"* UNCIISSIFIED I fLE C PSECURITY CLASSIFICATION OF THIS PAGE

j)Tf4(3CUMENTAM1N PAGEla. REPORT SECURITY CLASSIFICA'I• Eb. RESTRICTIVE MARKINGS

UNCLASSIFIED-- - ELECTE2a SECURITY CLASSIFICATION AU MAY 2 4 3. DISTRISUTON/AVAILABITY OF REPORT

Approved for public release; distribution Is unlimited.

AD-A 193 683 9 - . MONITORING ORGANIZATION REPORTN

6a. NAM U- - .......... - 6b. OFFICE SYMBOL 7a, NAME OF MONITORING ORGANIZATION

Naval Ocean Systems Center NOSC Naval Ocean Systems Center

So. ADDRESS (,SW#5P1Cod.) 7b. ADDRESS (ftShk&d1C0*d)

San Diego, California 92152-5000 San Diego, California 92152-5000

Sa. NAME C- FUNDtNQ/SPONSOFING ORG3ANIZATLON 18b. OFFICE SYMBOL 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER

" Naval Ocean Systems Center NOSC -

Bo. ADDRESS (SobafZ•C4•qj 10. SOUR'-C OF FUNPIN NUMBERS"""PROGRAM ELEMENT NO.1 " CT NO. TASK NO. AGENCYI ~ACCES•iON NO.

San Diego, Califonria 92152-5000 6275I'N SXB3 RW59S-1 DN888 715

11. TITLE (k•.*dS .i)

IR.SKY RADIANCE DISTRIBUTIONS IN TH.F MARINE BOUNDARY LAYER

I1C. IRSO:NAL AUTHOR(S)

DB. Law, et Al.13a.TYPE O R PORT ab T COVERED - -7* 5 PATTE 0! REPOiRT acW M~0' ¶.PG COUNT

PrcOIeSslOrmj 1A T -1FROM Apt 198'- TO Apr 1986 I April 1988as, SLPPLEMENTARV NOTATION

IF. 00SX - coveg i T -~ej Ent Ps-FIEL-O GROUP SUB-OROUP eleclromavnetci pboelc physics

___eltetro-optlcs

counletmrstu•ots

19, WT4II ACT Ca w .y.fo5v . .

• Infrated optical prSpertles of the mtarie boundary layer are baslc the performance of thermal Itmtaglng systems, such U forwardlooktin Infrared (j•iu ensors, over the oclan~lro eld IP e alualln$ ila ptrformance of these smnoors, spatial distributions of Infraredsky radiance in the 3-Sja mid-wavelength nfirated (4Wft41Jhd 8-WI lonI wavelength infrared (AAMNWtspectral bands were meas-uyetd sh'nulfisniously t flw *pVatlon ongles abovey the tea surface. CaLlibhatd AQA, Model 780, dual scanning systems functioned asImaging Infrared radlomelerth1frared Aky radiance avid meteovoiogical pifameters were recorded concurrenlty in a series of four dataset.durl~l 0 £.Ltttit•.J~yCltiarIng 15 April 1986 At 1500 Pacific Standard Time (PST) and unilng 16 April 1986 at 1)70 PST,

-. IRadlosondos were reltased from ,,oma,.bout 7.6 m. aba4utfat-oce njt a ,range of 5 km due west of thecoastal sensor site at Naval Ocean Systems CenteriA San Dieglo, CAgiM~nd speed, dhtocllon, iiE!1mperature, and cloud conditionswcre also recotded on board the 3hip, Sequential Images of radiance distributions provided control dalt for monitoring the stability orvariability of atmosphetic conditions th•oughout the time for radiosonde ascent to about 6 km altitude,

Measured IR sky radifiAce distributions were compared with corresponding cleariky radla•ce using the LOWTRAN 6 computercode. Cloud radiance and stattered olim fadilation restirted the compartioa to ieavatloas close to the optical horiton where aerosolatituuatlon would be greatest.

jnfrared aerosoJtfan•mhtance was Infernd from the ratio of m*2sured radiance to calculated cleat-sky radiance along the horftontine .. lght (L983)`btulvatent temperatures (or blackbody raditatce at the horizon were either less than or equal to the ambient airtemperature near the sea surface. except when the MWI band included racatteed solar radlation; cnaumqueaiiy, oWy the LWIR b&W,ooid be used to inter aesool transmittalnce reliably. ,- ....

