8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 1/55
Army Research Laboratory
Weather Effects on Target Acquisition
Part I : Sensor Performance Model
Infrared Algorithms
by
Richard C. Shirkey
Barbara J. Sauter
Computational and Information Sciences Directorate
Battlefield Environment Division
Rene V . Cormier
U.S. Air Force Research Laboratory
(Dynamics Research Corporation)
ARL-TR-821 uly 2001
Approved for public release; distribution unlimited.
2 0 0 1 0 8 0 2 3 4
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 2/55
NOTICES
Disclaimers
The indings n his eport re ot o e onstrued s n fficial Department f he Army osition, nless o esignated y ther
authorized documents.
Citation of manufacturers' or trade names does not constitute an official
endorsement or approval of the use thereof.
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 3/55
REPORT DOCUMENTATION PAGE Form Approved OMB No 0704 0188
Public reporting burden for this collection of information is estimated to average our per response, including the time for reviewing instructions, searching existing data sources,
gathering andmaintaining the data needed, and completing and viewing the collection information. end comments regarding this burden estimate or any other aspect of this collection of
information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway,
Suite 1204, Arlington, V A 22202-4302, and to the Office of Management and Budget, Paperwork Reduction (0704-0188), Washington, DC_20503_
1. AGENCY US E ONLY Leave blank) 2. EPORT DATE July 2001
3. EPORT TYPE AN D DA TES COVERED FINAL
4. TITLE AN D SUBT ITLE Weather ffects n arget Acquisition art : Sensor erformance Model nfrared
Algorithms
6. AUTHOR(S) Richard C. Shirkey
Barbara J. Sauter Rene V. Cormier, U.S. A ir Force Research Laboratory, Dynamics Research Corporation
7. PERFORMING ORGANIZATION NAME(S) AN D ADDRESS(ES) U.S. Army Research Laboratory Information Science and Technology Directorate
Battlefield Environment Division
ATTN: AMSRL-CI-EW White Sands Missile Range, NM 88002-5501
9. SPONSORING/MONITORING AGENCY NAME(S) AN D A DDR ESS(ES )
5.
UNDING
N UMBER S
8. ERFORMING ORGANIZATION REPORT N UMBER
ARL-TR-821
U.S. Army Research Laboratory
2800 Powder Mill Road
Adelphi, M D 20783-1145
1 0. SP O N SO R IN G / MONITORING A G EN C Y REPORT N UMBER
ARL-TR-821
1 1 . SUPPLEMENTARY NOTES
12a . ISTRIBUTION/AVAILABILITY STA TEMEN T Approved fo r public release; distribution unlimited.
12b . ISTRIBUTION C O DE A
13 . ABSTRACT Max imum 20 0 words) The U.S. A ir orce, Navy, nd Army re n he rocess f upgrading he lectro-Optical actical Decision A id (EOTDA). he EOTDA has been used to predict the impact of weather and time of day on target acquisition. he
upgraded rogram is alled he Target Acquisition Weather Software TAWS). ew eatures f the A W S will
include: automated data access; upgraded path radiance routines; replacement of separate infrared IR); television (TV) and night vision (NV) sensor performance models with Acquire; and a U.S. Army standard sensor performance
model, which has become a standard in Department of Defense fo r IR , TV, and NV systems. o quantify the effects
on redicted arget cquisition ange f upgrading he ensor erformance model, omparison f he AW S
Version 2 with Acquire has been undertaken. Weather effects on target acquisition are examined also insome detail.
1 4. SUBJEC T TER MS weather effects, target acquisition, Acquire, TAWS, clutter
1 7. SEC UR ITY CLASSIF ICATION OF REPORT
UNCLASSIFIED
18 . SECURITY CLASSIF ICATION OF THIS PAGE UNCLASSIFIED
1 9. SECURITY CLASSIF ICATION OF A BSTR A C T UNCLASSIFIED
1 5. UMBER OF PAGES 64
1 6. RICE C O DE 2 0. IMITATION O F A BSTR A C T
SAR
NSN 7540-01-280-5500 Standard Form 29 8 (Rev. 2-89) Prescr ibed by ANSI Std.Z39-18
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 4/55
Acknowledgements
The uthors would ike o cknowledge he many nd eneficial discussions with Dave Dixon, Training and Doctrine Command Analysis
Center-White ands Missile ange TRAC-WSMR), John Mazz, Army
Materiel Systems Analysis Agency (AMSAA), Melanie Gouveia, Logicon,
and av e chmieder, eorgia ech esearch nstitute GTRI) hat
occurred uring he ourse f his ork. heir id n roviding information and tracking down documents concerning both the Acquire
and TAWS sensor performance models was invaluable.
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 5/55
Contents
Preface iii
cknowledgements V
Executive Summary 1
1. ntroduction 5
2. ackground 7
2.1 Minimum Resolvable and Detectable Temperatures 7
2.2 ensor Performance Models 8
2.2.1 Acquire SPM and Methodology 9
2.2.2 AW S SPM and Methodology 14
2.2.2.1 Resolved Versus Unresolved Targets 16
2.2.2.2 Determination of SC R 17
2.2.3 Discussion 21
3 . Comparisons 23
3.1 cenarios 23
3.2 Model Runs 24
3.2.1 haracteristic Dimension 24
3.2.2 ensor Curves 26
3.2.3 Aspect Ratios 26
3.2.4 Backgrounds 27
3.2.5 Clutter 28
3.3 esults 29
3.3.1 Clutter/Complexity Effects 30 3.3.2 Weather Effects 33
3.3.2.1 Visibility 33
3.3.2.2 elative Humidity 33
3.3.2.3 ky Cover 36
3.3.2.4 og and Precipitation 37
vu
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 6/55
Figures
Tables
4. onclusions 9
References 1
Acronyms 5
Distribution 7
1 . quivalent bar pattern 0
2. arget-acquisition methodology 1
3a. nitial clutter temperature algorithm 0
3b . Current clutter temperature algorithm 0 4. robability of detection versus resolution for 50 percent
acquisition level using the T A W S and Acquire algorithms 1 5a. -80B 5
5b. -80U 5
6 . Comparison of the T A W S and Acquire M R T curves 6 7. Acquire versus the T A W S clutter comparison 8 8. he TAWS versusAcquire-detection ranges at 50 percent
probability of detection 9
9. cene complexity effects on target-detection ranges 1 10 . isibility impacts onthe TAWS-detection ranges 5 11. elative humidity impacts onthe TAWS-detection ranges 5 12. ky cover impacts onthe TAWS-detection ranges 7 13. og impacts on the TAWS-detection ranges 8 14. recipitation impacts on the TAWS-detection ranges 8
1 . 50 as a function of Acquire version 4
2. ummary of detection methods 7
3. Clutter temperature C T values 8
4. Algorithm fo r determining Sc 9
5. nitial T A W S implementation of CT algorithm 9 vm
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 7/55
6 . Current TAWS implementation of CT algorithm 0 7. Weather conditions used in th e study 3
8. roblems between TAWS and Acquire algorithms and
their resolution 4
9. -80 dimensions 5 10. mpact of a second background on detection range 2 11. Detection ranges as a function of various weather 4
IX
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 8/55
Executive Summary
Overview
The range a target can be detected on the battlefield is a valuable piece of information or he attlefield ommander. etection nd ecognition
ranges epend pon he arget nd ackground haracteristics,
atmospheric ropagation, nd ensor erformance. eather actical
decision aids provide information on sensor performance under adverse
weather conditions.
The Target Acquisition Weather Software (TAWS) is an updated version
of the U.S. A ir Force Electro-Optical Tactical Decision Aid. he TAWS
provides .S. ir orce, Navy nd Army mission lanners nd
warfighters with ppropriate nformation or ptimal ensor nd/or weapon systems selection, acquisition range determination, and mission routing under degraded weather conditions. he T A W S was originally
constructed to predict detection and lock-on ranges only. ecause the U.S. Army extensively makes use of recognition and identification ranges,
methodologies fo r adding this information to the T A W S were examined. The leading algorithm, and thus, a contender fo r the replacement of the
current algorithm, is Acquire.
Physics-based actical Decision Aids TDA)s, uch s AWS, mploy
physics alculations hat have heir asis n heory r measurements.
Thus, a physics-based TDA employs routines and physics that allow it to ascertain he robability f etecting iven arget t iven ange
under xisting r redicted weather onditions. he ffects nd methodology fo r determination of th e range a target is detected under
adverse weather conditions at infrared (IR) wavelengths is the subject of
this eport. he ensor erformance model Acquire nd he enor
performance model in TAWS both use what is known as the equivalent bar attern pproach; owever, he nderlying ssumptions f ac h
algorithm re onsiderably ifferent.
he urpose f this eport is o show th e ifferences between these two methodologies and to examine
weather effects on target-acquisition ranges.
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 9/55
Background
Typical performance prediction models for resolved targets a target is considered o be esolved when the arget-angular ubtense nominally
exceeds he ensor's ngular ubtense n both ertical nd orizontal dimensions at the range of interest) treat the target with the bar/target
equivalency riteria nd he ensor with he minimum esolvable
temperature MRT) unction. ar - nd arget-signal quivalency s
established by quating he bar attern emperature ifference o he
target verage emperature ifference. he etection range s harply
bounded in that it can never exceed the range at which the target ceases
to be resolved, that is the (detection) range «target size/resolution.
Models fo r predicting the
detection range
of unresolved targets typically rely trictly n arget-signal trength or etection. hese models
typically abandon both the bar- and target-equivalency criteria and the M RT approach. Unresolved target models are often called hot spot or
star detection models because they rely on high-apparent contrast fo r
detection. he target is a square or circle with dimensions matched to the high temperature target area of interest. his target spot detection
methodology pplies o ases he arget s iewed gainst niform background, and detection occurs when the ignal-to-noise ratio on the
display element that subtends the target exceeds that of the background. The methodology fo r spot detection applies only to the detection of the
target (its discrimination from the background), and not to levels of target discrimination. he ensor unction s he inimum etectable temperature (MDT).
