UNIVERSITY OF CALIFORNIA, SAN DIEGO
w
UC SAN DIEGO LIBRARY
..
...........
ww
WS
.
...
w wwwwwwwwwwwww
.
SER
...
Wwwwwwwwwwwwww
......
....
..
WWWSPA WAWA
w
wwwwwwwww
wwwwwwwwwwwwwwwwwwwwww
................
KASEGWAY W
.
ww.
......
.
w
......
.
..
3 1822 04429 0773
ATMOSPHERIC OPTICAL SYSTEMS
TECHNICAL NOTE NO. 207A
JUNE 1988
bu
Offsite
(Annex-JO
rnals)
QC
974.5
. T43
no. 207A
.....
AUTOMATED VISIBILITY & CLOUD COVER
MEASUREMENTS WITH A SOLID-STATE
IMAGING SYSTEM
(The As-Built First Generation)
R. W. Johnson
W.S. Hering

UNIVERSITY
CALIFORNIA
SAN DIEGO
The material contained in this note is to be considered
proprietary in nature and is not authorized for distribution
without the prior consent of the Marine Physical Laboratory
and the Air Force Geophysics Laboratory
-
www
Contract Monitor, Dr. H.A. Brown
Atmospheric Sciences Division
14
4
2
CORN
UK E
N
BAB
SCRIPPS
INSTITUTION
Prepared for
Air Force Geophysics Laboratory, Air Force Systems Command
United States Air Force, Hanscom AFB, Massachusetts 01731
OCEANOGRAPHY
MARINE PHYSICAL LAB San Diego, CA 92152-6400
SIS
****