(Ccatinued.)

Published In Proceedings of S. P.1.E. -The interna•ional Society for Optical Engineering Wave Propagation and Scattering in Varied

20. WtPARlUTaKtNAVAtAIJMuTY OF ABSTRACT 121, ABST"TRaSYSCU~I iV-7A8¶SW1AT10N

[] FWOrUAU&tiu0 mO E [MS• aS 'SPT oJ rTIC u:Lt3S Ur.CLASSIFIED

22s. N~AM W RSO NS6 A S E tQXM af TIELEP44ON 0kd4*4."l 22. OFFICS SISAOM

D.P. Law.. (619) $53-2630 Code 754

F3 A EDO MA u USE. Wir,. = D4UESTE UNCLUSSI lE-OD FORM 1473, 84 JAN ALL 0TE WT ARE OBSOLETe tC5T S8bATOOPT AJ

SECURITY CLASSIFICATION OF THIS PAGE (MMai DiEMws)

19. ABSTRACT (Co&Mwd

I Radiance along the optical horizon originated mainly in the lowest ;00 m of the atmosphere; therefore, reasonably accu-rate horizon radiance or transmittance predictions could be made from tneteorological data within this low altitude. Theseresults indicate that a LWIR aerosol transmissometer could be developed by computing the ratio of measured horizon sky radi-ance to calculated clear-sky radiance using local ambient meteorological data.

!I

Aoonaslon ?or

DTIC TA 0iLUiwounooe

D15trbibutiou/

Availablltty Codes

Met Speciaol

DD FORM 1473, 84 JAN UNCLASSIFIEDSECURMYh CLA5SSICATION OF T"4S PAGE (W%*VW(#jWIj

IR sky radiance distributions in the marine boundary layer

William J. Schade

Alcoa Defense Systems, Inc.16761 Via Del Campo Court

San Diego, CA 92128

David B. Law

Naval Ocean Systems CenterSan Diego, CA 92152-5000

ABSTRACI

Infrared optical properties of the marine boundary layer are bi€c in the performance of thermal imaging systems, suchas forward looking infrared (FLIR) sensors, over the ocean. To aid in evaluating the performance of these sensors, spatialdistributions of infrared sky radiance in the 3- 5 pm mid-wavelength infrared (MWIR) and 8 - 12 Am long wavelengthinfrared (LWIR) spectral bands were measured simultaneously at low elevation anglcs above the sea surface. CalibratedAGA, Model 780, dual scanning systems functioned as imaging infrared radiometers. Infrared sky radiance andmeteorological parameters were recorded concurrently in a series of four data sets during one diurnal cycle starting 15 April1986 at 1500 Pacific Standard Time (PST) and ending 16 April 1986 at 1730 PST. Radiosondes were released from thedeck of the USS POINT LOMA, about 7.6 m above the ocean, at a range of 5 km due west of the coastal sensor site atNaval Oceat Systems Center, San Diego, CA. Wind speed, direction, sea temperature, and cloud conditions were alsorecorded on board the ship. Sequential images of radiance distributions provided control data for monitoring the stability orvariability of atmospheric conditions throughout the time for radiosonde ascent to about 6 kn altitude.

Measured IR sky radiance distributions were compared with corresponding clear-sky radiance using the LOWTRAN 6computer code. Cloud radiance and scattered solar radiation restricted the comparison to elevations close to tie opticalhorizon where aerosol attenuation would be greatest.

Infrared aerosol transmittance was Inferred from the ratio of mnasured radiance to calculated clear-sky radiance along tliahorirzon line of sight (LOS). Equivalent temperatures for blackbody rwdIance at the horizon were either loss than or equal tothe ambient air temperature near the sea surface, except when the MWIR band ineluled scattred solar radiation;consequently, only the LWIR baId could be used to Infr aerosol transmittmrc reliably.

Radiance along the optical horizon originated inainly in the lowest 100 m or the atmosphe; therefor, reasonablyaccurate horizon mdiance or transmilttance predictions could be madn (mrm meworologal data within this low alltt.•e.11hese results indicate that a LWIR waosol tranrsmisometer could be developed by computi W rato of u•.eaukr hoL.t.nsky radluce to calculatel clr-sky radianca uing local ambient ametaological

1. INTRODIUCTION

The effects ofaerosols on the propagation of infrared radmti3ui through the marine boundary layer ae basic in the use ofboth active and passive remote scnsing systens over the ocean. The extent to which these effects can be evaluated andmodeled depends largely on tho methods for characterizing the atmospheric consritueats along tho optical path.