The D T s nly ppropriate or argets gainst niform r
unstructured background, fo r example, aircraft against a clear or overcast sk y r vehicles n a esert background. earching fo r anks gainst a varied terrain background requires the M R T approach. Additionally, the
M D T pproach nly epresents etection whereas he M RT pproach,
which may also be used fo r detection, is required fo r target recognition
and identification.
Target detection, recognition, and identification methodology applies to
situations in which the target is embedded in a non-uniform or cluttered background and it is necessary to separate the target characteristics from
the background. he target discrimination M R T methodology, based on
the Johnson cycle criteria in Acquire and Schmieder's criteria in TAWS, ca n be used fo r the prediction of target-acquisition range at
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 10/55
Conclusions
discrimination evels f detection, recognition, and dentification. oth
the TAWS and Acquire can use M DT to predict detection range also.
T he AW S urrently ses chmieder's work mplemented o redict
detection; igher isaimination evels, uc h s ecognition nd
identification, re not included. ecause Acquire s urrently both the U.S. Army and A ir Force standard fo r target acquisition, and it predicts
ranges or iscrimination evels f etection, ecognition, nd
identification it was decided to replace the urrent sensor performance
model SPM) esident n he AW S with cquire. owever, his
replacement aises he uestion f what ifferences may rise ue o
different methodologies between the SPMs in the T A W S andAcquire. o
answer his uestion or R ensors, omparison f tatic arget
discrimination methodologies and th e resultant target-acquisition ranges
produced by the TAWS and Acquire was undertaken. t should be noted
that this comparison in no way should be construed as a validation of the target-acquisition anges. ather we re xamining what, f ny,
differences that arise due to the underlying SPMs and the methods that
they implemented.
To ompare hese wo omplex arget-acquisition models equired
standardization of as many parameters as possible. o accomplish this,
one weather scenario was usedin conjunction with one sensor and target,
both with ixed rientations. hus, winter cenario was hosen nd examined using n xercised -80 oviet main battle ank against two
backgrounds vegetation and snow) at IR wavelengths. he sensor and tank were aligned such that the sensor always had a frontal view of the
tank; the ensor height was fixed at 300 t facing north. he date was
fixed t 21 December t a ocal time f 12N; the ocation was ixed t
latitude f 37°32' N, ongitude f 127°00' Seoul, South Korea). he weather onditions ncluded lear kies with arying isibility nd
relative humidity and overcast skies with varying visibility and relative
humidity. dditional ases ere un ncluding ight/heavy og
conditions, snow, drizzle, and rain.
The PM in the TAWS is based on Schmieder's image-based work and
thus requires additional algorithms to determine the effects of clutter in
the scene; these algorithms ar e not necessarily intuitive. ince Acquire's
target ransform robability unction nd chmieder's etection
probability s unction f esolution, gree easonably well nder
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 11/55
moderate lutter onditions, nd because Acquire s n ndustry, U.S.
Army and A ir Force standard, it was decided to replace the current SPM
with Acquire. Clutter levels can be accommodated in Acquire by varying the N50 parameter.
The incorporation of the Acquire SP M into the T A W S is scheduled fo r the
TAWS Version 3, which will be released fo r use in 2001. he comparisons
shown here, of target-acquisition range output from the current version
of T A W S with output rom Acquire, provide ositive eedback on the
benefits f his nhancement o AWS , while maintaining xisting
interfaces nd providing comparable arget-detection range predictions.
T he rimary benefit f this nhancement will e he bility o pecify
target acquisition discrimination levels, including detection, recognition,
and identification.
Because th e cases examined in this report are limited to a winter scenario in Korea, specific quantitative results of th e selected weather parameters'
impacts on target-detection range cannot be generalized to al l situations. However, these cases do highlight the importance of considering accurate atmospheric conditions in target-acquisition predictions. esults show a
smaller weather impact to Acquire detection ranges than predicted using
the current T A W S SP M under conditions of fo g or precipitation.
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 12/55
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 13/55
2.
Background
2 .1 Minimum Resolvable and Detectable Temperatures Typical performance prediction models or resolved targets a target is
considered o e esolved when the arget-angular ubtense nominally
exceeds he ensor's ngular ubtense n both ertical nd orizontal
dimensions t he ange f interest) reat the arget with he bar- and
target-equivalency criteria and the sensor with the minimum resolvable temperature MRT) unction. his methodology assumes that resolved
targets re etected, based n bserver attern ecognition. ignal strength only needs to be sufficient in order to define the pattern. he
signal s ypically efined s he emperature ifference between he
average temperature of the target and a uniform background temperature
as seen by the sensor. ar- and target-signal equivalency is established by equating he bar pattern emperature ifference o he arget verage
temperature difference. he detection range is sharply bounded so that it
can never exceed the range at which the target ceases to be resolved, that
is the detection) ange « target size /resolution. his is the maximum range t which eriodic arget an be aithfully eproduced; hus
target is considered unresolved if the projected sensor instantaneous field
of
iew
IFOV)
s
reater
han
0
ercent
f
he
arget's
ritical dimension. he percentage is taken to be 80 percent because other sensor apertures, n ddition o he detector IFOV, cause he ffective ystem
IFOV to be slightly larger.
Models fo r predicting the detection range of unresolved targets typically
rely trictly n arget-signal trength or etection. hese models typically abandon both the bar- and target-equivalency criteria and the
M RT approach. Unresolved target models ar e often called hot spot or star detection odels ecause hey ely n igh pparent ontrast or detection. he target is a square or circle with dimensions matched to the
high emperature arget rea f nterest. his arget pot etection
methodology pplies o ases n which he arget s iewed gainst a uniform background, and detection occurs when the signal-to-noise ratio
on he isplay lement hat ubtends he arget xceeds hat f he
background. hat s, ufficient mount f arget nergy eaches
detector lement o reate ot spot on he ystem isplay. For this
reason, pot etection s lso eferred o as tar etection.) The
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 14/55
methodology fo r spot detection applies only to the detection of the target
(its iscrimination rom he background), nd ot o evels f arget
discrimination. he ensor unction s he inimum Detectable
Temperature MDT). etection ange redictions re ot harply bounded with his nresolved arget methodology ince arget-signal
strength does not abruptly disappear after becoming unresolved.
T he DT s nly ppropriate or argets gainst niform r
unstructured background; fo r example, aircraft against a clear or overcast sk y r ehicles n a desert background. earching or anks gainst a
varied terrain background requires the M RT approach. Additionally, the
M D T pproach nly epresents etection whereas he M R T pproach,
which may also be used fo r detection, is required fo r target recognition
and identification. f a target is hot enough, the M D T approach predicts
target detection even though the target may be smaller than a forward- looking infrared (FUR) detector element. or al l practical purposes, the M DT pproach s ot sed n he Army's ombat imulations ince
recognition r dentification s equired before iring n arget. t should also be noted that in a cluttered environment the target would not
be the only hot spot.
2 .2 ensor Performance Models Target detection, recognition, and identification methodology applies to situations in which the target is embedded in a non-uniform or cluttered background and it is necessary to separate the target characteristics from
the background. he target discrimination M RT methodology, based on th e Johnson cycle criteria [4] in Acquire and Schmieder's criteria [5,6] in the TAWS, ca n be used fo r the prediction of target-acquisition range at
discrimination levels of detection, recognition, and identification. oth
the T A WS and Acquire can use M DT to predictdetection range also.
Acquire, eveloped y he .S. Army Night ision nd lectronic Sensors Directorate (NVESD), is an analytical model that predicts target- detection and discrimination-range performance fo r systems that image
in the visible, near-IR, and IR-spectral bands. anges and probabilities predicted y he model epresent he xpected erformance f n ensemble of trained military observers with respect to an average target
having a specified signature and size.
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 15/55
The U.S. A ir Force Research Laboratory's (AFRL) Electro-Optical Tactical
Decision Aid EOTDA) 7] was eveloped o rovide he ser with a
single iece f oftware o valuate he ombined ffects f target-to-background ontrast, tmospheric ransmission, nd ensor performance on the range at which a target ca n be detected by an imaging
device. he model treats detection by television, image intensifiers, and
thermal maging evices. he AWS, ri-Service rogram, s n
upgrade of the EOTDA.
The A W S urrently ses chmieder's work mplemented o redict
detection; igher iscrimination evels, uch s ecognition, nd
identification, ar e ot included. ecause Acquire s urrently both the
Army standard 8] nd a U.S. A ir Force standard fo r target acquisition,
and because t redicts anges or iscrimination evels f etection,
recognition nd dentification, t was ecided o eplace he urrent sensor erformance model SPM) esident n he A W S with Acquire. However, this replacement raises the uestion of what differences may
arise due to different methodologies between the SPMs in the TAWS and
Acquire. o answer this question fo r IR sensors, a comparison of static
target-discrimination methodologies and the resultant target-acquisition
ranges produced by the TAWS and Acquire was undertaken. t should
be oted hat his omparison n o way hould be onstrued s validation f he arget-acquisition anges er e. ather we re
examining what, if any, differences that arise due to the underlying SPMs
and the methods that were implemented.