*
....
...........
.....
...................
KVIRSIKUWA
w
A
N
ADAPA
wwwwwww..
UNIVERSITY OF CALIFORNIA, SAN DIEGO
3 1822 04429 0773
TECHNICAL NOTE NO. 207
AUTOMATED VISIBILITY & CLOUD MEASUREMENTS
WITH A SOLID STATE IMAGING SYSTEM
(The As-Built First Generation)
This Technical Note contains the annotated vu-graph representations, a sub-set of which were presented
to the Transportation Research Board Annual Meeting in Washington, D.C. on 11 - 15 Jan 1988.
w wwww
ww
.....
.................
........
...
.
.
....
.
WWW
.
MICE
..
.
.
.?
???
.....
..
..
.
.
..
.
.
.
.
.
.....
.
.
.
..
..
.
........
.
.
..
.
.
.
..
.
.
.....
...
.
.
.
..!!!
.
.
.
.....
.
....
....
..
.
.
..
..
.......
.
...
.
.
.
."
...
.
.
.
............
.
.
.
.......
...
1.1.
..
.
O
#.
.
.
..
.
.
.
.
.
.
.
ot.U
.
!
.
.
.
.
.
!
.
.
WOOD
+ + + . .
.
!
.
.
.
.
.
.
!
.
.
.++
+
UV
O
DO
DOLORO
W
OORDOPO
#100111 PUUDUVUUD
.++++
+
.. . !
200
POPUOLO
1010110
..!O
O
OUUOVO
+
.
.
#t
.
.
.
.
.
.
.
.
.
.
t
t
I
*
.
R
.
•.•.
..
.
0.
00
.
+0+*500001
. . + 1 +
0 + 0
• ..•.
9
! !
++
+
+
!
+
+
*
Out
.
....
.
....."
.
W
++
DO
2012
OOOO
.
.
2011..
!
+
..
.
.
OO
....
...O
.
5
O
..
SINVunosdo MONS NIVY_12
THOSH asya ano na
O
OTO+
1
!
.
0 .
COU
OOO10005
D
.
..
D
.
+
0.100.1
...
0
.
0
:
I !
.000.
+
.. . ..
. .... .. ...
OOOOOOOOOOOOOOOOOOO
!
....
.
.
0
1
.
.
Uttttt00.00
+
06.
00000000
ttu
+
.
1
+
.
+
.....
.
.
.
.
1
.
.
.
.
+0.10
.
.
.tttt
....
......
.
..
+
.
..
........
.
.
..
..
.
0
+
1
!
0
.
.
O
.
!
01.
5
.
0
UUD
!
.
.
0
$
.
.
OOOO!
.
.......
......
..t.000
.
:
.
..
0
.
$
.
.
..
...
..
..
!!
050
00
...
.
.
.
........
.
..
$
.
0
0
.
.
0
010000000000OOO
...
.
.
.
.
.
.
100.00
.
..
.
t
$
....!
DODO
.
O
.
#
.
2. CLOUD COVER
1. VISIBILITY
$
#
.
...
+
.
$
.
.
.
.
.
••••••••
+
.
...
to.
.
.
010
...
0
1000011
SA
.
.
.
.
..
.
.
5
.
.
$0.00
..
.
.
..
.
.
0
.
.
.
!
+
.
.
OOOO!
+0000
!
.
+
.
alized search, detect and identify scenarios.
cover. The multi-spectral imagery is equally applicable in more gener-
detection, cloud-free arc determinations, sector visibilities and total cloud
dependent, but current algorithm development is slanted toward cloud
the-atmosphere in which this sensor system operates are of course task
The quantities most desired for describing the optical state-of-
..
.
.
.
1
tel
100
000000
.
1
u
0.
0
.
+
0
0
_COULDW010010.
012
!12
.
1 1100010010..+1
+ 1
.
0_02_W
.
.
O_11_21.
#
. ..
#
.
7
.
U
1
.
.
.
.
KNXOOOOO
+
.
.
.
#
..
.
.
.
COMPONENTS OF ATMOSPHERIC STATE SPECIFICATION
+
.
.
.1.
.
.
1
0
.
.
DEN
.
.
DOLO
+
+
..
+
......
+
.
.
.
.
.
.
00000000
..OOOOOOO
. . !
SNORLIANO NYONO NON OL SNOISN
2011
:
28 aona
.
+
.
#.*.
*
#
.
.
.
.
.
..
.
.
..
.
.
.
.
.
..
0
tot
no
10600
000
6
:
4. METEOROLOGICAL REPORTING AND FORECASTING
1000
.
!
.
.
.
.
.
.
O
+
automatic assessment of the optical state of the atmosphere.
ing a visible spectrum, computer controlled imaging system for the
Laboratory, in support of several DoD sponsored programs, is develop-
The Atmospheric Optical Systems Group of the Marine Physical
.
#
ODO
2006
3. UNMANNED AIRFIELD REPORTS
.
...
...
....
OU.
06.
..
*
*
* *
.*. .
*
WWW
.
*
*
*
+.
.
..
.
000+
..
1000000
St.
SUOI NON INNOHAN
JUSOH
.
AIRBOANE AND SURFACE TACTICAL PLANNING
.
.
FIGURE 2
.
.
..
.
!
!
!
!
!
....
!
!
1
E
!
!
U21
.
.
.
...
..
.
.....
..
.
..
.
SITUATIONS REQUIRING ATMOSPHERIC STATE SPECIFICATIONS
tot
.
.
.
.
.
.
!
.
.
.
•
.
$
.
.
.
+
.
.
.
.
1
+
.
O
.
.
.
.
.
.
.
6001
.
.
.
.
.
+
.
.
.
+
.
.
O
.
.
.
.
!
!
+
.
.
.
.
!
.
.
.
.
0
.
.
.
.
.
.
+
.
.
.
.
.
+
.
.
.
.
.
.
.
.
.
.
.
+
.
6
.
.
.
0.0.
.
.
.
+
1
11111
1 1
.
# #
#
! ! # + . .
.
..
1
1
1
1
1
1
.0
+
+
+
.
0
.
University of California, San Diego
Scripps Institution of Oceanography
Institution of Oceanography.
Diego, Calif. and is administratively within the jurisdiction of Scripps
organization within the University of California system. It is located in San
The Marine Physical is a 100 person multi-disciplined research
Atmospheric Optical Systems
MARINE PHYSICAL LABORATORY
VISIBILITY
VISIBILITY
VISIBILITY
MSIBILITY
VISIBILITY
VISIBILITY
VISIBILITY
VISIBILITY
VISIBILITY
MSIBILITY
VISIBILITY
FIGURE 1


SAKA
|
E/O SYSTEM 5 - DISCUSSION OUTLINE
1. SYSTEM SCHEMATIC & PHYSICAL CHARACTERISTICS
FIGURE 3
2. BASIC CONCEPTS
The following annotated, illustrations describe the general
operational characteristics of the subject Electro Optical Camera System
3. COMPUTERIZED VISIBILITY ESTIMATES FROM VIDEO IMAGERY
4. TWO-COLOR CLOUD DETECTION & ANALYSIS TECHNIQUES
FIGURE 4
E/O CAMERA SYSTEM FOR
OPTICAL STATE OF THE ATMOSPHERE ASSESSMENTS
The schematic drawing of the multi-spectral imager illustrates
those features of the camera system that contribute to its general utility
as an automatic cloud detection system.
Fishoyo Lens Assombly
(174* FOV
.
.
..
.
.
WWW
.com
.
First, the Filter Changer Assembly is shown to contain two
independently controlled filter locations on the optical axis. In actuality,
this assembly contains two filter wheels, each holding four separate
filters in addition to the five lens optical relay. Each wheel can position
any one of its four filters into the optical path under either manual or
computer control. Thus, both spectral band and flux level can be
controlled either manually or by pre-determined computer program. This
is important to the cloud discrimination algorithm illustrated later in this
series.
www
WWW
.
...
AY
...
L5
VW
ZA
ZZZZZZAZ
Filter Changer Assembly
2710 Camera Assombly
Second, the Fisheye Lens Assembly allows the accumulation of
whole sky imagery in a single video frame which contributes to the
simplification of subsequent spectral ratioing techniques and related
image processing functions.



