4.18 CM H2

SO.I

Fig. 16 - Scattering coefficient a per kcm, forselected wavelength points away from absorp-tion bands, data from measurements o v e r

oIo i Z7.7-krn path at 19,000-ft altitude.

WAVELENGTH IMICRONS)

* RUN ML-3, 20 CM HOO RUN ML-6, 11C M H20


For comparison, the siupe of a pure Rayleigh scattering characteristic (a - X-1) isindicated on Figs. 13 through 16. These lines are placed on the graph in a position repre-sentative of pure sea-level air devoid of any contamination such as smoke or fog. Insuch an atmosphere the scattering is by the molecules of 02 and N2.

From an examination of Figs. 13 through 16, it will be seen that the scattering coef-ficient in the infrared is normally about half the value In the visible region. The reductionis more pronounced for a hazier-than-average condition and less pronounced for a clearer-than--average condition. In addition, there were anomalous results (run 57 in Fig. 14 andrun 40 in Fig. 15), but these were rare. The only two encountered are the two shown inFigs. 14 and 15, and there was nothing anomalous about these days, either visually ormeteornlogically, to aid in an explanation. It can only be construed that these days werecharacterized for some unknown reason by a more-than-average uniformity of particle size.

Since the present program was primarily concerned with long atmospheric paths, itwas not practical to Investigate highly opaque conditions such as dense fogs, rain, and snow.Through such conditions, the received radiation level was usually too low for measurement.Only two successful measurements were made in fouil-weather conditions, one through alight snowstorm and one through fog mixed with light rain (12). The results of these twomeasurements are listed in Tables 1 and 2, where the transmission ratios between selectedwavelength regions are shown. The actual transmission values fluctuated rapidly throughfog and snow, making it impossible to produce a meanir ° " spectrum. Therefore, thespectrometer was set successively to the several dis. .Le wavelength regions listed, andthe fluctuations averaged out to produce the transmission ratios. Both runs were madeover the 5.5-km path, and the slits were opened wide in order to achieve a measurableradiation level. Since the original slit program was abandoned, the actual transmissionat any one point can be estimated only from signal level and is therefore probably inaccu-rate by a factor of as much as two to five. In the case of the snowstorm (Table 1). thetransmission in the visible (0.54 11) was estimated at 0.004 percent, and in the case of thefog (Table 2), the estimated visual transmission was 0.04 percent. The impo:tant factorto remember in looking at these tables, therefore, is that while the transmission in theinfrared - around 3 to 5 p - improved over the visible region by as much as 20 for thesnow and 60 for the fog, this still results in absolute transmission values of only 0.08percent and 2.4 percent, respectively.

Table 1Ratio of Transmissions Over 5.5-km Path

During Snowstorm

Ratio of Transmittance at X1X X2 to Transmittance at X2

(microns) (microns)Rn1 Ru~n 2

1.26 0.54 8.9 9.0

1.54 1.26 1.58 1.63

2.18 1.54 1.06 1.2

3.98 2.18 1.15 1.46

4.69 3.98 0.77


Table 2Ratio of Transmissions Over 5.5-km Path

Through Fojg Mixed with Light Rain

At 1 '2Ratio of Transmittance at X.t(microns)1(miczrons)I to Transmittance at X,

1.04 0.54 8.14

1.21 1.04 1.45

1.59 1.21 1.45

2.27 1.59 3.0

3.64 2.27 1.25

There have been other experimental programs directly concerned with the meas-urement of transmission through fog and bheavyv weather, notably the work of Arnuif, et al.,In France (18), and the reader is referred to this source for more extensive results.

SELECTIVE ABSORPTION

Introduction

The most severe atmospheric attenuation In the infrared region of the spectrumresults from selective absorption by the gases In the atmosphere. Arranged in order ofdecreasing practical Importance, these gases are water vapor, carbon dioxide, nitrousoxide, .ozone, oxygen, methane, and carbon monoxide. This order differs from the orderof atmospheric abundance for these gases, because some of the absorption regions over-lap, and the rarer constituents may be masked by one or more of the abundant constitu-ents. For example, methane is more abundant than nitrous oxide, but its absorptionregions occur near intense water-vapor and carbon dioxide bands, so that its presenceor absence is difficult to detect without very high resolving power. It. contributes verylittle to the absorption encountered in the lower atmosphere. The relative abundanceof the infrared-active gases which make up the atmosphere are listed In Table 3 (takenfrom Ref. 19), where their concentrations are given in atmosphere-centimeters (atmos-cm), or the thickness of a column of gas at normal temperature and pressure which isequivalent to the amount of gas contained In a vertical column of atmosphere. To computevolumnetric concentrations, these numbers must be compared to the entire atmosphereequivalent, which Is 8.04 x 105 atnios-cm.