Current methods for measuring particle sizes and concentrations are limited to local sampling vohmrues at the senorsite which arm not characteristic of extended path lengths. Lidar menaurements have been used to compute til volumeextinction coefficients as a function of range; however. lidar datm are generally acquired atseoquential intervals along diffecrntdirections between which the meteorological conditions could clhage without notice, Additikl un taiutiendes uris whenlidar data at one wavelength are applied t otho spectral regions.

An altemative method lbr inferring the effects of aerosols on atmospheric transmittance 13 describe in this paper.Srmaal disLributions of Infrared sky iadiance in dIe 3- to S-irm mk-wavcleogth infrared (MWIR) and 8- to 12*jim loWg

wavelength infrared (LWIR) spcCtral bands wcre measured at low clevation angles above the sea surface. These distributionsrepresent the integrated radiance propagated along contiguous optical paths through the entire atmosphere within the field ofview (FOV) of the calibrated thermal imaging sensor. Since the FOV of the AGA Thcrmovision 780 is scanned in 40 ms,the data for each distribution arc almost simultaneous.

The extent to which the measured radiance differs from the corresponding clear-sky radiance can be used to imply theeffective transmittance. The clear-sky radiance was computed with the LOWTRAN 6 computer code using radiosonde:measurements of meteorological parameters concurrent with the measured radiance over the ocean. 1 This technique offers aviable method for evaluating the total aerosol effect through the entire marine boundary layer by remote sensing.

The data were obtained during one diurnal cycle starting 15 April 1986 at 1500 Pacific Standard Time (PST) and ending16 April 1986 at 1730 PST. The sensor LOS was directed along an azimuth of 270 degrees true radial from NOSC,building T323, located at 32.699 degrees N latitude, 117.253 degrees W longitude, at a height of 33 m above sea level.

Radiosondes were released from the deck of the USS POINT LOMA (AGDS 2) about 7.6 m above the sea surface at arange of 5 kin from the sensor site. Wind speed, direction, and sea temperatures were also recorded on board the ship, aswell as cloud conditions. A complete description of the shipboard data is available.2

Throughout this report, clear-sky radiance refers to the atmosphere without aerosols. This corresponds to IHAZE = 0in the LOWTRAN 6 computer code when IS = I in the LOWTRAN calculations for sky radiance (where IS is transmittancedue to scattering). This is regarded as the reference condition for comparing the respective measurcements in the LWIR andMWLR spectral bands.

1. ATMOSPHERIC MEASUREMENTS

Infrared sky radiance and meteorological measurements were recorded simultaneously in a series of four data sets, TableI lisq the dates and dimes for the measurements along with the solar positions. Surface meteorological measurements weremade at the NOSC sensor site as well as shipboard, while the vertical profiles were measured by radlosondes released tomnthe USS POINT LOMA. The air and sea tempertuares, wind speed, and direction are also liUsd in table 1.

Table 1. Autospheric measurements series for meteorological parameters.

Dan -Set 1 2 3 4

Date 15 Apr 86 16 Apr 86 16 Apr 86 16 Apr 86Radlosondo Time (PST) 1945 0845 1245 1645Solut Position (degrees)

AMdmuth 295 106 212 .70Zenith 140 49 26 71

Air Temp. (O") 15.0 15.4 15.s 15.7Wind SIped (knou) 10 9 20 20Dirzmcon (degres) 300 330 280 310

Sea Temp, (C) . 17,8 16,7 172 17.2

Figures 1, 2, 3, and 4 shuw dte rudiosonde measurements for temperature and relative humidity plotM with respect toaltitude. These dm- wiro used in the LOWTMAN 6 computer code to calculte clar-sky nubuince as a function of clevationangle above the horizon. The data in figure I were used Uo evaluate the meteorlogical scnshividt", of dih radiancecalcultidons. This set presnts thd nighttime data fat a wellkdined cloudl.wtp boudasy ad go adiabailc lapse rate of9.85Cii the rfut kilometer above sea level.