2.2.1 Acquire SPM and Methodology
During he 950s, he ilitary eveloped lectro-optical mage
intensifiers, which rovided nhanced isual urveillance apabilities
under conditions of limited visibility. he complexity of these intensifiers and ssociated arget-acquisition ystems equired methodology or
evaluating performance haracteristics. ohn Johnson, of the then U.S. Army ngineer esearch nd evelopment aboratories ERDL)
(currently VESD), resented esults f xperiments with uman
observers conducted at ERDL to determine the resolution required of a system to perform certain target-interpretation processes such as target
detection and recognition. e referred to these as decision responses and said hat the processes were dependent upon the haracteristics of the optical message, he roperties f he ntensifier evice, nd he physiological response of th e human readout processes. hrough a series
of xperiments sing rained bservers ooking t argets nd ar
resolution diagrams, Johnson [4 ] developed a method relating the decision
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 16/55
Figure 1. quivalent bar pattern.
response o he umber f bars line airs) ormalized o he hortest
target imension hat n bserver eeded o ee o make ecision
(detection, recognition, etc.).
T he methodology developed by Johnson was simple and straightforward.
In the laboratory, scale models of various military targets were moved to
a istance where hey could just be detected as viewed by an observer
through an image intensifier. ar charts, with the same contrast as the
scale models, were then placed in the observer's field of view at the same
range s he arget. he patial requency line airs) esolved by the
observer was hen etermined s unction f ontrast. his ame
methodology was used or determining he ine airs equired by the
observers to recognize the object seen as a tank. he spatial frequency of the pattern was specified in terms of the number of lines in the pattern
subtended by the object's minimum dimension as illustrated in figure 1 .
Figure 1 shows three bars across the tank's shortest dimension as seen by an observer; this is the riterion Johnson used or ecognizing that the object was a tank. urther, Johnson found that the normalized line-pair resolution equired or articular decision esponse was early
constant fo r the group of nine military targets he employed. n the case
of target detection, that was 1.0-line pairs per shortest target dimension. These ormalized to he hortest arget imension) ine air alues
required or ecision were ound o be ndependent f contrast nd scene signal to noise ratio as long as the contrast on the bar chartwas the
same as the target contrast.
he results showed that decision levels fo r military argets may e onsidered quivalent o ar atterns f
appropriate spatial frequencies.
10
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 17/55
This methodology, which provided arget-discrimination riteria based
upon resolution, gained widespread acceptance within the industry and
became he ccepted riteria or erformance measurement f ptical systems. hese riteria ere eferred o s he ohnson Criteria r
equivalent bar pattern approach.
In 1974,16 years after publishing his original paper, Johnson modified his
original work nd xtended t o over R ystems 9]. his aper
emphasized hat he alues or he arious ecision evels re
representative values, essentially average values required fo r 50 percent
probability, and must not be onstrued as rigid or optimum values fo r
specific targets and target-aspect views. he values associated with the various ecision evels emained he ame, e.g., .0 ine airs or
detection) except fo r recognition, which changed slightly. his paper also
recommended and provided procedures fo r usingthe concept of MR T fo r thermal sensors. ohnson's methodology is schematically represented in
figure 2. [10]
Figure . Target- acquisition methodology.
Sensor Resolution
m i m HI nf^
mi m utfini. e U jj/lii: r
Spatial Frequency
cyrles/mrad)
Probability
ID
O jS
DMClfDI
Ibci0iriifl
£ 02.
lhrillEidc£
D IW L
The M RT s efined s he emperature ifference between uniform background nd he ars f our-bar attern, ac h bar aving
7:1 aspect ratio (so the overall pattern will be a square), which is required
by rained bserver o ust esolve ll our bars when iewing he
pattern hrough n mager. 11] he ef t side f the igure hows he
resolvable emperature ifference contrast) ersus he aximum
resolvable bar pattern (spatial frequency) as a function of contrast. or a specific target-contrast level, the maximum resolvable spatial frequency is th e ighest patial requency t which uman bserver an till
recognize he our istinct bars nd not ne r wo blobs. hus, he
temperature difference required to resolve the four bars increase, as the bars et maller. This maximum esolvable patial requency s
11
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 18/55
function of contrast visual or thermal) nd s the minimum resolvable
contrast (MRC) in the visual or M R T difference curve in the thermal. n
the figure, the bars represent the generic formulation, whereas the solid line would epresent he arget ontrast nd ensor esolution or
specific sensor. ith knowledge of the target's contrast (AT in the E R ) , critical dimension, range, and the atmospheric attenuation, the number of resolvable ycles, N, cross he arget's ritical imension an e
determined by
N =
f ^f->. where is the maximum resolvable patial frequency of the sensor in
cy/mr) t he pparent Ta usually n K), Hta r g s he arget-critical dimension (in meters) and R is the range (in kilometers). he apparent thermal contrast is determined by
AT a = AT T(R), 2)
where T(R) is th e atmospheric transmission. Under the assumption of a
homogeneous tmospheric path he ransmission may be ound using Beer's law
T(R) = e-ßR 3)
where s he tmospheric xtinction oefficient determined rom n atmospheric propagation code [12,13]. W e now need a way to correlate N
with the iscrimination level: etection, ecognition, and identification.
Johnson did this by establishing the so-called target transform probability
function TTPF). 9] he TPF, hown n he ight-hand ection f
figure 2, was erived rom aboratory sychophysical xperiments n which the ability of observers to perform a particular discrimination task
as a function of resolvable cycles across the target-minimum dimension
was measured. or a given discrimination task, the TTPF represents the
50-percent point, referred to as N50, as determined from this ensemble of
12
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 19/55
range. hus, if the range is unknown, but information is available about
the target-critical dimension, the target/background AT , and the sensor
response urve, then a olution ca n be ound through iteration fo r the range at a predetermined Pd.
A s LIRs dvanced rom irst o econd eneration, heir esolution
increased o hat here was early qual esolution long he ca n
direction and perpendicular to it. his, in part, necessitated a change in
the riginal ne-dimensional ID ) ersion f Acquire. n pdated version of Acquire, discussed in detail in reference 11 , was issued in June
1990. he modelupdate consisted of two parts:
1 . LIR90, which predicts laboratory measures and
2. two dimensional (2D) version of Acquire, which predicts field performance.
In LIR90, redicted r measured orizontal nd ertical MR T s re
averaged at a particular temperature using
/ff=(x* ) * -4
) *
This effective spatial frequency applies to the effective M RT used along
with modified alues f N50 or he ifferent iscrimination asks o predict range performance. he N50 values fo r ID or 2D are presented in table 1 . he N50 fo r a particular task, using the more recent 2D version of the model, is found by multiplying the original ID N50 values by 0.75. The amount of shift was determined by requiring the range predictions
fo r he D model o orrectly redict he esults f field ests, which
served s alidation or he riginal ID ) model. he iscrimination levels in figure 2 are fo r second generation FLIRs. s an example, if we
have a target with a critical dimension of 3 m at a range of 3 km, with an
apparent contrast that would ive a maximum resolvable frequency of 3 cycles/mrad, his would ead o robability f ecognition, or
second generation
FLIR,
of 50
percent; with a
first generation FLIR,
this would lead to a probability of recognition of 25 percent.
* The discussion in this paragraph relies heavily on reference 11 .
13
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 20/55
apparent contrast that would ive maximum resolvable frequency of 3 cycles/mrad, his would ead o robability f ecognition, or
second generation FLIR, of 50 percent; with a first generation FLIR, this would lead to a probability of recognition of 25 percent.
Table 1 . N5o as a function of the Acquire version Acquire version ID 2D
Detection 1.0 .75
Recognition 4.0 3.0
Identification 8.0 6 .0
2.2.2 AWS SPM and Methodology
The TAWS SPM is derived from work done in the 1980s at the U.S. Air
Force vionics aboratory now art f FRL), which ed o he development of a research grade TD A. he main difference between
the AW S nd he Avionics esearch rade ode ies rimarily n he
underlying thermal model, the Thermal Contrast Model 2-TCM2, with T A W S using a scaled-down version of that model. he SP M resident in TAWS uses the equivalent bar-chart approach but differs from Acquire
by irectly ncorporating lutter ffects or 0 ercent robability f acquisition. he T A W S predicts lock-on range based on signal-to-noise ratio hresholds, hot pot detection based n M D T methodology, nd discrimination etection ange based n M RT methodology. Thus, ll
other arameters being qual, he AW S etection anges hould be
approximately qual o hose redicted by Acquire t he 0 ercent
probability of detection level fo r moderate clutter (see discussion below).
Clutter is automatically computed, based on an empirical algorithm [14], and is a strong factor in determination of the number of cycles on target.
During the 980s, Dave Schmieder t the GTPJ examined he ffect of
clutter on target detection. 5,6] He found that the amount and nature of
background clutter had a significant impact on the probability of target
detection. chmieder first looked at scene radiance standard deviation as a clutter measure. owever, this measure had the deficiency of giving
large lutter values o elatively uncluttered cenes when hose cenes possessed several intensity modes. Moreover, this definition, like many other amplitude measures [5] , lacked a weighting factor based on target
size. oth amplitude and target size measures appeared to be required to
14
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 21/55
predict bserved rends. ince xisting efinitions ppeared
inappropriate, Schmieder redefined th e erm. he clutter definition he
used was a scene radiance standard deviation computed by averaging the radiance ariances f contiguous cene ells ver the whole cene nd
taking the square root of th e result. his is formulated as
clutter = [fVf INf1, 5)
i
where \ is the radiance standard deviation fo r the ith cell and N is the
number f ontiguous ells n he cene. his efinition mplicitly
included both target size nd ntensity measures nd produced higher
values fo r scenes that looked more complex and cluttered. t also avoided
yielding large lutter value or relatively uncluttered cenes that still
contained variations in intensity. dditionally, it accounted fo r clutter object izes lose o he arget ize weighing ore n he lutter calculation. However, as with other definitions of clutter, this definition
introduces ts wn et f roblems. cene magery s equired o
adequately determine o\ and eq (5) is not scale invariant but depends on the cell size elected fo r calculation. chmieder took the cell size to be square in shape with side dimensions of approximately twice the target
height. However, if the scene under examination contains more than one type of target (man, tank, bridge, etc.) the cell size must be redefined fo r
each target examined, resulting in different values fo r ci and N.