Of these constituents, oxygen, carbon dioxide, nitrous oxide, and methane appear tobe uniformly mixed In the atmosphere and to have stable concentrations. Water vapor,of course, is highly variable in concentration and is pronouncedly stratified. Figure 17illustrates this large vertical gradient in water-vapor concentration, with the resultsof three series of measurements taken from balloon-borne instruments (20). More than90 percent of the water vapor lies In the atmosphere below 15,000 to 20,000 ft, and it Isin this region that the most pronounced changes in concentration occur. Above 20,000 ftthe water-vapor concentration tends to follow the air-density curve, indicating a morenearly uniform volumetric concentration. Ozone, which absorbs appreciably in a strongband at 9.6 I-L, Is confined principally to a layer lying from 15 to 35 km above the earth'ssurface. It exerts little influence on optical paths which do not include this layer, but it


Table 3Molecules Identified in Earth's Atmosphere*

Atmos-CmMolecule Molecular Weight (NTP)

0, 32.0 167,600

Ollb 34.0 659

O'101 44.0 320

0o1b17 33.0 134

C1 0 2sO 45.0 1.5

CH 4 16.0 1.2

C10"Go's 46.0 0.67N20 44.0 0,4

0, 48.0 0.3

CO 28.0 0.06 - 0.15

42O 18.0 103 -104

HDO 19.0 0.4 - 4.0

R~e f. 19.

Is discernible in the results of the present study (21). Its concentration at sea level isonly a few parts per million and is highly variable, being generally more evident by daythan by night.

Atmospheric Windows

The striking feature if the infrared absorption by gases Is the localization of regionsof high absorption into marked bands. Generally these bands are vibration-rotation bandsof the particular gas molecules. It might be mentioned that the homonuclear molecules,such as 0 rtid N2, do not have absorptl, a bands of this type. The near-infrared 0,absorption structure (0.78 p) corresponds to an electronic transition.

As can be seen from any of the spectra in Figs. b through 11, the selective absorp-tion bands are scattered through the .nfrared spectrum. Between the absorption bandslie regions of moderate or no selective absorption. These regions of relatively goodtransmission have been named "atmospheric windows" and have the generally accepteddefinitions listed in Table 4.

In spf,.ifying the transmission of an optical path at wavelengths which cover atmos-pheric windows, a very useful quantity is the "selective-window transmission." Thisconcept was first introduced by Elder and Strong (22) and has become widely used. Theselective-window transmission for a given window is defined as the ratio of the energythat actua1ly penetrates the atmosphere between the wavelength limits of the window,


F100'

MEASUREMENFS FHOM

HIGH-ALTITUDE BALLOON

*-.. '"- I JULY 1949 lC'_UP .PLEV, V0- - 26 AUG. 1949J W.... x..-. 7 JAN. 1950 ST.0.UIS, MO.

-' I.O ,

ot

It ) . , I-=

-:1

0-0' a

0.01- JO

AIR 0-

DENSITY I- U.

Cr

110C I0 1001 IOWC

ATMOSq•HEIC PRESSURE WMILLI5ARSI

I I I I I 150 40 30 20 10 0

ALT!TUDE (KILOMETERS) I

Fig. 17 - Examples of the vertical distributionof water vapor, data from Ref. 20


Table 4Atmospheric Window Definition

Window Number Wavelength Limits

(microns)

I 0.72 to 0.94

II 0.94 to 1.13

III 1.13 to 1.38

IV 1.38 to 1.90

V 1.90 to 2.70

VI 2).70 to 4.30

VII 4.30 to 6.00

VIII 6.0 to 15.00

iX 15.0 to 25.0

divided by the energy that would be received in the absence of any selective absorbers.The selective-window transmission then accounts for only selective absorption and doesnot take a, o consideration scdttering losses.

As part of the present program, an effcrt was made to obtain selective-windowtransmission as a function of water-vapor concentration, in order to extend the dataavailable in Ref. 22.

Experimental Re sults

Figures 18 through 24 show the selective-window transmission for windows II throughVIII plotted as a function of the precipitable water vapor in the path for a broad crosssection of the results obtained in this program. Window I has been omitted because of Atsvery slight dependence on water vapor concentration and window IX because it becomesessentially opaque through only 5 or 6 mm of water vapor. Figure 25 illustrates theapparent precipitable water in the Hawaiian path vs time as measured by the absolutehumidity on Mauna Loa and Mauna Kea. These values, which apply to the time duringwhich the data in Fig. 10 were being collected, emphasize the. need for continuo"s meas-urement during data runs.