The uncertainty in rotative humidity Is estimated to be S percent, which provides a ber W apptoximaion to the watervapor density than any of the standard moe atmospheres. The sensitivity of the radiance calculations to this uncetiainty Inhumidity was examined by Incremcntlng the relative humidity profile :5 percent whiie keeping the temperaztw profileunchanged from the measurcd values. Similarly, the se, itivity to thL Icaiperutur uac.ily o ±0,.5 T was evalnated fordie observed elative humldity pmrd fotf igure 1.

8f 8 F

15 April 1986 16 April 1986T RH 1946 PST T RH 08465 PST

6 - - - ---

2 2 ------- ------

2 '2.- - - - -- - -- - --

01 0 1 -1_RH 0 20 40 . 60 80 100% RH 0 20 40 60 80 100%

-20 -t0 0 10 20 300 C T -20 -10 0 t0 20 30 OC

TEMPERATURE. T TEMPERATURE, TRELATIVE HUMIDITY, RH RELATIVE HUMIDITY, RH

Figure 1. Radiosondo mcasuremns for data set 1. Figure 2. Radlkoa mcasurcmnts for data set 2.

* 8•

16 Aptd 1986 RH 16 Afid 1986T RH 1;46 PST T 164 -p T

I.. ..

4. ' .

",,,dhr '.;,,,4. .,.

-----------

RH 0 20 40 60 t0 1o% RN 0 20 40 60 so 100%T -20 -10 0 to 20 30 O ? -2t0 -10 0 tO .20 0c

TEMPIRATURE 1 TEMPERATURE. tRELATIVE UMIOJTY. RIB RELA4IVE 1m4UarM1 , 0R

figure 3. Radlosonem~c amwemon (or dat saL 3. fisuw~4. Rad~osoiadomcA!4mt=cnis ord~asd4,

Figure 5 shows the comparative effects of deviations in temperature and relative humidity. The calculated clear-skyradiance, for MHAZE = 0, is plotted as a function of elevation angle. Radiance discrepancies resulting from relative humidityerrors are less than those resulting from temperature errors throughout the range of elevation trom 0 to 2 degrees. Forelevation less than 0.5 degr.ze, the effects of errors in relative humidity are negligible compared with those in temperature.A temperature error of ±O0.5*C changes the radiance by about ±0.033 mwl(csn2 - sir) in the LWIR band and about ±0.0033mw/(cni2 - sir) in the MWIR band.

3.14

S311

2.99- 0 R14, T '0.5 OC

0 0.2 0.4 0.8 0.8 1O 1.2 C4 t'd i's 2.0

6LUVACION ANGOLE (dog)

Figure 5. C~alculawd C, - jk tbutieas of LWIR radiancec above the horizon forcomparotive errors in tcmpe ate (T ± 0.50C) awl feahtive, huridity (RHR M 0)

Sky railiative data were acquired In the 3.5- to 5.S-I~m and 8- to 124Am spectral bands siniultancously With' the AGA780 dual scanning systems functioning as imaging radiometers. The operational characeteristics of the sen"o systemis weclisted In table 2I Thw anlog video signals In fte 3- by 3'dogree FOV ame Interfaced through a flataInk anatog-to-digiwta,converner wo a digital inioge processo. The fomat of the digital display is 128 by 128 pixels with $-bit resolotton. Ttk,ruinhmmn dtctccb~cte mpeituree diffeucnce (Mt)TD) between 4djacen kimsmaneous field 0(~w (IEOVs) L% O.11C.

7the total system response rot ecub spectra band is cal ibrated in isotherms which represet fte vide sgnlt piuefor radiation fromn a blackbody Ut 2 measured temperaure and close iunge. Sets of convesion cuarves zteatiig isoiheMa 110objact tempirature (with cmissivity *a 1) awc obtalaed with th exptession

I t C II(CS - xp (C2fr -1). 1where T is the blackbody temperature in degrees Kelvin. Tte coa~tanz C1. C2. and C3 ate dcaivc for the calibrtiosource at SOveaUl derent teMVheUawL

Table 2. AGA Thcrmovision 780 dual scanning systems operational characteristics.