Based pon his efinition, chmieder erformed xperiments with
observers which howed that clutter could be categorized according to the ignal-to-clutter atio SCR) where he ignal was he emperature difference between the maximum/miitimum target temperature as seen
by ensor nd he background emperature where he lutter was defined as above. chmieder found that high clutter exhibited a SC R of
< 1, moderate clutter an SC R of 1 to 10 , low clutter an SC R > 10 and < 40 ,
and no clutter effects could be assumed if the SC R > 40 . urthermore,
Schmieder's data howed that fo r a detection probability of 50 percent, the number of cycles (line pairs) required fo r detection of a target varied
between 0.5 fo r low clutter to 2.5 fo r high clutter, with moderate clutter
requiring approximately 1.0 cycle.
Schmieder oncluded hat cquisition evels re trong unction f clutter s well s esolution nd hat ange rediction models must
include lutter ffects. ecause he umber f ine airs er arget-
angular ubtense ecessary or etection s nversely roportional o
detection ange, hanges n CR an be xpected o ignificantly alter
target-detection range predictions.
15
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 22/55
The TAWS detection methodologyuses Schmieder's SCR algorithm at the
50 ercent probability of detection evel o determine whether MRT or
MDT ormalization hould be used nd o alculate arget-acquisition ranges. he SCR may be hought f as he atio f hot/cold pot ATT
(i.e., maxmium/minimum arget emperature ackground ean
temperature) to clutter equivalent temperature (CT), thus
SCR = ATT/ CT- 6 )
The value of CT s determined by the IR scene complexity (a measure of
the umber f bjects n he cene ompeting with he arget) n
conjunction with he ackground cene ontrast Sc) n he arget's
vicinity. ince detection methodologyand acquisition range both depend
upon SC R (discussed in some detail below).
The AWS etermination f cquisition ange s ixed t 0 ercent
probability f detection. hus, s mentioned bove n he ection n
Acquire MRT methodology, a solution for the range can be found if the
target's ritical imension, arget nd ackground emperatures, nd
MRT sensor curve re ll known. hile TAWS uses this methodology,
there re ignificant ifferences rom Acquire—most otably n he
determination of the number of line pairs on target, N.
2.2.2.1 Resolved Versus Unresolved Targets
Whether he AWS omputes he etection ange ia MRT r MDT
methodology depends on whether the target is resolved or unresolved at
a given sensor to target range and the SCR at that range.
For resolved targets, if the SCR > 40, indicating little or no clutter, MDT is
used with he arget ontrast ased n he emperature ifference
between the hottest (or coldest) facet of the target seen by the sensor and
a given background; this is referred to as ATMAX. ven though the target
is esolved, Schmieder did not ecommend using MRT because he elt
that with uch high ignal o lutter > 0), t does not matter f the target is resolved or not, and MDT (hot/cold spot) detection would give a
longer detection range than MRT. f the SCR is < 40, MRT is used with
the target contrast, A T A V G , being the difference of the average temperature
of the target facets seen by the sensor and the background temperature.
16
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 23/55
If the target is unresolved at a given range, it is either detected with the
M D T methodology or is not detectable at all. f the unresolved target is
detectable using M D T , then SC R > 40 . n that case, ATMAX is used fo r the target and background contrast. f SC R < 40, and the target is unresolved, then the clutter is to o high, and no detection occurs at this range and the
range must e ecreased or etection o ccur. hese esults re
summarized in table 2.
Table 2. ummary of detection methods
SCR Line pairs on target Method
<1 2.5 M RT
< 1< 10 < 1.0 M RT
10<40 0.5 M RT
>40 - M D T
For alculating he CR n he bove rocedures, chmieder
recommended using the ignal ATMAX ather than ATAVG, fo r the reason
that in his 1982 SC R study [6] the signal component was defined as ATMAX
rather than ATAVG-
2.2.2.2 Determination of SC R
In the absence of actual imagery of the target area, Schmieder estimated
clutter from a combination of scene complexity and Sc. W e present his
methodology in some detail below.
Schmieder's SC R algorithmaffects the acquisition range by modifying the
number of line pairs (N) on target. At 50 percent probability of detection,
Schmieder's results can be formulated as [15]
N -- 1.64 H
arg
(log(50?) + 1.2)1M
R (7 )
17
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 24/55
fo r CR .1 . he A W S uses he arget height fo r the arget-critical
dimension, Hta r g. n order to determine SC R from eq (6 ) both the target
contrast ATT) nd he lutter emperature, Cr must be known. T, n turn, equires knowledge f both cene ontrast and cene omplexity.
Because the T A W S does not produce or use target- and background-scene
images, it cannot compute clutter as originally defined by Schmieder. o
rectify his roblem, chmieder 16 ] ecommended hat lutter e computed based n T . e ssumed hat n he verage, he arget
contrast, AT T, which represents the temperature difference between target
and background, would be qual to °C. chmieder then used eq (6),
with he verage TT C, oupled with CRs f 0 , 0, , nd ,
representing clutter states of none, low, moderate, and high respectively,
to etermine alues or r. chmieder tates 16] , Clearly, ther
assumptions f arget ignal nd CR would ead o ther ssociated values. he values shown (in table 3) ar e used as a starting point to be used until the results of further research can yield a better basis fo r these
choices.
A method fo r relating scene contrast (Sc) to the clutter levels is needed.
Schmieder hose 17] c anges f o , o , o , nd o correspond o he lutter evels f one, ow, moderate, nd igh, respectively. hese quantities (SCR, CT, Sc) and their relation to clutter
level are summarized in table 3.
Clutter Sc range level SCR CT degrees C) degrees C)
none 40 .0 5 <1.0
lo w 1 0 .2 > 1.0-£2.0
moderate 4 .5 >2.0-<4.0
high 1 2 >4.0
To bviate he eed or he ser o elect lutter evel, he A W S
implemented Schmieder's lgorithms to determine CT based on a user
input of IR-scene complexity and a calculation of scene-thermal contrast by the AW S IR model. 18] he lgorithm fo r determination of Sc as
found in TAWS, is presented in table 4. n this table, TB s a background
temperature nd he ubscripts nd 2, H and efer o he irst and
second, high nd ow background emperatures, espectively. able 5
and figure 3a show the original implementation.
18
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 25/55
Table 4. Algorithm for determining Sc Number of c value backgrounds degrees C) 1 .5
2 BI TB 2
3 BH - TBL
Table 5. nitial T A W S implementation of Or algorithm
Scene
Complexity
Sc degrees C)
Otol >lto2 >2to4 >4
r degrees C) None 05 05 05 05
L ow 05 2 2 2 Medium .05 .2 .5 .5
High .05 .2 .5 2.0
In operational use it was found that the step function shown in figure 3a
resulted in occasional discontinuities in detection range. o mitigate this,
the tep unction was eplaced with he ontinuous inear unctions
shown n able nd igure b. ence, n onjunction with cene
complexity, he arger he emperature ifference mong backgrounds
(i.e., the higher the scene contrast Sc), the higher the value of Or, which in
turn owers he alue f SCR o c Cr1)- he ower SCR s, he ewer
equivalent ine pairs here re per arget see q 7]), which ffectively
reduces the detection range. n general, as the number of backgrounds
increases he detection range will change, enerally decreasing in value
somewhat. his s easonable ince y dding ackgrounds, ne s
effectively increasing the clutter and making the target more difficult to
detect. hus, the lgorithm in TAWS for deterrriining Or is reasonable;
however, he boundaries or Or nd c, hosen by Schmieder 16,17]
using his experience and expertise, are subject to review. As indicated in
a previous paragraph or Or; nd s Schmieder pointed out 17 ] or Sc
These values result from heuristic and judgmental considerations. hey
have not been derived from sensitivity trades, which rigorously calculate
the lutter evels hat re btained rom arious background ontrast
conditions. uch omprehensive tudy will be ventually needed o
arrive at more fully supported values.
19
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 26/55
E
I 1 . 0
S c e n e Comple x i t y
— High
- -»Low
2 Sce ne Cont ras t Sc ( de gre e s )
2 S c e n e Cont ras t Sc ( de gre e s )
Figure 3a. nitial clutter temperature algorithm. Figure 3b. Current clutter temperature algorithm.
Table 6 . Current T A W S implementation of Or algorithm
Scene
complexity
Sc degrees C)
Otol >lto3 >3to4
CT degrees C)
>4
None
L ow
Medium
.05 .05
.05 +0.15 Sc .2
.05 +0.15 Sc
.05
.2
.5
.05
.2
.5
High .05 +0.4875 Sc 2.0
This clutter algorithm has an unexpected effect on MRT detection ranges.
When only one background is selected by the user (i.e., no scene-thermal
clutter), or if the temperature difference among the selected backgrounds
is < 1.0, the acquisition range found is the same whether the user selected
scene omplexity s ow r medium. ote he middle cene ontrast
categories now break at 3° rather than 2° for medium-scene complexity.
This change
is not significant since the
impact on clutter temperature
is at
most 15°, and he original distinction between low and moderate scene
contrast ategories was n rbitrary election etween he hreshold
values of 1 for none and 4 for high.
20
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 27/55
2.2.3 Discussion
Johnson's aper mentioned he ssue f ackground lutter. e
emphasized hat is etection riteria 1.0-line airs) were pplicable under onditions hat equired ome egree f arget hape
discrimination n rder o etect he arget rom ther bjects n he
background, (i.e., where background clutter was present). He also stated
that he umber f line airs equired o ttain a particular detection
probability ould ary ignificantly depending upon the ature f the
background lutter. his aised umber f uestions s o he
significance of background clutter on the validity of the Johnson criteria,
especially since Johnson did not provide a definition of clutter.