In each of Figs. 18 through 23, the slope of the appropriate curve from Elder andStrong's report (22) has been reproduced as a dashed line. The points from the presentstudy have also been identified as to the individual path - 0.3, 5.5, 16.25, or 27.7 km -over which the result was obtained. The points for the different paths will be seen to fallon different straight lines, almost all of which i:le slightly below the lines extrapolatedfrom Ref. 22.


• "-z

70r

£ 8. "' ,>,PEF 2?

2n 0

Fig. 18 - Selective transmsission of +Window 11 (0.94 to 1.13 iz) as a funr- >

tion of precipitable water in the path -j IPRECIPITABLE WATER VAPOR IN PATH

a 0.3-KM PATHSEA LEVEL* 5-5-KM PAT14,SEALEVELO 16.25-KM PATH. SEA LEVEL+ 27.7-KM PATH, 10,000-FT. ELEVATION

so ,!o ,0

R --- RF 2 2ia70-

Fig. 19 Selective transmission of

Isop ____ ____ ____ ____ Window III (1.13 to 1.38 p) as a fune-d jt 1 0 tian of precipitable wate~r in the path

o 16.25-KM PATH, SEA LEVEL+ 2-7-KS PPTH,1O0OOO-FT. ELEVATION


80 - r

70 A1

0

0• ] • REF 22

Ii ½1Fig. 20 - Selective transmission oi

SWindw IV(1.38 to 1.90 g) as a func-fintion op water in the path

40 -0

,L 1[0 100PRECIPITASLE WATER VAPOR IN PATH

6 03-KM PATH.SEA LEVEL* 5.5-KM PATH, SEA .EVELo 1625- KM PATH, SEA LEVEL+ Ze" -KM PAT:., 0,0-00-FT ELEVATION

80z'x

70

S60--~------REF 22

Fig. 2 - Selective transmission ofWindow V (1.90 to 2.70 W) as a func- OL

tion of precipitable water in the path +N 20

01 [.0 10 100PRECIPITABLE WATER VAPOR IN PATH

A 0.3- KM PATH, SEA LEVEL* 55 -KM PATH, SEA LEVELO 16.25- KM PATH, SEA LEVEL+ 27-7 - KM PATH, 10,000-FT ELEVATION


s2 so

W 70

60.

2E 2'1 50

o40

Fig. 22- Selective transmission of 2 4Window VI (Z.7 to 4.3 ,u)as a function ',_of precipitable water in the path .

01I 1O00PPECIPITABLE WATFR VAPOR IN PATH

A ."3-KM PATHSEA LEVEL* 5,5-KM PATH, SEA LEVELo 16.25 -)KM PATH. SEA LEVEL+ 2'?.-KM PATH,,10,000-FT ELEVATION

6c

502 N

40

S30 -"' - .. e .2

0 N201

' Fig. 23 - Selective transmission of> 10o Window VII (4.3 to 6.0 a)as afunction

of precipitable water in the path

'-0 10 10PREC[PITABLE WATER VAPOR IN PATH

4 0.3-KM PATdiSEA LEVEL0 5-5-KM PATH, SEA LEVELo F6.25-KM PATH, SEA LEVEL+ 27.7-KM PATH, tO,000-FT ELEVATION


3°°Co I50-1

20o Fig. 24- Selective transmission of

+Window VIII (6.0 to 15 d)as afunctionWc p r€ •eripi a bl e- water in the path

> 10ki 01

0 01.0 10 100

PRECIPITABLE WATER VAPOR IN PATH

A 03-KM PATH,SEA LEVEL* 5.5-KM PATH, SEA LEVELO 16 25-KM PATH, SEA LEVEL+ 2717-KM PATH,I10,000-FT ELEVATION

I,.

20 1 1 I

16•. MAUNA LOA

914 -

12

1 r[ MAUNA 9EA--

Fig. 25 - Precipitabie water vapor in theHawaiian path as a function of time cornk (A.

puted from observatiohs at Mvauna Loa and 6 1 fll ,Mauna Kea. The se conditions prevailed 2M ME M O TEUduring~ -u L6(i.i0) TIME OF MEASURlEMENT FOR THEduring run ML-6 (Fig. 10). VARIOUS SPECTRAL WINDOWS

I0 RM. II I2 1A.M.

SEPT.6 SEPT. 7TIME


This separation of lines with path length indicates that the transmission of an atmos-phere containlng a fixed number of water molecules is not a function of the precipitablewater alone. There are two phenomena which would give rise to this type of etfect, andit is probable that both are present in these data.