Scanner SWIR LWIR Both

Spectral Band: lni 3.5 to 5.5 8 to 12Detector I each InSb HgCdTeTemperature: *K 77FOV (az by el): degrees 3 by 3IFOV (ax by cl): mrad 1.2 by 0.9

Telescope ObjectiveDiameter: cm 10.6Focal Ratio 1.8

FOV Scan ModeHoriz. Line Time: ms 0.4Vert. Field Time: ms 40

Field Rate: Itz 25Frame Rate: Hz 6.25

IFOV dwell: ps 5.3Signal Processing: AC + DC Restorcd

MDTD: *C/IFOV 0.1Temp. Range: .C .20 to 1000

The original image is analyzed with the microcomputer by assigning the equivalent blackbody temperature to thecorresponding isotherm value for slekted pixels in the FOV, tirough the tramposhion of equation 1,

T C/In u, (2)whom

u= ((C 1i1) + W)C3.

Rudiance values awe computed for each blackbody temperature through the Integration of PLack's equation rot theappropriute spectrul band. 1U real fIm of tho cquation for raldlanc as a functio of i.uahens Is

L(S'T) m a . I()- h. (3)where SB is the spectral band, a is dte slope, b is the intercept. wid I(M) is a nonlinear (unction oft in equation 1. Thiconstants for converting isothemis to radiaxce In cqualwlo 3 are listed in tLtib 3 for each •s uuI band, as well as thecalibration source 'hwarisucs.

Table 3. Calibration constaits for converting Isothents to rdiantcz.

Scaner " SWIR -. . LIR Both

CI: Lsothenns 383140.4 1514.4

C2: OK 2945.0 1294.2

C3: dimensionless 1.0 0.503a: mw(cm 2 Str sr- iso im) 9,695. 10"6 1.002 104b: mwlcm 2 Ur) 1.556 •105 1.002. 1O4

Calibration Sou=cDismance: 01 3.0

Emissivity: ±0.01 0.98Temrpemture: -*0,!C ... .... .. _ __________I-t1O1O

The main source of unccrtainty in the conversion to radiance is in the isltherm values Wor the video signal amplitudeS.Isuhcunis am calult=d W1 the mi a~omputc to within A0.1. so a the empeatum of t Ca€lwi source is

reproduced within ±0.20 C. This uncertainty in the temperature causes a radiance error of:±0.01 mw/(cm2 - str) in the LWIRband, and ±0.001 mw/(cm2 -str) in the SWIR band.

In each of the four data sets, both infrared sensor systems were referenced to a calibration source at ambient airtemperature within 1 hour prior to the radiosonde ascent. Thermal images of sky radiance were recorded continuously onanalog video tape and sequentially on digital diskette during 1 hour preceding and following the radiosonde measurements.These data provide an unambiguous verification of either the stability or variability of the sky radiance distributions duringmeteorological measurements. No measurable changes were observed during each of the data sets reported here, except fortemporal variations in cloudiness.

3. RADIAINCEF DISTRIBUTIONS

Figure 6 shows the four LWIR radiance distributions as a function of elevation angle above the optical horizon. Zerodegrees elevation corresponds to the IFOV tangent to the sea surface where the radiance is maximum. The elevation IFOVangular subtense of 0.05 degree is shown in the figure as well as the radiance uncertainty corresponding to a temperaturedeviation of TD = ±0.2?C.

T tC)

5 1 - 1170

320 1- 16.0

3 15 LWIlR SKY RAO (8.12 bm) - 15.0"• 310 - 14.2

3.05 13.23.00 2132.95 11.4

6 .24 104

,'2a5 94

276 -7A8

0.2a 0,4 0 6 04 10 12 14 1.8 1.4 2.0ELEVATION ANGLEI(dgj

l;g-ure 6, LWIR sky radince di~tributhmnsabove the horizon-. plot 1: 15 April $6. 1945PST-.phn2: 16 April86.0845 PST-.pknt3: I6April86.1245.PST~pio&4: 16April86. 1645 PST.

Cloud radiance contributions 010 appartnt at elevations gteaxet t 0.4 degee S d solar radiatio rrom clouds iscspecia•iy notable in daa sot 4. when the sun is abou •20 degrees above the sensor LOS.

The radiunces fur data sets 3 and 4 axe equal within dte mcasucment uncertainty, (oc low elevations bitween 0 and 0.3degree: however, the solar zenith angle changed froi 26 degrees in dAta set 3 to 71 dcgrc in dta set 4. These con"dionsimply that arosol scattering or solar radiation is negligible in the LWIR band s de mtnemeoogical par et aremerly the same for these two dat s as shown in figures 3 ard 4.