Schmieder noted that the Johnson criteria fo r target detection (1.0 cycles
or line pairs) correlated well with his moderate-clutter category (implying that uch evel f lutter was robably resent n ohnson's work, although ohnson id ot measure t). omparison f Schmieder's
detection probability as a function of resolution with Acquire's TTPF (see
figure 4) clearly shows that Acquire's formalization compares favorably with chmieder's oderate lutter. ther lutter evels an e
accommodated in Acquire by varying N50.
Figure 4. Probability
of detection versus
— Scluueder: L-
— — Schmierer: H
olution fo r 50 it acquisition
3 the T A W S 1 —| _ — ..... Acquire - r algorithms. a .« —
S -w ÄO i.4 —
0L , •
y
0. 1 —
i <s
1 z 3
Resolution Line Pairs / Target)
21
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 28/55
3.
Comparisons
3.1 cenarios
To ompare hese wo omplex arget-acquisition models equires
standardization of as many parameters as possible. o accomplish this,
one weather scenario was used in conjunction with one sensor and target,
both with fixed orientations.
A winter cenario was hosen nd xamined sing n xercised -80
Soviet main battle tank against two backgrounds (vegetation and snow)
at IR wavelengths. he sensor and tank were alignedsuch that the sensor always had a frontal view of the tank; the sensor height was fixed at 30 0 ft
facing orth. o minimize hadow ffects, he ate was ixed t 1 December at a local time of 12N. he location was also fixed at latitude of
37032' N/ longitude of 127°00' E (Seoul, S. Korea). he weatherconditions
include see able ) lear kies with arying isibility nd elative
humidity, nd vercast kies with arying isibility nd elative humidity. dditional ases ere un ncluding ight/heavy og
conditions, snow, drizzle, and rain.
Table 7 . Weather conditions used in the study
Relative Humidity
Precipitation
30 % none
50 % none
80 % none
100% light
fo g
100% heavy
fo g
80 % snow
90% light
rain
90% moderate
rain
Cloud clear clear clear clear clear
Cover overcast overcast overcast overcast overcast overcast overcast overcast
2km 2 2 2 2 2 2 2
Visibility 5km
10 km
5
10
5
10
5
10
rr; ---:-.- 5
10
5
10
15 km 15 15 15 \ 15 15
23
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 29/55
3 .2 Model Runs T he AW S was un using winter limatology along with the weather
conditions isted n able . omparisons were made with he cene
complexity nitially et t low. o educe he ossibility f rrors,
Acquire was initially run separately; however, this procedure was fraught
with problems (see table 8). o alleviate this, the Acquire algorithm was
programmed directly into TAWS , thereby, insuring that the many values
(AT, tmospheric ransmission) ere dentical n oth rograms.
Discussion onthe problem and their resolutionappear in table 8.
Table . roblems etween he AW S nd Acquire lgorithms nd heir
resolution
Problem Routine
TAWS Acquire
Resolution
Characteristic
dimension
Target height V(Xeff*Yeff) Target height
Sensor curves Horizontal (ID) Horizontal and
Vertical (2D)
ID
Aspect ratio V (7/2 Xeff/Yeff) <3 V (7/2 Xeff/Yeff)
Backgrounds 3 allowed N/A 1 used
Scene
complexity
None, low,
moderate, high
N/A+ Low
* ee section 3.2.>
3.2.1 Characteristic Dimension
The AW S nd Acquire se ifferent arget-characteristic imensions.
For M RT calculations, T A W S uses the target height; whereas, 2D Acquire uses the square root of the target's projected area as seen by the sensor. When Acquire was rogrammed nto he AWS , hereby using D formulization, the characteristic dimension was changed to use the target
height.
In th e course of this study the T A WS and Acquire target T-80 Soviet main
battle tank databases were examined; the results from T A W S and Acquire databases, nd dditional ources, re resented n able . he CASTFOREM is the U.S. Army's entity level warfare simulation; World
Wide W eb ( W W W ) 1, 2, and 3 were taken from various, unsubstantiated, W W W ources or omparison urposes. he alues elected or he
various imension izes robably epresent ifferent onfigurations f
24
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 30/55
the -8 0 cf. igures a nd ). sing he un-forward ength s not
representative of the actual target size and will produce overly optimistic
detection ranges. he TAWS database has been subsequently changed to reflect he more ccurate alues. nce Acquire was oded nto he
TAWS, both algorithms used the same database.
T-80B T-«CU
TT-V^v^i;.
Figure 5a. -80B Figure 5b. T-80U
Table 9. -8 0 dimensions
Source Length (m) Width m) Height m)
Acquire^ - 3.59 2.64
T A W S 9.1 4.64 3.73
CASTFOREM 6.75 3.55 1.5**
W W W 1 9 .7 3 .6 2.2
W W W 2 9.9/7.4§ 3 .4 2.2
W W W 3 7.01 3 .6 2.20
X values are fo r projected area
§ gun forward/hull
* * does not include turret
25
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 31/55
3.2.2 ensor Curves
Because he AW S was onstructed during he 980s, it uses D M R T curves, whereas, Acquire uses 2D M RT curves. nitially, before coding
the Acquire lgorithm nto AW S nd he wo odes were xecuted separately, it was necessary to ensure that th e two E R sensor curves that
were being used were the same. his was accomplished by multiplying
the TAWS abscissa by the normalization factor of .7 5 (see section 2.2.1);
the esulting omparison s presented n igure . s hown, the wo
sensor curves re dentical. hen Acquire was coded into T A WS and
thus used TAWS M RT curves, N50 fo r detection was changed from .75,
appropriate fo r th e 2D algorithm, to 1.0, appropriate fo r the ID algorithm.
Figure 6 .
Comparison of the
TAWS and Acquire
MRT curves.
GD I L — —
2£D lD W ID inn
3.2.3 Aspect Ratios
In ID Acquire, the target was described by its minimuin dimension; thus, targets with dimensions (length x width) of2mx2mor2mx4mor2mxl6m
and identical AT would al l be equally detectable according to ID Acquire. The aspect ratio adjustment compensates fo r this unrealistic result. his
is an important issue fo r large aspect ratio targets such as battleships but
not fo r th e ypical aspect ratios of ground vehicles. n 2D Acquire, the
target is described by the quare oot of its area; thus, the 2 m x 6 m
26
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 32/55
3.2.4 ackgrounds
target will have a higherestimated probability of detection than the 2 m x
2 m target. herefore, in moving from ID to 2D Acquire, we have gone
from th e simplifying assumption that al l targets of the am e height ar e equally detectable (given th e same AT) to the simplifying assumption that
all targets with the same presented area are equally detectable; however,
aspect ratios > 3 are not recommended fo r Acquire. [3 ] or the typical
aspect atios f round ehicles, his s easonable implifying
assumption fo r 2D Acquire.
When sing D cquisition models uch s AWS , he irst tep conventionally performed in range prediction is to convert the laboratory
M RT o ak e nto ccount he cene bject spect atio maximum o
minimum object dimension) if use is to be made of the concept that an object is more readily discerned if the aspect ratio is greater than unity. The laboratory M RT is computed or measured with a bar aspect ratio of 7.
It can be shown [19] that
MRTfieid MRTj7/ 2Ne), 8)
where s is the scene aspect ratio.
For xample, or etection nly ne ycle s equired N 50 ),
yielding MRTfieid = MKlJl/2e Within TAWS , the M RT is adjusted
by a factor oi lXeff /2Yeff where Xe f f (along-track) and Yef f (cross-track)
are the projected target dimensions at a given range. he cross-track and along-track dimensions, viewed in a plane coincident with the sensor, ar e the bscissa nd rdinate, respectively. ince he alculation of aspect
ratio s ntegral o AW S alculations, o hanges were made o he
algorithm.
The AW S l lows alculations or cenes hat nclude p o hree backgrounds; whereas, Acquire does notconsider differing backgrounds,
primarily because Acquire ccounts fo r clutter through variation of the
parameter N50. he TAWS backgrounds ar e intimately connected with
the clutter calculations (see section 2.2.2), allowing ATT in eq (6 ) to vary
as the backgrounds ar e cycled. hus, eq (6 ) can be rewritten as
27
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 33/55
SCR AT / CT (9 ) 7= 1
3.2.5 Clutter
Figure 7 . Acquire versus the TAWS clutter comparison.
where he ubscript efers o he ackground nder onsideration
(considered primary) nd he ubscript; efers to the number of (user- entered) backgrounds NB maximum of 3) . hus, the T A W S detection
ranges fo r all backgrounds ar e calculated with the displayed background,
as printed on the output, considered primary. discussion of the lack of
backgrounds n Acquire, which re ntimately ied o he lutter
calculations in TAWS, is deferred to the clutter section below.
Schmieder's work, as implemented in TAWS, accounts fo r various levels
of lutter; whereas, Acquire nly ccounts or lutter by arying N50.
However, s hown n ection , Acquire's lgorithm, with 50
(detection) ompares avorably with Schmieder's t moderate lutter levels. o effectively represent Schmieder's low and high clutter cases, N50 takes on values of 0.5, and 2.5 respectively (see figure 7) . ince clutter is a subjective measure, othervalues of N50 may be chosen.
A— Schjnieder: L«w
Schmied er: Moderate Sclwieder: High Ac««xre,N50 = 1 .0 Ac«ni», N S O = O. S Acquire, NSO = 2.5
Resolution Line Pairs / Target)
28
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 34/55
3.3 Results
The December 1
ases were un sing ow-scene omplexity
with
single vegetation background in TAWS, and with N50 = 0.5 in Acquire. Results showing the detection ranges predicted at noon in each case by
TAWS andAcquire ar e shown in figure 8.
Examination f he ata hows maximum alue or he A W S
calculations t 2.1 m or he egetation background ue o ensor
optical resolution limits. ecause this limits scales with target height, a
determination an e ade sing alues aken rom eference
18 tables A.l-2. or the sensor and target chosen th e range limit turns out
to be 32.8 km in good agreement with the TAWS values. his same cutoff
fo r a maximum detection range was applied to the Acquire results.
