Aircraft measurements for calibration of an orbiting spacecraft sensor


Hovis, W.A.; Knoll, J.S.; Smith, G.R.

Applied Optics 24(3): 407-407

1985


Aircraft
measurements
for
calibration
of
an
orbiting
spacecraft
sensor
Warren
A.
Hovis,
John
S.
Knoll,
and
Gilbert
R.
Smith
The
best
method
of
establishing
a
calibration
for
a
satellite
sensor
is
to
make
simultaneous
measurements
along
the
satellite
view
vector
from
a
calibrated
instrument
on
board
a
high-altitude
aircraft.
The
satellite
sensor
prelaunch
radiance
calibration
can
then
be
compared
to
the
in-orbit
radiance
values
produced
by
comparison
with
the
simultaneous
aircraft
data.
Aircraft
measurements
were
made
over
the
Atlantic
Ocean
off
the
east
coast
of
the
United
States,
and
comparison
of
the
prelaunch
and
in-orbit
data
indicated
a
25%
degradation
in
channel
1
of
the
Nimbus
7
Coastal
Zone
Color
Scanner.
I.
Introduction
The
Nimbus
7
Coastal
Zone
Color
Scanner
(CZCS)
after
9
months
in
orbit
began
to
show
signs
of
degra-
dation
in
response,
especially
in
the
blue
channel
cen-
tered
at
443
nm.
The
degradation
became
obvious
when
the
Rayleigh
correction
based
on
prelaunch
cali-
bration
began
to
produce
values
of
upwelled
radiance
at
the
surface
that
were
too
small
after
atmospheric
correction.
The
onboard
calibration
lamps,
used
to
calibrate
the
sensor
only
beyond
the
postoptics
(after
the
large
scan
and
telescope
mirrors),
indicated
no
such
degradation
in
response
and
hence
could
not
be
used
to
recalibrate
the
entire
sensor.
In
an
effort
to
determine
the
degradation
and
recalibrate
the
sensor,
an
aircraft
program
was
initiated
to
fly
on
board
a
high-altitude
Lear
jet
a
double
monochromator
specifically
config-
ured
to
cover
the
wavelength
region
of
the
CZCS
channels
in
question.
The
monochromator
viewed
the
same
surface
area
of
the
ocean
through
the
atmosphere
as
being
viewed
by
the
spacecraft
during
near
simulta-
neous
measurements.
The
results
showed
a
very
strong
spectral
characteristic
to
the
degradation.
A
large
degradation
was
indicated
in
the
443-nm
channel
of
the
CZCS
which
dropped
off
quite
rapidly
in
amount
in
the
longer
wavelength
channels.
This
result
has
implica-
tions
both
for
the
CZCS
and
other
optical
instruments
in
space
that
may
operate
into
the
blue
region
of
the
spectrum.
The
strong
spectral
characteristic
of
the
degradation
will
change
the
shape
of
the
spectral
re-
sponse
of
a
band,
degrading
the
blue
side
more
than
the
red
side,
as
well
as
reducing
the
overall
responsivity
of
the
channel.
The
authors
are
with
NOAA
National
Environmental
Satellite,
Data,
and
Information
Service,
Washington,
D.
C.
20233.
Received
3
August
1984.
II.
Onboard
Calibration
Status
Instruments
such
as
the
CZCS
on
Nimbus
7
and
the
Multi
Spectral
Scanner
(MSS)
used
on
Landsats
1,
2,
3,
4,
and
5
as
well
as
the
Thematic
Mapper
flown
on
Landsat
4
and
5,
all
have
calibration
lamps
internal
to
the
system.
In
all
the
above
the
calibration
lamps
do
not
illuminate
the
large
collecting
optics
and
scan
mir-
rors
that
are
most
exposed
to
the
atmosphere
of
the
spacecraft
(schemes
illustrated
schematically
in
Fig.
1).
The
CZCS
actually
has
two
calibration
lamps
with
a
beam
splitter
so
that
either
one
of
the
two
can
be
used
alternatively.
The
Thematic
Mapper
has
three
cali-
bration
lamps
that
can
be
used
in
various
combinations
to
monitor
the
performance
of
the
detectors
and
am-
plifiers.
The
MSS
also
has
two
calibration
lamps
that
can
be
used
alternatively.
The
first
of
the
MSS
in-
struments
on
Landsat
1
also
had
a
solar
calibration
mirror
that
was
intended
to
monitor
sensor
degradation
by
viewing
the
sun
during
a
pass
over
the
Pole.
The
usable
calibration
mirror
diameter
was
limited
to
0.5
mm
to
reduce
the
direct
solar
image
intensity
on
the
detectors
to
a
level
that
would
not
damage
the
detectors.
This
calibration
method
was
never
successful,
and,
in
fact,
when
the
Landsat
1
MSS
was
activated
on
orbit
21,
—35
h
after
launch,
it
was
found
that
the
solar
calibra-
tion
mirrors
were
profoundly
degraded
by
the
solar
ir-
radiance.
In
the
shortest
wavelength
channel,
500-600
nm,
the
signal
reaching
the
detectors
was
—9%
of
that
expected.
The
degradation
was
somewhat
less
severe
in
the
longer
wavelength
channels
but
nevertheless
so
severe
that
this
method
of
calibration
was
found
to
be
unusable.
After
5
1
/
2
yr
in
orbit,
a
test
utilizing
the
alternate
calibration
lamp
of
the
CZCS
showed
degradation
of
less
than
one
digital
count
on
average
from
that
mea-
sured
before
launch.
This
degradation
included
the
optics
subsequent
to
the
insertion
of
the
lamp,
the
de-