Figure 7 shows the MWIR sky radiance disributions. These four plots ar= concunrent with those of the LWI data.On the radiance =calc used in figure 7. plot.s 1, 2, and 3 am approximately equal, within d.-a wmperature deviadon TID±0.2*C. Plot number 4 shows a very large increase in radiance resulting from scattered solar radiation in the clouds andlower Atmospheric levels. The equivalen badbody twpcwube scales ae shown Alm& ft ordinates on the right of rig=6 and 7.

T (C)

0.21

0.20 4 - 28.8

0.19 27.3

018 - 25.8

0.17 MWIR SKY RAD: (3-5 Aro) - 2420.16 -

UZ 0.16 - Tom 0.2 6C. zz IFOV: H4 0.14 -- "

013 - IA.2030- 171

012- - 15.1

011 12.90 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

ELEVATION ANGLE ("gM

Figure 7. MWIR sky rWi4=c di4tributions above ate horizon and coincident with theLWIR distributions.

Figures 8 and 9 compare the observed radiance distributions and the corresponding calculated clcar-iky rudiances.Figures 8. b. c. and d show ft dlsutibutions for the LWIR band, and figures 9a, b, c and i show Otos for the MWR band,

Ilia clear-sky radiances were calculated using atmosphcric models with 30 layers in the LOWTRAN 6 computer code.The meteorological enties were provided by die radiosondo data. The concentrat•os of uniformly mixed gases and ozonewere those In thde 12 U.S. Standud model amtosphere.

The importance of the humidity and temperature lape rates in caculuting sky radiance distributions in the boundarylayer was noted prcvlously? Accordingly. th,, first kilometer above e level was divided Into 10 levels with about l C-incremnents for cacti 100 rn between level•. 11m remaining aunosphwe was divid.-d into 20 levels betwee I aMd 5.5 km.TIh radiosonde data terminated at about 6 kin.

The LOWMAN experssioin for calculating atmosphcr radiance (Kneizys ot a.. 1980) is

N-1!

+ 8 (SO. T (b)) 1h. (4)wher the s•munmaion along a specified LOS is over (N - I) uidwr yallrs at tacmperaum T (i). and

SB = spectral band: 3 to SWnor8to 12pmla = band avcrage transmitonce due to absoqukul,

= bid average transmittance due to single scattering,

ha bnd average tnansuiaLxnce from scnw to cloud boundary.D (SUMT) PLnk blackbody fuAcdwi ora spectral band and average impe = Totan atmospheic

layer.

To avoid the effects of cloud radiance, the compaison of observed-to.-calcdtked rdiance is restricted to the horizon LOSwhere the atmosphcric optical depth and aerosol attenuation are greatest. Along zero etavmion, tM LWIR motecuta4r(absoption) transmittancc, M, Is less than 1 x 104 through the 137- km hoifr-cnial optical path to the cloud base. With acloud-basc weinpcrauro of 6.8C. tf nasimium LWIR cloud radianc is 2"7 mw/(cm2 - su), which becomeS 'elig.ible a1th ca sens rtmjew.

31 3 20

36L491Api86316 LWIR tfiApoilill314 301245 PST

312~ 3054310 -LOWMAAN UWAE 0

E 308 -E 295 SKSKYS

4( 304 395

302 .42 70

2so-2 700 02 08 06 08 to 12 14 16 is 20 0 02 04 06 08 10 12 14 IS 1G-20

ELEVATIO ANGLE (dig) ELEVATION Ah401. 6 Wooa. Data set I c. Data Set3

3 '.o 2318 LWIS 1641A " 1iM l5vd6

S316 00445PST 1 64 S

EWý 3073 12 LOWIRAIN OU4AU '030LOMNWE.

E310SK

S30630.1 SKY 07

302 ;V

300 27

0 02 Oil 06 00 10 13 14 16 id 20 a 02 04 04 a$ to 12 14 I. to 20

b. Goa set 2 d. Data sol4

Figure 8, Compaiison of moasuucd LWIR sky radlno= distributions with c*ULtcuzc ckýsy r.Alion. (tamuLOWTRAN 6: ULAZE 0.