For detection ranges less than 10 km the agreement between Acquire and TAWS was also good. etween these high and lo w detection ranges, i.e.,
midrange alues etween 0 nd 0 m, cquire redictions were
10 to 15 percent lower than the TAWS. pecifically, the cases with heavy
fo g or precipitation have identical values below 10 km, while the cases
with lo w or moderate humidity and high visibilities have values of km. Thus, e ay raw he onclusion hat he ontribution rom
approximately 2 lutter s rrelevant n nstances when he weather conditions ar e either extremely unfavorable (low visibility) or extremely
favorable (high visibility).
Figure 8 . he S versus
lire-detection iges at 5 0 percent
probability of detection.
1 0 0 0 0 TAWS Detection Range (km)
29
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 35/55
Some of the discrepancy between the T A W S and Acquire in the midrange
may be explained by the fact that the clutter temperature in the T A WS
runs with one background and low-scene complexity is 0.12S °C (tables 4 and 6 ), which equates to an N50 value of approximately .3 , while the value
used or hese Acquire uns N 5o .5 ) ould be ssociated with
low-clutter alue f 0.2 C. owever, note hat Acquire does not use
clutter temperature directly. ncreasing the clutter temperature to 0.2 °C in the TAWS, by including an additional background with an appropriate
scene ontrast emperature, id ower he AW S ange ut nly
accounted fo r less than half of the discrepancy between the T A W S and
Acquire values. When additional Acquire runs were made with N50 se t
= .3 , which is not strictly analogous to calculating a clutter level in TAWS,
the results were within 5 percent of the T A W S values.
3.3.1 lutter/Complexity Effects
Because he ser an elect cene omplexity evel f one, ow,
moderate, or high in the T A WS (specified as Clutter/Complexity in the
User Interface), it is important to understand how this selection affects the
resulting etection ange. nly ne background was sed or hese comparisons in order to hold constant th e cene contrast component of
clutter. hese omparisons howing he ffects f the hoice f scene
complexity n A W S nclude he omplete ourly ata rom he
December 1 uns, ather han just he oontime ata. he tandard
deviations of the difference between T A W S and Acquire detection ranges throughout each daily un re mall, ndicating that diurnal variations
behave similarly in both T A W S and Acquire. n general, most cases using
one background displayed a decrease in detection range of approximately
5 ercent when cene omplexity was ncreased rom none o ow r
moderate, and nother 5 ercent ecrease when scene omplexity was
increased from low or moderate to high. However, there is a fair amount
of ariation rom his ypical esult, s hown by he ample f runs
plotted in figure 9. n this figure, the abscissa represents the cases run
(generally in order of cases with low humidity on the left to cases with
high humidity and precipitation on the right), and the ordinate showing
the percent of the resulting detection range with
a given
level of scene complexity compared o he ange ound with scene omplexity et to
none. Note from tables 4 and 6 that with only one background low- and
moderate-scene complexity will return the same values.
30
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 36/55
Figure 9 . cene complexity effects on target- detection ranges.
Low/Moderate High ComplexH
86.0
C a s e s with I n c r eas ing Hum id i t y /P r ec ip i t a t i on
The primary exception to th e general impact of varying scene complexity
occurs both when TAWS detection ranges ar e very long and when they
are ery hort. or xample, esults lotted oward he ef t ide f
figure 9 include cases where humidity is 50 percent or less, and there is
no fo g or precipitation. hese conditions result in long detection ranges,
which ar e not substantially decreased when the scene complexity level is
increased. dditionally, ases lotted n he ight ide f he raph provide very short detection ranges due to fo g or precipitation; so that
increasing
he
omplexity
evel
as
o
orresponding
ecrease
n detection range, since the detection range is already so restricted. Other
deviations re elated o he elected isibility n ach ase, with increasing cene omplexity ausing mpacts more han he sual
5 percent when visibility is great. In order to highlight the general trend,
cases with visibilities of 2 and 5 km have been omitted from figure 9 , which would otherwise show even greater fluctuations. he cases potted with visibilities of 5 and 10 km reflect some fluctuation relatedto whether
th e cloud cover was entered as clear or overcast, which accounts fo r the
up-and-down nature of th e graph.
Another esult ccurs when multiple backgrounds re elected n he
TAWS. he previous results were based on cases using vegetation as the only background. f econd ackground s dded, t ffects he
detection range predicted fo r the first background if the complexity level is et o nything ther han one. he ffect f dding econd
background s hown n able 0, sing he oontime alues nd
excluding the cases reflecting extreme high or lo w detection ranges. A s
31
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 37/55
expected, detection ranges fo r the vegetation background decrease when
the econd background f now s dded, ince he lutter has been
increased. Additionally, detection ranges fo r the vegetation background increase when now s ntered s he rimary ackground with vegetation s the econdary background. his is expected because the
thermal contrast (AT) betweenthe target and the background is calculated
using the first, or primary, background entered in TAWS; the secondary
background does not interact with the target. ecause snow has a higher
albedo han he egetation, elatively reater mount f he olar
radiation is reflected onto the tank surface raising its temperature; and,
thus, producing the larger contrast value.
Table 10. mpact of a second background on detection range
TAWS Range km)
Scene
complexity
Vegetation
background
only
Vegetation Vegetation
background background
primary with secondary with
snow background snow background
secondary primary
None 22.46 22.46 23.30
L ow 21.46 21.36 22.22
Moderate 21.46 21.07 21.95
High 20.39 19.87 20.81
A s discussed in section 2.2.2, rather than a second choice of a background
fo r a what if capability to see two separate background results in a single model un, dding econd background erves o dd lutter o he
scene, with he reatest mpact n onjunction with he election f moderate-or-high complexity. Note that when multiple backgrounds are selected, there ca n be a difference between detection ranges based on low and moderate omplexity, s well s ven reater ifferences between moderate and high complexity than seen with just one background. his is easonable result, and has been highlighted in the updated T A W S user ocumentation Version .1 nd reater). n cases with multiple
backgrounds, the clutter temperature Or will not necessarily be consistent fo r each time in a 24-hour run, or between cases with varying weather
input, ue o he ifferential eating f he arget nd he ifferent
background types. However, to get an idea of the total effects possible
under different clutter amounts, a single example at a single time shows a
32
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 38/55
decrease in detection range from 19.69 km based on a single background
and no cene omplexity to 5.75 km based on three backgrounds and
high-scene complexity.
3.3.2.2 Weather Effects
Although one purpose of this report is to compare the predicted target-
detection anges utput by he AW S nd Acquire, t s lso worth examining he ffects f weather n he redicted anges. arying
visibility, relative humidity, cloud cover, fog, and precipitation resulted
in imilar impacts o etection ranges in both the T A W S and Acquire,
consistent with at least qualitative expectations of how these atmospheric
properties affect IR sensors. [20,21]
The
ollowing
xamples
re
ased
n
airly
ealistic
winter-weather scenarios fo r Seoul, Korea. limatological temperature values varied 6 °C over he 4-hr eriod, with ess ariation n ew-point emperature
values, resulting in a typical increase in relative humidity values in the
early morning and lower relative humidity in the afternoon. hese same temperature values were used whether or not clouds, fog, or precipitation
were included. hese examples ar e based on low-scene complexity and
one background in the TAWS, and N5o = 0.5 in Acquire. s discussed above, ncreasing he omplexity evel o igh rovided omewhat shorter detection ranges in most cases, but yielded similar results in terms
of weather mpacts n etection-range redictions. able 1 ists he noontime detection ranges calculated by the TAWS (number in upper left
of ell) nd Acquire number n ower ight f cell) or ac h weather scenario. s previously discussed, the Acquire values ar e not resolution limited to th e T A WS' appropriate maximum of 32 km in cases with high
visibility and no fo g or precipitation. Other cases, which include fo g or precipitation seem to return detection ranges around 7 km using Acquire, while he AW S rovides much ower etection anges. therwise,
Acquire esults how imilar mpacts based n weather nd iurnal
effects s he AWS, nd ubsequent iscussions il l ighlight he
specific-weather impacts using the T A W S data.
33
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 39/55
Table 11 . Detection ranges as a function of various weather scenarios
Detection Ranges (km) TAWS
Acquire
RH Precipitation
30% none
50% none
80% none
10 0 % light 100% heavy
80% snow
90% light rain
90% moderate
rain
Cloud Cover
S - 4 - > cß c d u 1 1 0
> o
- 4 - » e n c d C D > O
c
4- * c d Ö > o
•S i c d u >
3 ovec
le r
ovec
le r
ovec
le r
ovec
? 2 1 1 8 2 0 1 8 1 9 1 6 5 5 2 2 2 5 3
1 « 1 6 1 7 1 5 1 6 1 4 5 5 2 2 2 5 S 5 3 2 2 8 3 0 2 7 2 8 2 4 1 2 1 0 4 5
? 9 7 5 2 7 3 2 2 3 2 4 2 0 1 0 9 4 5
1 0 3 2 3 2
3 1 3 2 2 8 1 8 1 5 7 6
• 1 H 3 2 3 2 3 2 2 9 3 0 2 5 1 5 1 4 7 6 tß > 1 5 3 2 3 2 3 2 3 2 3 2 2 9 2 1 1 8 1 0 6
6 3 2 3 2 3 2 3 2 3 2 2 7 1 8 1 6 9
3.3.2.1 Visibility
Although R ensors re seful or etecting argets t anges
beyond the distance visible to the unaided eye, he same
atmospheric onstituents, which educe isibility, will reduce R
detection anges, lthough o ifferent mount. Figure 0
illustrates he mpact of reduced visibility on TAWS. Under the
benign conditions of low-relative humidity (around 30 percent at
noon), clear skies, and 15 km isibility, TAWS provides a
maximum ensor-detection range imit of 32 km throughout the
24-hour period. The other examples with clear skies result in an
increase n etection anges during aylight ours, s olar
loading heats the target more than the background, resulting in a
greater T or ifferential n arget/background emperatures).