Image Acquisition & Analysis System
Hardware Block Diagram E/O System 5
SONY PVM 1271Q
MONITOR
FIGURE 5
GE 2710
SOLID STATE
VIDEO
CAMERA
TMI COMPUTER
(IDMAT CLONE)
The major E/O SYSTEM 5 sub-assemblies are physically inter-
related as illustrated in this block diagram. The as-built configurations of
the exterior sensor housing and the interior electronic control console are
illustrated in the figures which follow.
AUTOMATIC
EQUATORIAL
SOLAR
OCCULTOR
ASSY.
VIDEO
IMAGE
PROCESSING
SUB-SYSTEM
(ITI FG 100)
ARCHIVAL
1/0
SUB-SYSTEM
K(SEAGATE 65 M bytel
H.D.)
REMOTE
CONTROLLED
IRIS
ASSY.
EXABYTE EXB - 8200
2.2 G byte 8mm CARTRIDGE TAPE SYSTEM
This diagram illustrates the system as used for automatic cloud
cover archival and analysis. It is somewhat more powerful than the
system required for visibility determinations.
ANALOG
ACCESSORY
CONTROL PANEL
REMOTE
CONTROLLED
OPTICAL
FILTER
ASSY.
STOWED KEYBOARD
EXTERIOR SENSOR
INSTALLATION
INTERIOR CONTROLLER
INSTALLATION
FIGURE 6
The as-built fisheye system used for Whole Sky Imagery. Pro-
vides weatherproof housing and solar occultor forunattended operation.
Yields multi-spectral imagery customized for automatic cloud descrimi-
nation studies.
FIGURE 7
und
REAL ACTS*** OWN AN
The system control console required for automatic cloud discri-
nation illustrated here photographically. This highly efficient package
requires only 28 inches of panel height in a standard 19 inch instrument
rack. It provides IBM/AT compatible computer capability, image proc-
essing, archival, and display characteristics as previously summarized
in the preceeding block diagram.
COLOR MONITOR
ZENITH
MODEL ZVM-136
FIGURE 8
ACCESSORY CONTROL PANEL
MOTOR DRIVE MANUAL OVERRIDE
COMPUTER
ZENITH
MODEL ZF241-81
PRECISION ROTARY TABLE
DAEDAL INC.
MODEL 20801 P
.... .........
DAEDAL MODEL MD-23
USTEPPER MOTOR DRIVE
The E/O SYSTEM 5 instrumentation is used in a modified con-
figuration to address the task of automatic visibility determinations. This
alternate configuration is illustrated in the figures which follow.
IMAGE
PROCESSING
SUB-SYSTEM
W
ITI PEG-8
PC VISION, REVC
ARCHIVAL
& VO
SUB-SYSTEM
640 kbyte RAM
360 kbyle FLOPPY
1.2 Mbyle FLOPPY
40 Mbyte H.D.
(REMOTE UNIT).
EVEREX FT-60
1/4" STREAMER
W/CONTROLLER
This diagram illustrates the system as used for automatic deter-
minations of sector visibilities and related local apparent contrasts.


DAEDAL MC5000
INDEXING MOTOR
CONTROL BOARD
WEATHER PROOF
ENCLOSURE
MODEL EO3-3900
VIDEO COAX
STOWED KEYBOARD
COMPUTER TO TABLE HOMELMT SWX)
SONY
AC PWR
DISTA
RECIAC
FAN
+15VDC
CAM. PWR
AUTOSALON
COSTACAR SORT
BRED EXTENDER
es EBS
VIDEO CAMERA
CID 776
115VAC ACCESSORY POWER
COMPUTER TO MD-23 (DIR. & STEP
FIG. A
AUTOMATIC VISIBILITY SYSTEM
HARDWARE LAYOUT - 1 MAY '88 AS-BUILT
INTERIOR CONTROL CONSOLE