For calculation of ctcar-sky radlincc. L-, I (at aJI aumosphecnc laytrs. Along the hotizon LOS. the sky radlaticeoriginates piucdominarntly in the lowest toyer which aplvoxima mu a horizonatly horn-ogeneous mmnosjkhre about 100 inh141h, Ibis comlition Is clealy evidcol in tnble 4 which list% the LWIR todimanc Wn band vcaebsptrwin tlasittano.Ia. (or ranges along the horiwon LOS to five levzds in iho r~st nift anospheodc layers. Over 99 percent or the ctceir-skyWsadiwce along fti hoii~on LOS is awdtriutblo to the lowcsa 100 m of the =a4:ilw=xc W"ti this lowest bap. equaiioa 4rtwiuicesto

From table 4. Ia (1) r. .004, so the cinissivity or the lowest taycr. e (1) mt - Mn (1). is Very close to unity throughapi~laso o Kichfi~ lw. hu, te ~c~~sy Maianec in fti lowest layer should equal the blackbody rdicet

ambienit air tenipcralurc as repomtd previously.A-Vi heffects of werosals on inkrarcd sky ridiancc can be inrcrtcd fromcqiwtions 4 =nd 5 by including fth trw smitimcc duc to scatteriag. The musauitmid rudian W0%,t tho hori"o Los the

wk L (SB) is fthoclar-sky raod cgivcni by equ ati Scx dcl to all 30 kayms

01190 •i•1O 20 1 1 1 '

"U WiR: 15 ArU 86 Oils 0 -MWLR: 16ApUO86-0.1614 S 1245 PSTSK 0 118 --

U U01133 01l4 SKY

E E01100 - 0112

U~ tOLOWTIAAUWEZ0 2

c€ 0 top, -- 0108 -

01,0.o 01o60 02 04 06 08 tO 12 14 16 i8 30 0 02 04 06 0a 10 12 14 16 18 20

ELEVTiON ANQA1E Ideq) ELEVATON~ A*NGIE fdg)L. Oa3se01 c. Daaset 3

oz v A Ii 1 I" 0205

1UMR: 1166

0116 SKY

0 02 0. 08 01 10 13 1d I, tl 20 00 4O 1s •t i

EL8VA1I0N M4,NW tid |ILVAIi,23I AM~8Idegb. Oa1a12 d. OgasI4

Figure 9. Canomp of ncisurcd •W• -sky wI, L di•uAi•os with CAkUCd Ckr.sU-ky rMiA4 G(M

LOWTRAN 6 A -0.

Taie 4. LWIR mdtLLm And tinsmiu•t.nce 41og th bmize- LOS ro" dat set 1. LOWTRAN 6 MtIAZE 0

w wp to tp Rzvtc % Ntux A•I.V'•(ml ,m mwkin.2 -s.r• • T•.nimilu.•nc.

143 67 3.163 99.7 0.0036

263 84 3.171 99.9 0.0011

3 383 97 3.172 99.9 0.00045 575 114 3.173 100 0o0o019 1016 145 3.173 100 0.00,

Absorption by aerosols would tend to bring thc emissivity of the atmosphere closer to unity through ia. which wouldchange by less than 1 percent from the transmittance due to molecular absorption for clear-sky conditions. Molecularscattering, however, is negligible in the spectral range of 3 to 12 tIn, so ja can be assigned primarily to aerosols. Thus, thescattering transmittance. 1s was calculated with equation 6 using the measured sky radiance for L (SB)'. which rcpresenLs thetransmitted radiance along the optical LOS fron the sae receiving aperture to space.

A summary of the horizon sky temperatures and aerosol transmiuance is presented in table 5 for each set of radiosondedat The temperatures corresponding to the measured horizon sky radiance and calculated clear-sky radiance are listed for theLWIR and SWIR bands. Tiw radiosonde base temperatures and sensor (local) air temperatures are also given. Within cachdata set, the temperature differences are less than the uncertainties in the measurements except for the equivalenttemperatures of the measured sky radiance. Reduced equivalent temperatures in the LWIR band correspond to aerosolscattering loss in radiance, while the increased temperature in the MWIR indica additional radiance from scattered solarradiation in dam set4.