The first phenomenon is that of pressure broadening, which results in an increasein the absorption of a fixed quantity of gaseous absorber as the total pressure of the gasincreases. In this phenomenon it happens that self-broadening, or the effect of absorbermolecules, is greater than foreign-gas broadening, or the effect of unlike molecules.Therefore, for a fixed total pressure and a fixed quantity of absorber, the absorption willincrease with the partial pressure of the absorber. This would be manifested as a higherwindow transmission when a fixed amount of absorber is spread out over a longer path.In Figs. 18 through 23, the short-path points should fall below the long-path points ifthis phenomenon is present.

The second phenomenon is the presence of absorption due to atmospheric constituentsother than water. The effect of these constituents, which generally have uniform concentra-tions and exist in any optical path in a quantity prnporttonal to path lengths, Is to producegreater absorption in the longer paths. For this effect, the long-path points in Figs. 18through 23 would lie below those of shorter paths.

Window 11, Fig. 18, whose selective window transmission is almost completelydependent upon water vapor (it contains one weak oxygen band), is seen to follow thepressure-dependence pattern, since the points for the 16.25-km path lie above those forthe 0.3- and 5.5-km paths. It is surprising to observe, however, that the points from the27.7-km path, which should show even higher transmission, because this lower pressurepath lies at an elevation of 10,000 it, lie more nearly on the 5.5-km curve. There is iioimmediate explanation for this anomaly.

Window III, which contains only one moderate oxygen line, should be similar toWindow U, but it shows a strong separation of lines characteristic of the presence of:.appreciable absorption due to constituents other than water vapor. There is no immediateexplanation for this.

Window IV, which embraces some weak absorption bands of C02 and CH 4 , showsa very pronounced separation of the lines - rather surprisingly large - and a moreshallow slope than the Elder and Strong curve, It is interesting to note, however, that asingle line drawn through all the points would have a slope s!milar to that from Elder'and Strong.

Window V contains a large number of moderately strong lines of C02 and some veryweak ones due to CH 4 and N20. It shows a pronounced separation between the 0.3-kmand 5.5-km paths but not between the 5.5-km and longer paths.

Window VI is bounded on one edge by a very strong C02 band and contains consider-able absorptidn due to weaker CO bands and moderately strong bands bf N2 O, CH 4, andCO. It shows the most pronounced separation of lines of all the windows except for thefact that the 27.7-km path falls on the .16.25-km path as it tends to do in almost all thewindows' It is probable that the effect of absorption by constituents other than watervapor is partially offset by the slight increase in transmission due to the reduced atmos-pheric pressure in this path, the two effects working against one another.

Window VII contains very strong CO 2 absorption and some very weak O and COabsorption. The line separation is quite pronounced, although between the longer paths itis not so obvious because the transmission is so low. This window is almost completelyopaque through 40 cm of water vapor.


Window VIII is bounded by a strong CO2 band and contains absorption by O, N2 0,and CH 4 but shows no tendency for separation of the paths. It is probable that it goes tozero too rapidly for the effect to be manifest with the accuracy of the present study.Over the 16.25-kin path (Fig. 8), through 38 cm of water vapor, no radiation could bemeasared from the source, indicating a transmission everywhere less than 0.5 percent Inthe window.

ACKNOWLED GMENTS

The authors wish to express their gratitude and appreciation to the many peoplewhose efforts have contributed either directly or indirectly to the success of this experi-mental program. To Dr. L.F. Drummeter, many thanks for patient guidance, and toMr. Thomas Cosden, our appreciation for his persistent and effective efforts to keep theequipment running whatever the time, location, or conditions. Mr. Lloyd Knestrick con-tributed to the experimental program and the data analysis, and Mr. Joseph Curclo filledin when the field team was reduced by illness.

'The operation in the mountains of Hawaii would hardly have been possible withoutthe generous assistance of the Hawaii National Guard and, in particular, Col. MichaelRoman and his unit. In the difficult tasks of moving and installing heavy equipment athigh altitudes and-in rough terrain, Major John D'Arujio and his men frequently exceededthe requirements of their normal dutius. The spirit and proficiency of th1is NationalGuard Unit are exemplary.

The cooperation of the Weather Bureau in freely making available the facilities oftheir new station on Mauna Loa materially aided this program. Mr. Roy Fox of theHonolulu office and Mr. Jack Pales, head of the Mauna Loa Observatory, did everythingwithin their capabilities to assist the program. Their efforts -are most gratefullyacknowledged.