FIGURE 9
The as-built horizon scanning system used for the automatic
determination of local sector visibilities. The weather proof housing
contains a precision rotary table upon which is mounted the solid state
video camera and it's telephoto lens. Computer control of the table
rotation enables reproducible horizon scans and the computerized
assessment of pre-selected target properties.
FIGURE 10
The rack mounted version of the system control console re-
quired for automatic visibility determinations is illustrated here photo-
graphically.
The rack mount configuration is very similar to that used in the
automatic cloud discrimination mode, but is operationally much simpler.
Note that this illustration of the rack mount configuration has
substituted a TMI 2001A computer for the original Zenith A-241 to
simplify the mechanical assembly.
Wh
Sex
S
RELATIVE SPECTRAL RESPONSE
Oriel Broadband Interlerence Filters with GE2505 Camera
www
blue
groen
red
Peak Bandwidth -
À @ 0.5 Peak
456 56.7
550 54.7
654 71.1
FIGURE 11
Spectral bands associated with automatic assessment of at-
mospheric optical properties.
Relative Response, S(2)T()
The blue and red bands are used for the discrimination of cloud/
no cloud regions, and the green band is used in the determination of
surface sector visibilities.
0.0 R
400
450
500
550
600
650
700
750
800
Wavelength (nm)
SOLAR
SPECTRAL
IRRADIACE
FIGURE 12
ww
w
.
....
.
.
..................
ATMOSPHERIC OPTICAL EFFECTS
"
.
...
..
.
.....
......
...
.
..........
........
...
...
DIRECT & OIFFUSE
SURFACE IRRADIANCE
SCATTERED SKYUGHT
(PATH RADIANCE)
..
The process of daytime visibility determination either instrumen-
tally or by a human observer ultimately involves the resolution of the
distance from pre-selected targets that the apparent contrast, Cr,
reduces to some minimum (threshold) value needed for detection. The
elements to be addressed in the solution of this or any target detection
and identification process are illustrated in the adjacent flow chart.
SOLAR
NEW
ANGLE
GEOMETRICS
BACKGROUNO
REFLECTANCE
OBJECT
REFLECTANCE
ا 4,اما
The notation is adapted from Duntley et. al (1957), "Image
Transmission by the Troposphere 1", JOSA, 47, 499-506.
INHERENT
BACKGROUND
RADIANCE
INHERENT
OBJECT
RADIANCE
ATMOSPHERIC
RADIANCE
TRANSMITTANCE
APPARENT
OWECT
RADIANCE
APPARENT
BACKGROUNO
RADIANCE
NHERENT
CONTRAST
APPARENT
CONTRAST
The stepwise derivation of the fundamental equations for con-
trast transmittance as given by Duntley, et. al (1957) is likewise shown
schematically in the adjacent flow chart, along with the definition of the
individual terms in the expressions.


Coslehoo Lollobo
Cp = (hp-b
CONTRAST
TAANSMITTANCE
to = C/C. = lololololt
Visual Range vs. Apparent Contrast Relationships
*
*
*
FIGURE 13
v
in{s(e.) /[1-(8.) (1-s)]}
in{s ()/(1-() (1-s)]}
The computations implied in the preceeding flow chart can be
illustrated by the corresponding diagnostic equation for the calculation of
visual range, V, in the form shown in the adjacent figure.
Where
Specified: r, €
Estimated: S,CO
Measure: Cr
Derive: V
E
= the threshold contrast for detection
S = Lg/blo = the so called "sky-ground ratio"
(Duntley 1948)
Under optimum conditions of horizontal homogeneily, this diagnostic expression reduces
to the familiar Koschmieder relationship whenever one uses black targets viewed against
a clear, horizon sky, in which case,
h
= the point source function (equilibrium
radiance) of the directional path
radiance L*r.
Co=-1
and
S = 1.
Equilibrium Radiance Relationships
FIGURE 14
bbs = 6Lo T, + 3 (1 - T)
The notational equivalencies between Duntley's Equilibrium
Radiance, Lq , and the Chandrasekhar's Source Function, 3 , are
illustrated in the adjoining figure. The source function, Lq , may be
estimated from the clear-sky horizon radiance for a path of sight having
the same scattering angle with respect to the sun as the target line of
sight.
= Apparent Radiance
= Inherent Radiance
Ref:
= Radiance Transmittance (over pathlength "p")
Duntley, S.Q. (1948), "The Reduction of Apparent Contrast by the
Atmosphere", JOSA, 38, 179-191.
= Equilibrium Radiance (Chandrasekhar's Source Function)
Chandrasekhar, S, (1960), "Radiative Transfer", Dover Publications,
180 Varick, N.Y. 14 N.Y.