The calculated aerosol transmitances for both infrared bands are listed in table S. The LWIR values show a slightaerosol effect in data sets 3 and 4. but the MWIR values arc inconsistenL Although the values for both infrared bands agreein the nighttime and morning runs when the aerosols were negligible, the MWIR band cannot be used reliably to indicateaerosol attenuation when scattered solar radiation must te included. As noted above, the horizon LWIf values wereunaffected by solar scattcring for zenith angles from 26 to 71 degree

Table 5. HoriTon sky temperature and aerosol transmittance.Data Set 1 2 3 4

Date 15 Apr 86 16 Apr 86 16 Apr 86 16 Apr 86Tfte (PSI) 1945 0845 1245 1645RAOB V=s (±O.50C) 15.0 15.4 15.5 15.7Local Air (+0.3 0C) 14.6 15.2 15.1 15.8LOWT•AN 6: Cler-Sky

MWIR 14.6 15.1 151 15.2LWiR .15.4 15.9 15.9 16.1

Ni,-•sed Sky (±01"Q)MWIR 14.5 14,2 145 23.7LWIR 14.7 14.2 13.1 13.4

Aerosol Teasmu=WccMW[R .97 .98 1.38, .W R . .. .. 97 .... .95

4. -ONCLUISIONS

The AGA Thetmvision 7130 dual scanning system wit used to mrasure infrared sky nrad distributions in theMWIR and LWIR spectlal bawdU sltmultaneously. Sequtcatial images of radaet d4s1tibutiors pmvidcd CO•trol da=a ftm04aa4i4ing t•h abiliy Or vaiabilty of aU rc& condiions thiou- the coWuUMc dtie for nWIOSd ascenL

Mcasured infraxcd sky tadiace distibutions; were compared with the Corresponding clear-sky rndimce using theLOWTRAN 6 computer code. The ptesence of cloud radi=nce and scatered sWr tadi• •.,cictc•b. compuson toCle'vations close to fth horuimo Wthmr arosolW atznauion would be grc=4cL

=h W ,fraced aero sol trwm anstitrce was Wli tfrd oni fth. tatio of mecasured rd iacd to c.*cu Late cleAr- sky radiahcc awogtehonon LOS, The euivalcnt wnpcaturts for blackbody radiac at the hocizon were either less tW or cqul to theAmbichta ir temperature near the sea surface, except when tf MWIR band included s.c.jWed solar tadintion. waus-this, only th LWIR b.nd ca= be used, o infer aerosol tuasitta= tclaNy.

"The rwdiante along the horizon LOS originas predminm•dy In the Iowat 100 m at the aunmbspre,; teretoe,.asonably accumi horizon radiance or uanSmiuancc predictions could be made from meowwtoogical data within this low

altudVe. Th coclusion sqggesi tha a LWtk aetosig trami"somesr could W, developed by comp•ting the rao d the

measured horizon sky radiance to the calculated clear-sky radiance using the local ambient meteorological data. Such apoint-measuring system would only require a calibrated nonimaging radiometer to record the radiance along the horizonLOS.

5. ACKNOWLEDGMENTS

The authors gratefully acknowledge the radiosonde data provided by H.G. Hughes and R.A. Paulus.

6. REFERENCES

1. Kneizys, F.X., E.P. Shettle, W.O. Gallery, J.H. Chetwynd, Jr., L.W. Abreu, J.E.A. Selby, R.W. Fenn, and R.A.McClatchey (1983). "Atmospheric Transmittance/Radiance: Computer Code LOWTRAN 6," Air Force GeophysicsLaboratory AFGL-TR-83.0187.2. Jensen, D.R. (1986). "Temperature Measurements of the USS Point Loma for FUR Evaluation Studies," Naval OceanSystems Center TD 971.3. Ben-Shalom, A., A.D. Devir, and S.G. Lipson (1983). Qt.gL Eng. 22:460.4. Bennett, H.E., J.M. Bennett, and M. R. Nagel (1960). "Distribution of Infrared Radiance over a Clear-Sky," LQnLSo.Am, 50(2):100-106.

5. Bell, E.E., L. Eisner, J. Young, and R.A. Oetjen (1960). "Spectral Radiance of Sky and Terrain at Wavelengthsbetween I and 20 Microns. II. Sky Measurements," OtScAm. 50 (12):1313-1320.6. Kneizys, F.X., E.P. Shettle, W.O. Gallery, J.H. Chetwynd, Jr., L.W. Abreu, J.E.A. Selby, R.W. Fenn, and R.A.McClatchey (1980). "Atmospheric TransmittancelRadiance: Computer Code LOWTRAN 5," Air Forc GeophysicsLaboratory AFGL-TR-80-0067.