Visibility decreasing from 15 to 10 km generates rmnimal impacts,
but oing rom 0 o m auses 5 ercent ecrease n
detection range, while going from 5 to 2 km results in a 35 percent
decrease.
34
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 40/55
Figure 10. Visibility impacts on the TAWS- detection ranges.
— — — — -— — . ? <_- ^—, D E X ^v S ? - a c o
yf ^ V > yT N^_
^^f — —
DC — 15
10
R H = Low Clear Skies
isby=15km isby=10km isby=5km isby=2km
ID 11 12 13 14 15 16 Local Time IB 20 21 22 23 54
3.3.2.2 elative Humidity
Atmospheric moisture absorbs IR signals. ecause the time series
of emperature alues s onsistent n ac h ase, hanging he
relative humidity is equivalent to changing the amount of water vapor or absolute humidity available fo r attenuation of the sensor
signal. igure 1 hows n xample f ncreasing umidity
resulting n ecreasing etection anges. ince he old winter
temperatures used in these cases do not allow the atmosphere to contain ignificantly more moisture, the etection range s only
decreased y 0 ercent s umidity s aried rom ow
(approximately
0
ercent;
quivalent
o
/m
3
)
o
igh (approximately 80 percent; quivalent o /m3). owever,
comparable runs made using summer temperatures show noon-
time etection ange redictions f 8 m n ow umidity
(approximately 30 percent; equivalent to 7 g/m3) falling to 13 km
in igh umidity approximately 0 ercent; quivalent o
19 g/m3), reflecting more than a 50 percentdecrease.
35
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 41/55
C le a r S k i e s Visby=5km Figure 11. Relative humidity impacts on the
TAWS-detection ranges.
35
— 0 i 0)
g>25 is a
1 3 0 a a ; a 15
10
RH=Lo w (30 a t n o o n )
H = M o d e r a t e ( 50 at n o o n )
H=High ( 80 a t n o o n )
2 3 4 5 6 7 8 9 2 3 4 Lo c a l T i m e
3.3.2.3 ky Cover
The primary effect of cloud cover above the sensor path is based on the
cloud's nfluence n eating nd ooling f he arget nd he
background. s shown in figure 12 , this example exhibits slightly longer
detection ranges during the night with overcast skies compared to clear
skies. lthough both the arget and the background temperatures are
affected by he louds, o hat he T emains maller han when o clouds ar e present, less radiational cooling allows the thermal imager to
detect the warmer target at slightly longer ranges than under clear skies.
Cloud cover has a reater impact during the day, as the reduced solar
loading provides only a 5 percent detection range increase, while the case with no clouds produces a twenty percent greater detection range during
the daytime.
36
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 42/55
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 43/55
Figure 13 . og
impacts on the
T A W S - detection
ranges.
RH=High Overcast Visby=2km
-LIGHT FO G (Radiat ion)
—HEAVY FO G (Advect ion)
9 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 1 2 2 2 3 24 Local Time
RH=High Overcast isby=5kn[ Figure 14 . Precipitation
impacts on the TAWS-detection
ranges.
LIGHT RAIN (2 m m / h r )
-MODERATE RAIN (5 m m / h r )
- S N OW
2 0 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 2 1 22 23 24 L o c a l Time
8
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 44/55
4 . onclusions
It is important that Acquire be merged into the T A W S so that the services
ca n redict arget cquisition f ground argets using ecognition nd
identification in addition to detection. he incorporation of the Acquire
SP M into the T A W S is scheduled fo r the TAWS Version 3, which will be
released or se n 001. hese omparisons f the arget-acquisition
range output from the urrent version of the T A W S with output from
Acquire provide positive feedback on the benefits of this enhancement to
the AWS , while maintaining he xisting nterfaces nd roviding
comparable arget-detection range predictions.
he primary benefit
of this nhancement il l e he bility o pecify arget-acquisition
discrimination levels, including detection, recognition, and identification.
Decisions will need to be made on an efficient and appropriate selection
of N5o fo r use in the Acquire SPM.
Because the cases examined in this report ar e limited to a winter scenario
in Korea, specific quantitative results of the selected weather parameters'
impacts on target-detection range cannot be generalized to al l situations.
However, these cases do highlight the importance of considering accurate
atmospheric onditions n arget-acquisition redictions. esults
showing maller weather mpact o Acquire-detection anges han
predicted sing he urrent AW S PM nder onditions f og r precipitation warrant additional investigation.
39
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 45/55
References
1 . auter, D., M . orres, .D . Brandt, t al., The ntegrated Weather Effects Decision Aid: A Common Software Tool to Assist in Command
and Control Decision Making/' roceedings f the Command Control Research Technology Symposium, Newport, RI, (June 1999).
2. ouveia M.J., J.S. Morrison, R.B. Bensinger, et al., T A W S and NOWS:
Software Products fo r Operational Weather Support, roceedings of the Battlespace Atmospheric an d Cloud Impacts on Military Operations Conference, Colorado State University, Fort Collins, Colorado, (25-27 April 2000).
3. Acquire Range Performance Model for Target Acquisition Systems, Version 1
User's Guide, U.S. Army CE COM Night Vision and Electronic Sensors
Directorate Report, Ft. Belvoir, VA , (1995).
4. ohnson, J., Analysis of Image Forming Systems, Proceedings of the Image ntensifier ymposium, .S. rmy ngineer esearch nd Development Laboratory Ft. Belvoir, (A D 22 0 160), (October 1958).
5. chmieder, D.E., and M.R. Weathersby, Detection Performance in Clutter with ariable Resolution, IEEE Transactions on Aerospace and Electronic
Systems, Version 19 , pp. 622-630, (July 1983).
6 . chmieder, D.E., M.R. Weathersby, W.M. inlay, t al., lutter an d Resolution ffects n Observer Static Detection Performance, U.S. Air Force
Wright Aeronautical Laboratory, echnical Report AFWAL-TR-82-1059 (A D B071777), Wright-Patterson AFB, OH, (June 1982).
7. ouart, C.N., M.J. Gouveia, DA. DeBenedictis, et al., Electro-Optical Tactical ecision id EOTDA) ser's anual, ersion , echnical Description, U.S. Air Force Phillips Laboratory, Technical Report PL-TR-
93-2002 (A D B172088L), Hanscom AFB, M A , (June 1994).
8. Army Modeling an d Simulation Office, Standards Category Acquire, web site, http://www.amso.army.mil/.
9. ohnson, J., and W.R. Lawson, Performance Modeling Methods and
Problems, Proceedings of the IRIS Specialty Group on Imaging, pp.
105-123, Infrared nformation nd Analysis enter, RIM, Ann Arbor, M I,
(January 1974).
10 . Mazz, J., Acquire Model: Variability in N50 Analysis, Proceedings 9th
Annual round arget odeling nd alidation onference, ignature
Research, Inc., Calumet, M I, (August 1998)
41
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 46/55
11. Howe, J.D., Electro-Optical Imaging System Performance Prediction, in The Infrared an d Electro-Optical Systems Handbook (U), volume4., M .C. Dudzik, Ed, Infrared Information Analysis Center and SPIE Optical Engineering
Press, (1993).
12. erk, ., .S. Bernstein nd . . obertson, MODTRAN: Moderate Resolution Model for LOWTRAN , U.S. Air orce Geophysics
Laboratory, echnical Report, GL-TR-89-0122 AD A214337), Hanscom
AFB, M A , (1989).
13. hirkey, R.C., L.D. Duncan and F.E. Niles, The Electro-Optical Systems Atmospheric ffects ibrary, Executive Summary, Atmospheric ciences
Laboratory, echnical eport, SL-TR-0221-1, White ands issile
Range, NM, (October 1987).
14. iggins, G.J., .F . Hilton, . hapiro, t l. , perational actical Decision Aids (OTDA)s U.S. Air Force Geophysical Laboratory, Technical
Report, GL-TR-89-0095 (A D B145 289), Hanscom AFB, M A , (March 1989).
15. Higgins, G., D.A. DeBenedictis, M.J. Gouveia, t al., lectro-Optical Tactical Decision A id EOTDA ) Final Report, U.S. Air orce Geophysics Laboratory, echnical eport, L-TR-90-0251 I) AD 153311L),
Hanscom AFB, M A , (September 1990).
16 . chmieder, D.E., echnique for ncorporating High Resolution arget Signature Predictions into Sensor Performance Models, U.S. A ir Force Wright Aeronautical Laboratory, Technical Report, AFWAL-TR- 87-1055, Wright-
Patterson AFB, OH, (July 1987).
17.
chmieder, D.E., Interim Contrast Categories for 'Rule of
Thumb' Clutter
Computation, Georgia Technical Research Institute, Technical Transmittal,
TT-4828-004, Atlanta, GA , (March 1988).
18. ouart, C.N., M.J. Gouveia, D.A. DeBenedictis, et al., Electro-Optical Tactical ecision id EOTDA ) ser's anual, ersion , echnical Description, ppendix , hillips aboratory echnical eport PL-TR-93-2002 (II) (A D B171600L), Hanscom AFB, MA, (January 1993).
19 . oist, .C., lectro-Optical maging ystem erformance, CD
Publishing, Winter Park, FL , (1995).
20 . uantitative Description of Obscuration Factors for lectro-Optical an d Millimeter ave ystems, ilitary andbook, OD-HDBK-178(ER),
(July 1986).