Under meteorlogical conditions characterized by
horizontal homogeneily and increasing opacity
as T, + 0
then bLs → 3
s
FIGURE 15
Sky Ground Ratio, S
Duntley (1948) · Duff (1972)
*
|
*
SA
The Sky Ground Ratio illustrated in the adjacent figure is a
concept established by Duntley (1946 & 1948), repeated by Middleton
(1952), and discussed at length by Duff (1972). It is a powerful concept
for modelling and computational techniques but it is difficult to determine
experimentally, and is often misused.
Sky Ground Ratio, s, =
Equilibrium Radiance
Inherent Background Radiance
Lg (2,0,0)
blo (21, 0, 0)
Ref:
•
Duntley's Equilibrium Radiance, Lg,
Chandrasekhar's Source Function, 3
•
Equilibrium Radiance & Inherent Background Radiance must be defined for the same
Detector/Target path of sight.
Duff, E.A., "Atmospheric Contrast Transmission Application to the
Visual Detection and Electro-Optical Lock-on Problem", Master
of Science Thesis, School of Engineering Air University, USAF,
Wright-Patterson AFB, Ohio (1972).
For some paths of sight, Equilibrium Radiance cannot be measured directly.
•
•
Clear Horizon Sky radiances L* (z, 90, $) are direct measures of
Equilibrium Radiance La (7, 90, ) along these horizontal paths.
Duntley, S.Q., Visibility Studies and Some Applications in the Field
of Camouflage, Summary Tech. Rept. Div. 16, NDRC, Vol. 2,
NTIS No. AD 321 102, (Columbia Univ. Press, New York, 1946).
•
A Clear Horizon Sky radiance is the Equilibrium Radiance for any path of sight
having the same scattering angle B from the sun.
Duntley, S.Q., "Reduction of Contrast by the Atmosphere", JOSA,
38, 179-191 (1948).
FIGURE 16
The general solution to the expression for V/r shown previously
is illustrated here for the case where the threshold contrast is .05 and the
inherent contrast is -1.
VISUAL RANGE / TARGET RANGE (Log Scale)
* Note in particular the changes in calculated visual range
associated with departures of sky-ground ratio from the
clear horizon-sky background value of 1. In general, S
varies from about 0.2 for very bright backgounds to 5 or
more for low surface reflectance and upsun paths of sight.
The error bars shown for the S=1 result, indicate the
relative errors in the calculated visual range due to a
+20 percent uncertainty in the estimated inherent contrast.
* Note that the errors in calculated visual range due to errors in
estimated inherent contrast increase markedly with increas-
ing apparent contrast of the designated target.
20.06
ADJUSTED TO THRESHOLD CONTRAST - .05
INHERENT CONTRAST - 1.0
14.05
SKY/GROUND
RATIO
>l-
5.0
4.0
3.0
2.5
2.0
1.8
1.6
1.2
1.0
7
¢
¢
À
u
kt
05
APPARENT CONTRAST (Log Log Scale)
* In other words, reliable calculations of visual range are
especially difficult using the apparent radiance contrast of
close range targets in good visibility conditions.