RE FERENCES

1. Fowle, F.E., "Smithsonia-n Miscellaneous Collections," Vol. 68, No, 8 (1917)

2. Fischer, Heinz, "The Influence of the Spectral Transmission of Optics and theAtmosphere on the Sensitivity of Infrared Detectors," USAF Technical ReportF-TR-2104-ND, Oct. 1946

3. Strong, .J.D., 'Atmospheric Attenuation of Infrared Radiation," OSRD Report 5986,Nov. 30. 1945

4. Gebbie, H.A., Harding, W.R., Hilsum, C., Prycc-, A.W., and Roberts, V., "AtmosphericTransmission in the 1 to 14 Micron Region," Gr-eat Britain A.R.L./R.4/E-600, Dec.19.49, also Proc. Roy. Soc. A. 206:87 (1951)

5. Benesch, W., Elder, T., and Strong, J., "Absorption Spectrum of Cie Lower Atmos-phere from 2.5 to 4.0 Microns," Progress Report on ONR Contract N5-ori-16B95,Johns Hopkins U., Sept. 15, 1950

6. Migeotte, M., Neven, L., and Swensson, J., "The Solar Spectrum from 2.8 to 23.7Microns, Part I: Photometric Atlas, Part 11: Measures and Identifications,' U. ofLiege Final Reports Contract AF61 (514) - 432, 1957

7. Mohler, O.C., Pierop,`A.K., McMath, R.R., and Goldberg, L., "?.hotw•itriL' At'.aof the Near Infra-Red Solar Spectrum, k8, 465 to N25, 242,' U. of Michigan Press,Ann Arbor, 1950

8. Talbert, W.W., Templin, H.A., and Morrison; R.E., "Quantitative Solar ectralMeasurements at Mt. Chacaltaya (17,000 it)," paper presented at Eastern SvctlonIRIS, NOL, 5th meeting, Feb. 3, 1958

9. Benesch, W., Strong, u., and Benedict, W.S., "The Solar Spectrum from 3.3 to 4.3

Microns," Progress Reporf onONR Contract N5-ori-16605, Johns H|opkins U.,Aug. 1, 1950

10. Howardi J.N., Burch, D.L., and Williams, D., "Near Infrared Trankmliston tumn%- Synthetic Atmospheres," AFCRC-TR-55-213 Geophysical Research PaWprs No. 40.

Geophysics Research. Directorate, AFCRC, Air Research and Devel•pmeni Cwnma•• d,"Nov. 1955

11. Drummeter, L.F., Jr., "The Infrared Absorption of Water Vapur at 1.6 Micro",*

P Thesis, Johns Hopkins U., 1949

12. Taylor, J.., and Yates, H.W., "Atmospheric Transmission in thkn nrvrvd, MLReport 4759, July 2, 1956, also J. Opt. Soc. Am. 47(3):223 (1957)

13. Yates, H.W., "The Absorption Spectrum from 0.5 to 25 Microns of a IWOd- FiwiAtmospheric Path , at &a Level," NHRL Report 5033, Sepi. 1957

14. Goldstein, E., "The Measurement of Fluctuating Rdatlo•ai Components in the ty andAtmosphere - Part 1," NRL Report 3710, July 1950

45

A,

*46 wei NAVAiprr. S1L REERCH-u 41 LvABORATORY

15. Duwnl, A .. Mai zocai. J hI.C: i. PU.cl, ad WrT. , C., C Trawford, i t B. , ' I f '. it ,of nfrredPr~i Sp1ermlters, a J.Op. S'wdAt. 43(11:94 Am.~ '61:)

16. ns~tT.HWRordin.\ du4rlml ArrnspJ.l Phtrsziss2,4120 ul IA~;

ax.JDa, .t Retw uýir 457 .. Jur,' 1955Cttar .JMe tm

at. TAkeu~a L, mrzuia Atteuation4 Put Ultran-v jlAan ilbe :

In.fltA. iriars tdi.. (Arj tug UarW4 V-rtt 4". 1~aii~ssu F

-.3

IA 14 *...±R,...-..~ lt~zrrowt

Naval Research LaboratoryTechnical Library

Research Reports Section

DATE: October 24, 2000

FROM: Mary Templeman, Code 5227

TO: Code 5600 Dr Giallorenzi

CC: Tina Smallwood, Code 1221.1 / A"/'SUBJ: Review of NRL Reports

Dear Sir/Madam:1. Ple re ew NRL Report 5453 for:

Possible Distribution Statement

EI Possible Change in Classification

Thank you,

Mary Templeman(202)767-3425maryto~library.nrl.navy.mil

The subject report can be:

Changed to Distribution A (Unlimited)

["] Changed to Classification

EI Other:

Signature h lte

unclassified ad number · ed befense ... parabola designed for optimum use with the perkin-elmer...

Documents