21 . Federation of American Scientists, METOC Effects Smart Book (U), web site, http: / /www.fas.org/spp /military/program/met/metocsmarttbook.htm
42
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 47/55
Acronyms
ID 2D
A MSA A
AFRL
CASTFOREM
EOTDA
ERDL
FLIR
GTRI
IFOV
IWEDA
IR
M DT
M RC
M RT
NVESD
SC R
SPM
TASC
TAWS
T CM2
TDA
TRAC
one dimensional two dimensional
Army Materiel Systems Analysis Agency
U.S. A ir Force Research Laboratory
Combined Arms and Support Taskforce Evaluation
Model
Electro-Optical Tactical Decision Aid
U.S. Army ngineer esearch nd Development
Laboratories
forward-looking infrared
Georgia Tech Research Institute
instantaneous field of view
Integrated Weather Effects Decision Aid
infrared
minimum detectable temperature
inimmum resolvable contrast
minimum resolvable temperature
Night Vision and Electronic Sensors Directorate
signal-to-clutter ratio
Sensor Performance Model
The Analytic Sciences Corporation
Target Acquisition Weather Software
Target Contrast Model 2
tactical decision aid
TRADOC Analysis Center
43
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 48/55
T R A DOC raining and Doctrine Command
TTPF arget Transform Probability Function
W W W orld Wide W eb
44
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 49/55
Distribution
Copies NASA MARSHALL SPACE FL T CTR
ATMOSPHERIC SCIENCES DIV
ATTN DR FICHTL HUNTSVILLE AL 35802
NASA SPACE FL T CTRATMOSPHERIC SCIENCES DIV
CODE ED 411 HUNTSVILLE AL 35812
USARMYSTRATDEFNSCMND
CSSD SL L ATTN DR LILLY
PO BOX 1500 HUNTSVILLE A L 35807-3801
US ARMY MISSILE C M N DA M S M I R D AC A D ATTN DR PETERSON REDSTONE ARSENAL AL 35898-5242
US ARMY MISSILE C M N D
A M S M I RD AS SS ATTN MR H F ANDERSON REDSTONE ARSENAL A L 35898-5253
US ARMY MISSILE C M N DA M S M I RD AS SS ATTN MR B WILLIAMS REDSTONE ARSENAL AL 35898-5253
US ARMY MISSILE C M N DA M S M I RD DE SE ATTN M R GORDON LILL JR REDSTONE ARSENAL A L 35898-5245
US ARMY MISSILE C M N D
REDSTONE SC I INFO CTR A M S M I RD CS R DOC REDSTONE ARSENAL AL 35898-5241
US ARMY MISSILE C M N DA M S M I REDSTONE ARSENAL AL 35898-5253
45
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 50/55
PACIFIC MISSILE TEST CTR GEOPHYSICS DIV ATTN CODE 3250
POINT M U G U C A 93042-5000
N A V A L O C E A N SYST CTR
CODE 54 ATTN DR RICHTER SA N DIEGO CA 52152-5000
METEOROLOGIST IN CHARGE KWAJALEIN MISSILE RANGE PO BOX 6 7 A PO SA N FRANCISCO CA 96555
DEPT OF C O M M E R C E CTR
M O UN T A I N ADMINISTRATION
SPPRT CTR LIBRARY R 51
32 5 S BROADWAY BOULDER C O 80303
N C A R LIBRARY SERIALS NATL CTR FO R ATMOS RSCH PO BOX 3000 BOULDER CO 80307-3000
DAMIPOI WASHINGTON DC 20310-1067
LEANDER PAGE D A M I POB WASHFATGTON DC 20310-1067
M IL ASST FO R ENV SC I OF C OF THE UNDERSEC OF DEFNS FO R RSCH & ENGR R & A T E LS PENTAGON ROOM 3D129 WASHESTGTON DC 20301-3080
US ARMY INFANTRY ATSH CD CS OR ATTN DR E DUTOH FT BENNPMG GA 30905-5090
A IR WEATHER SERVICE TECH LIBRARY FL4414 3 SCOTT AFB IL 62225-5458
USAFETAC DNE ATTN M R GLAUBER SCOTT AFB IL 62225-5008
46
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 51/55
HQ AFWA/DNX
10 6 PEACEKEEPER DR STE 2N 3 OFFUTT AFB NE 68113-4039
PAULTATTELMAN AFRLVSBE
29 RANDOLPH ROAD HANSCOM AFB M A 01731-3010
DRANDYGOROCH N A V A L RESEARCH LABORATORY
MARINE METEOROLOGY DIV CODE 7543 MONTEREY C A 3943-5006
MELANIE GOUVEIA
TASC 55 WALKERS BROOK D R READING M A 1867
US ARMY
MATERIEL
SYST ANALYSIS ACTIVITY
AMXSY ATTNMRJMAZZ A PG M D 21005-5071
US ARMY MATERIEL SYST
ANALYSIS ACTIVITY AMXSY AT ATTN M S R KISTNER APG M D 21005-5071
US ARMY MATERIEL SYST ANALYSIS ACnVITY
AMSXY APG M D 21005-5071
US ARMY RESEARCH LABORATORY AMSRLD 2800 POWDER MILL ROAD ADELPHI MD 20783-1145
US ARMY RESEARCH LABORATORY AMSRL O P CI SD TL 2800 POWDER MILL ROAD ADELPHI MD 20783-1145
US ARMY RESEARCH LABORATORY
AMSRL SSSH ATTNDRSZTANKAY 2800 POWDER MILL ROAD ADELPHI M D 20783-1145
47
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 52/55
US A R M Y RESEARCH LABORATORY
AMSRLIS ATTNJGANTT 2800 POWDER MILL ROAD ADELPHI MD 20783-1197
US A R M Y RESEARCH LABORATORY AMSRL 2800 POWDER MILL ROAD ADLEPHI M D 20783-1145
NATIONAL SECURITY AGCY W 21 ATTN D R LONGBOTHUM 9800 SAVAGE ROAD FT GEORGE G M E A D E M D 20755-6000
ARMY RSRC OF C
ATTN AMXRO GS DR BA C H PO BOX 12211
RTP NC 27009
DR JERRY DAVIS NCSU PO BOX 8208 RALEIGH NC 27650-8208
USARMYCECRL CECRL GP ATTNDRDETSCH HANOVER NH 03755-1290
US A R M Y A R D E C SMCARIMIIBLDG59
DOVER N J 07806-5000
US ARMY COMMUNICATIONS
ELECTR CTR FOR EW RSTA AMSRL EW D FT M O N M O UT H NJ 07703-5303
US ARMY COMMUNICATIONS
ELECTR CTR FOR EW RSTA AMSRL EWMD FT M O N M O UT H NJ 07703-5303
US ARMY DUGWAY PROVING GR D STEDPMTDAL3
DUGWAY UT 84022-5000
US ARMY D U G W A Y PROVING GR D
STEDPMTM ATTN M R BOWERS DUGWAY UT 84022-5000
48
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 53/55
DEPT OF THE AIR FORCE OL A 2D WEATHER SQUAD MAC H O L L O M A N AF B NM 88330-5000
PL W E KBRTLAND AF B NM 87118-6008
USAF ROME LA B TECH
CORRIDOR W STE 26 2 RL SUL 26 ELECTR PKWY BL D 10 6 GRIFFISS AFB N Y 13441-4514
AFMC D OW WRIGHT PATTERSON AF B OH 45433-5000
US ARMY FIELD ARTILLERY SCHOOL ATSFTSMTA FT SILL O K 73503-5600
US ARMY FOREIGN SC I TECH CTR
C M 220 7TH STREET NE CHARLOTTESVILLE V A 22448-5000
N A V A L SURFACE WEAPONS CTR
CODE G6 3 DAHLGREN VA 22448-5000
US ARMY OE C CSTE EF S PARK CENTER IV 4501 FORD AVE ALEXANDRIA VA 22302-1458
US ARMY CORPS OF ENGRS ENGR TOPOGRAPHICS LA B
ETLGSLB FT BELVOIRVA 22060
US ARMY TOPO ENGR CTR
CETEC ZC 1 FT BELVOIR VA 22060-5546
US ARMY NUCLEAR CM L AGCY M O N A ZBBLDG 2073 SPPJNGFEELD VA 22150-3198
US ATRADOC A T C D F A FT M O N R O E VA 23651-5170
49
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 54/55
US ARMY TRADOC ANALYSIS CTR
ATRC W SS R WSMRNM 88002-5502
DTIC 8725 JOHN J K I N G M A N RD
STE0944 FT BELVOIR VA 22060-6218
US ARMY MISSILE C M N D
A M S M I REDSTONE ARSENAL AL 35898-5243 US ARMY DUGWAY PROVING GR D
STEDP3 DUGWAY UT 84022-5000
W S M R TECH LIBRARY BR
STEWS M IT WSMRNM 88002
US MILHARY ACADEMY DEPT OF MATHEM ATICAL SCIENCES
ATTN M D N A MAJ DON ENGEN
THAYER HALL WEST POINT NY 10996-1786
ARMY MODELDvTG & SIMULATION OFFICE
DDCSOPS ATTN D A M O ZS 40 0 ARMY PENTAGON WASHINGTON DC 20310-0450
US A R M Y RESEARCH LABORATORY
AMSRLCIEW ATTNDRSHIRKEYINFO SC I & TECH DIR WSMRNM 8002-5501
US ARMY RESEARCH LABORATORY
AMSRLCIEW ATTNBSAUTER INFO SC I & TECH D IR WSMRNM 8002-5501
M R RENE CORMIER AFRLVSBLDRC
29 RANDOLPH ROAD HANSCOM AF B M A 01731-3010
50
8/13/2019 Weather Effects on Target Acquisition
http://slidepdf.com/reader/full/weather-effects-on-target-acquisition 55/55
MRDDIXON
TRAC ATTN ATRC W BC W S M R N M 8002
Record Copy
95 TOTAL