Cr

FIGURE 17
The solid-state imaging system provides highly accurate and
continuously updated measurements of the apparent contrast of all
identifiable objects within its field of view. Early experience indicates that
consistent and reliable estimates of visual range can be obtained from
these target-background ensembles as well as with isolated targets.
o
Whenever practicable, as illustrated in the adjacent figure, the
actual selection of visibility targets should be adjusted for each scene in
accordance with observed conditions. In the absence of ideal dark
targets viewed against the background horizon sky, priority should be
shifted as necessary to objects at distances close to the limit of detection,
where the most consistent and representative calculations of visual
range can be made.
The important consideration is that the objects in a selected area
ensemble be located at approximately the same range and as close in
range as practicable to the limit of visibility.
Diagnostic Display for Automatic Visibility Assessments
FIGURE 18
Target A Target B Target C Target D Target E Target F
AP CR = Cr
RANGE = r
EPSON = E
206
6.000
1.030
.186
5.000
2.173
5.000
1.030
2353 2.209
4.000 4.000
.030 .030
269
4.000
.030
The stylized computer display illustrated here typifies a situation
where an interactive operator has selected six different visibility targets
i.e. A through F, and asked the machine to estimate the sector visibility
for diagnostic assessment.
.
9.030
NNG
OOON
SO
A
.
Note that this example is NOT for the scene illustrated above
and that in full automatic mode, the machine would display only the best
estimate, and not the total selection illustrated.
NNN
co co
XXXX
.50
14.24
13.43
3. 12.80
12.30
11.89
.55
.60
65
.70
ON
Dono
19.03
17.77
16.81
16.06
15.45
14.94
14.52
14.15
13.84
13.56
13.32
.75
.80
.85
13.24 14.9412.93 18.21
12.56 13.72 12.05 16.31
12.03 12.81 11.39 14.97
11.61 12.129 10.86. 13.97
11.26 11.56 10.44 13.20
10.96 11.11 10.09 12.58
10.71 110.7490 9.80 12.07
10.50 10.42 9.559 11.64
10.31 10.15 9.33 11.28
10.14 9.91 9.14 10.97
9.99 9.70 8.97 * 10.70
190
.95
1.00
11/26
11.01
10.79
10.60
10.43
DO YOU WANT A NEW TARGET SET AND NEW IMAGE?
YES, NOx0
23
PERCENT ERROR IN COMPUTED VISIBILITY DUE TO HORIZON CLOUD
100
TEST BOGEY-
TEST BOGEY -
PERCENT ERROR
PERCENT ERROR
PREFERRED OPERATING REGION.
FIGURE 19
PREFERRED OPERATING REGION
The proprietary extraction and analysis algorithms embedded
within the control computer enable a broad variety of self-test and
optimization routines to be exercised at the discretion of the operator.
28
-0,6
0.4 0.8 12 1.6 2 24
. CLOUD-SKY BRIGHTNESS RATIO
-0.8 -.75
INHERENT CONTRAST
.
T.T
An example of several runs performed to evaluate the influence
of a cloudy horizon upon human mimicking visibility determinations is
shown in the adjacent set of graphic displays.
TEST BOGEY
10.4
PERCENT ERROR
Dlagnostic Caveats for Obscured Horizon Sky
Avoid to Minimize Error
• Extremely Bright Clouds
• Clouds Very Near Target Range
• Non-Black Targets
• High Observer to Target Transmillances
PREFERRED OPERATING REGION
1.2
1.6
3.6
2 2.4 2.8 3.2
CLOUD-TARGET RANGE RATIO
ABSOLUTE CALIBRATION PROCEDURE
FIGURE 20
BAFFLE
STANDARD OF
RADIANT INTENSITY
W2= 28540K
CALIBRATE TARGET
Rx-BASO4
LAMP -
EIO CAMERA
ASSEMBLY
+ ROUND TARGET
Whereas many useful algorithms for the determination of atmos-
pheric properties can be devised to require only the input of the relative
values of radiant flux fields, it is generally true that far more redundant and
reliable methodologies are available when absolute values of radiance
are available. Thus, to enable an optimum selection of techniques for
analytic applications, the camera systems described in this note are all
calibrated against standards of radiant intensity traceable to N.B.S. using
standard radiometric procedures in association with optical calibration
facilities established at the Marine Physical laboratory.
.
www
www.
.
.
.
www
www
CONTROL
COMPUTER
-
**.
ww
WWW.
W
ten
NU
-
w
.
...
....
..
..
...
3 METER .
OPTICAL BENCH
& ACCESSORIES
TAPE
ARCHIVE
A schematic representation of those calibration facilities is
shown in the adjoining figure.


.
REGULATED
POWER
SUPPLY
_

FIGURE 21
24
A variety of techniques for objective cloud/clear-sky discrimina-
tion with the solid-state imagery system are being investigated. Single
image (monochromatic) analysis with simple brightness contrast and/or
edge depiction techniques often provides excellent results. The cloud
radiances are in generallarger than the background clear-sky radiances,
especially in the red portion of the visible spectrum.
b) Cumulus clouds
Four examples of monochromatic (650nm) whole sky imagery
taken under various different cloud conditions are shown in the adjacent
figure.
d) Cirrus clouds
Fig. 4-1. Imagery from EO Camera Il for a variety of sky conditions. Images were digitized from video, piped
through image processing, displayed into matrix camera.
FIGURE 22
SINGLE WAVELENGTH DISCRIMINATORS: CAVEATS
a. Sky & Cloud Radiances Vary over Large Excursions
b. Threshold Radiances Require Continual Analytic Updates
c. Dark or Shadowed Clouds mimic Background Sky Radiances
Although the qualitative brightness contrast is readily apparent
in most monochromatic images, obvious difficulties are involved in the
development of objective depiction criteria using single wavelength
image analysis. As summarized in the list of caveats in the adjacent
figure, the cloud and clear-sky radiances vary over broad limits, requiring
complex analytic adjustments to the appropriate cloud/clear-sky thresh-
old radiance values. Furthermore, dense and/or shadowed cloud
radiances are comparable and sometimes become less than the corre-
sponding background sky radiances.
a. Cloud Imbedded in Cloud Background Yields Redundant Boundaries
b. Fails as Scene Approaches Uniform Overcast
09-04-
a) Clear sky
C) Alto cumulus
1. RADIANCE THRESHOLD TECHNIQUES
2. EDGE GRADIENT TECHNIQUES
Barce
BLUE-RED SKY RADIANCE RATIO
FIGURE 23
Visibility
km
Equinox 10:00 AM
Lat 30 N
40
Experimental results clearly demonstrate the effectiveness of
color contrast, as derived from multi-spectral imagery, for objective sky-
cover analysis.
RADIANCE RATIO
• Dedicated systems with the dual filter wheel option can easily be
configured to acquire, sequentially, narrow-passband blue and red
imagery with variable neutral density control.
East
West
0.07
-80
· Objective cloud depiction algorithms based upon the ratio of the all-sky
blue and red radiance fields provide good specification accuracy over a
broad range of sky and visibility conditions as illustrated here. The blue/
red spectral radiance ratios are characteristically near 1 for the white or
grey cloud elements in contrast with significantly larger values for the
background clear sky.
-60
-40
-20 0 20
ZENITH ANGLE
40
60
80
BLUE-RED SKY RADIANCE RATIO
Visibility
km
FIGURE 24
Equinox 7:00 AM
Lat 30 N
RADIANCE RATIO
As in any new experimental system, there are operational
regions which cause difficulty and require additional system sophistica-
tions for adequate performances. For the current configuration of EOI
CAM SYSTEM 5, early morning eastern skies as illustrated in the
adjacent figure show an aggravating ratio cross-over which must be
resolved by the application of supplementary decision algorithms.
West
0.0 ton
-80 -60 -40
East
nonton
-20 0 20 40 60 80
ZENITH ANGLE
.
.
.
.
.
.
.
..



FIGURE 25
Illustrated here is a typical set of cloud depiction imagery. Shown
for comparison, are the observed brightness field acquired with the red
passband filter, and the red/blue radiance ratio field, which has been
false colored for illustrative clarity. The false color key identifies the
relationship between the calculated blue/red ratio as derived from the
systems multi-spectral imagery, and the display brightness used to
activate the monitor.
This depiction of a broken Strato Cumulus cloud field clearly
illustrates the ability of the machine's cloud discrimation algorithm to
mimic the monochromatic depiction which is used as the test bogey.
255
FIGURE 26
DISPLAY BRIGHTNESS
A graphic representation of the false color translation key is
shown on the left. It should be noted that the cloud detection threshold
can be adjusted by the system operator to optimize decisions, or to
investigate other near-threshold phenomena.
The four regions near the ratio value of 1.0, which are blocked
in along the horizontal axis, are typical of those near-threshold regions
currently under investigation.
CLOUD DISCRIMINATION ILLUSTRATION
Broken Strato Cumulus
MONOCROMATIC @ 654 nm
BLUE/RED SPECTRAL RATIO (THRESHOLD @ .85)
SPECTRAL RATIO VALUE
.85 1.3
3.0
65 105
DISPLAYED BRIGHTNESS VALUE
FALSE COLOR CODE
FALSE COLOR CONVERSION CHART
(display brightness vs. blue/red ratio)
1.0
3.0
2.0
SPECTRAL RATIO
(Derived from calibrated blue and red images.)

B & W
.61
.73
FIGURE 27
An example of alternative threshold selections applied to broken
Strato Cumulus imagery are illustrated in this six element display.
Interactive analysis using this variable threshold procedure has
been highly productive in technique development.
Preliminary results indicate that the cloud analysis based upon
the radiance ratio imagery might be extended effectively beyond a simple
yes-no representation. For example, studies are underway to explore
the correspondence between the measured radiance ratios and the
opacity of thin, semi-transparent cloud elements.:
Note that the numerical values shown with each image represent
display brightness values which must be related to blue/red ratios using
the preceeding false color conversion chart.
TYPICAL ANALYSIS FOR VARIABLE THRESHOLD EFFECTS
CLOUD DISCRIMINATION ILLUSTRATION
Thin Cirrus
FIGURE 28
Experimental results have clearly demonstrated the effective-
ness of color contrast, as derived from multi-spectral imagery, for
objective sky cover analysis over a broad range of sky and visibility
conditions.
This image pair illustrates an application of the blue/red cloud
discrimination algorithm to a thin cirrus situation. The basic technique is
the same as that used for the broken strato cumulus example shown
earlier.
MONOCHROMATIC @ 654 nm
BLUE/RED SPECTRAL RATIO (THRESHOLD @ .85)
SPECTRAL RATIO VALUE
.85 1.3
The discrimination of these thin, tenuous cloud types is one of
the most demanding tasks faced by observers, and one for which the
automatic system is demonstrably well suited.
.84
-96
1.07
3.0
65 105
DISPLAYED BRIGHTNESS VALUE
255
FALSE COLOR CODE

B&W
.61
.73
FIGURE 29
Similar to the preceeding Strato Cumulus set, interactive analy-
sis using variable thresholding techniques is applied in this case to the
thin cirrus imagery.
The sublety of the ratio detection discrimination algorithm is
clearly illustrated, evoking an exciting new concept for future research.
.84
.96
1.07
TYPICAL ANALYSIS FOR VARIABLE THRESHOLD EFFECTS
Sky Cover Determination
vs. Spectral Radiance Ratio Threshold
FIGURE 30
. An attractive technique for reducing the data acquisition rates
thus enabling a higher temporal sample rate is illustrated on the left. In
this procedure, rather than digitize and store the entire 512x512 image,
the system is instructed to digitize and store only the data contained in 32
preselected rows and 32 preselected columns. This reduced sub-set is
then operated upon by the red/blue ratio algorithm and the resultant
cloud/no cloud decision for each point along the rows and columns is
superimposed over the original image enabling rapid monochromatic
visual analysis.
Using this technique the existance of the grid overlay indicates
pixel below ratio cutoff and thus identifies that pixel as cloud. Note that
the higher the cutoff selected the better the cloud discernment of thin and
semi-transparent cloud elements becomes.
R = 1.0
R = 1.75
R = 450nm (blue) / 650nm (red)

Sky Cover Determination
vs. Spectral Radiance Ratio Threshold
FIGURE 31
Huwa
650nm
Rx 1.0
In this multi-image display, the use of the grid overlay technique
is illustrated using three separate cut-offs. Note the clearly improved thin
cloud discernment as the identification cut-off is increased from 1.0 to
1.75.
nh,
R 1.50
R* 1.75
Spectral Ratio (R) - 450nm (blue) / 650nm (red)
Sky Cover Determination
vs. Spectral Radiance Ratio Threshold (R)
FIGURE 32
The multi-image display shown here illustrates a comparison
between the multi-threshold false color evaluation format and the mono-
chromatic screen technique.
R = 1.25, 1.50, 1.75
R=1.0
Whereas the false-color imager is a more asthetically satisfying
medium for some analysts, the screen overlay technique also has it's
advantages. This is particularly true when data storage or processing
limitations are being stressed by the operational tasks, or reliable sub-
sets are required for quick quality control assessments.
R = 1.50
R = 1.75
R = 450nm (blue) / 650nm (red)

FIGURE 33
The false color technique is used in this image pair to illustrate
broken alto-cumulus sky cover with wind-blown whispy cirrus clouds
visible in the intervals between the middle-clouds.
The opaque cumulus cloud elements have blue/red ratios less
than 1.25. The associated ratios of the upper thin clouds are in the range
of 1.25 to 1.75.
Sky Cover Determination
A clear sky situation with fresh narrow contrail patterns is shown
in the imagery to the left. Much of the detail in the contrail is lost in the
photographic reproduction; however, these images illustrate that con-
trails which are both optically thin and geometrically small are readily
identifiable in the ratio imagery.
One should note that the threshold cut-offs used for detection
must vary depending upon the directional scattering properties and
optical depth of the contrail element.
Note also the potentially misleading spectral ratio distributions
in the solar aureole region.
Sky Cover Determination
vs. Spectral Radiance Ratio Threshold
650nm
450nm (blue) / 650nm (red)
R = 450nm (blue) / 650nm (red)
1.25 1.50 1.75
Spectral Ratio Code
vs. Spectral Radiance Ratio Threshold
FIGURE 34
650nm
450nm (blue) / 650nm (red)
R = 450nm (blue)/650nm (red)
1.25
1.50 1.75
FIGURE 35
Automatic Observing Systom for Wholo Sky and Horizon Imagory
.
me
en internatione
m renderemo Oncinar
Shuttor
Assointly
-
Fishoye Land
1
Tolaplolo i ens
1911
.
From our experience to date, it is clear that a small, relatively
automatic system for the determination of cloud cover and related
statistical characteristics is now available to the experimental and
modelling communities. The adjacent figure illustrates our conception of
an automatic system designed to accumulate the imagery required to
address not only the specification of local cloud cover, but also the effects
of directional contrast transmittance, and related optical phenomena.
wer
Tavoal Scono
.
w wers
40 (2007
..
................
Heated
Observation
Windows
- I lortzon
Www
.
*
Fior Assembly
- 50/50 Noam Splittor
do + y - 150 ann, all. Tukownola.--
didy - 30 nom, olt. I'ishayo
*
*
*
'
..
menamenwen Locollorizon
Insulatod
Woathor-Prool
Housing
www
IR AFW
mome
www.17
nommen werden Concora
vintuvu
:
-..
w
ve
...
The application of this prototype system to the task of multi-
purpose, local area measurements will begin shortly and should be the
- harbinger of a new standard of consistent and reliable weather related
observations.


Montenentemente come
To Control Panol