tectors,
and
amplifiers.
Nevertheless,
the
images
1
February
1985
/
Vol.
24,
No.
3
/
APPLIED
OPTICS
407
1
1
1
0
\1
I
1
1
I
11
I
I
1,1
I
CALIBRATION
LAMP
FLD
ST
IE
OP
NEUTRAL
DENSI
TY
ILTER
F
SOLAR
CALIBRATION
MIRROR
MULTI
SPECTRAL
SCANNER
CALIBRATION
LAMP
FIELD
STOP
1.0
.9
.8
0
a
.6
CC
.
5
To
3
CC
.2
0
COASTAL
ZONE
COLOR
SCANNER
Fig.
1.
Calibration
schemes
MSS
and
CZCS.
produced,
especially
with
the
443-nm
channel
of
the
CZCS,
show
an
obvious
degradation.
This
indicates
that
loss
of
response
is
due
to
loss
of
reflectance
in
the
large
mirrors
prior
to
insertion
of
the
calibration
source
in
the
optical
path.
The
mirrors
are
fabricated
with
a
beryllium
base
over
which
there
is
a
Kanigen
coating
covered
by
an
evaporated
silver
film
that
is
protected
by
a
proprietary
coating
of
the
Samson
Co.
The
mirrors
most
exposed
to
the
environment
of
the
spacecraft
are
the
scan
mirror
and
the
primary
and
secondary
area
of
the
Cassegrain
telescope
system.
The
exact
mechanism
of
this
degradation
is
unknown
at
this
time
with
spec-
ulation
centered
around
possible
particulate
contami-
nation
due
to
outgassing
from
the
spacecraft
itself
or
micrometeorite
impact
roughening
the
surfaces
of
the
mirrors.
Until
one
of
these
instruments,
or
a
test
mirror
recovered
by
the
shuttle,
is
returned
to
earth
there
ap-
pears
to
be
no
way
of
determining
the
mechanism
of
the
degradation
of
the
mirror
surfaces.
III:
Aircraft
Measurements
A
number
of
methods
are
possible
for
estimating
degradation
of
sensors
in
space
including
calculations
based
on
ground-based
measurements
such
as
done
by
Gordon
et
a/.
1
One
fundamental
problem
exists
with
such
measurements
in
that
the
backscattering
charac-
teristics
of
the
aerosols
cannot
be
measured
directly
from
the
surface
except
possibly
with
the
use
of
a
mul-
tiband
laser
measurement.
Measurements
from
a
high-altitude
aircraft
which
can
fly
above
most
of
the
atmosphere,
and,
hence,
most
of
the
aerosols,
empiri-
cally
include
the
optical
charactereistics
of
the
aerosols.
It
is
important
to
carry
out
the
spectral
measurement
simultaneously
with
the
spacecraft
overflight
and
to
view
the
same
ground
surface
area
since
the
solar
ele-
vation
is
changing
rapidly
with
time
when
the
Nimbus-7
spacecraft
passes
overhead.
A
double
monochromator
2
was
chosen
for
this
par-
ticular
measurement
since
measurements
down
to
433
nm
with
a
single
monochromator
are
highly
susceptible
to
scattered
light
within
the
instrument
itself.
The
420 460
500
540
580
620
660
700
740 780
820
Wavelength
in
Nanometers
Fig.
2.
CZCS
spectral
bands.
1
/
8
-m
double
monochromator
(spectrometer)
used
a
cam
driven
sine
drive
to
scan
the
grating
every
5
sec.
The
average
spectral
resolution
of
the
instrument
is
7
nm.
The
spectrometer
is
calibrated
in
radiance
in
the
labo-
ratory
using
a
76.20-cm
diam
white
sphere
source
3
il-
luminated
by
twelve
internal
quartz-halogen
lamps.
The
sphere
output
is
traceable
to
the
NBS.
The
spec-
trometer
foreoptics
includes
a
quartz-wedge
depolarizer
and
a
lens
system
which
defines
the
surface
field
of
view
at
—2
X
2
km.
Some
distortion
in
the
shape
of
the
field
of
view
is
caused
when
the
look
angle
is
set
at
other
than
nadir.
The
field
of
view
of
the
CZCS
is
—800
X
800
m,
and
the
aircraft
spectrometer
field
of
view
provides
an
average
over
several
CZCS
fields
of
view.
The
CZCS
data
when
viewed
with
an
image
analyzer
were
averaged
over
an
area
of
approximately
the
same
size
as
the
air-
craft
spectrometer
footprint.
It
should
be
noted
that
the
spectrometer
was
chosen
deliberately
rather
than
trying
to
attempt
to
build
a
multichannel
radiometer
simulating
the
CZCS.
The
rationale
for
this
decision
was
that
it
would
be
difficult
to
duplicate
the
CZCS
spectral
bands
exactly,
whereas
they
were
well
measured
before
launch
as
shown
in
Fig.
2.
It
is
then
a
relatively
simple
matter
to
integrate
the
upwelled
radiance
measured
at
the
aircraft
over
the
measured
spectral
response
of
the
instrument.
Band
5,
from
700
to
800
nm,
is
relatively
insensitive
in
the
calculation
of
the
characteristics
of
water,
and
no
at-
tempt
was
made
to
recalibrate
it.
The
double
monochromator
chosen
for
this
mea-
surement
was
designed
for
NOAA
by
W.
G.
Fastie
of
the
Johns
Hopkins
University
and
is
shown
schematically
in
Fig.
3.
It
was
produced
for
NOAA
by
Research
Support
Instruments
Inc.
(RSI)
of
Cockeysville,
Md.
The
double
monochromator
was
flown
on
the
NASA
Lear
jet
aircraft
of
the
NASA
Lewis
Research
Center.
It
was
mounted
in
a
specially
constructed
downward
observation
pod
so
that
the
monochromator
was
ad-
justable,
and
the
optical
axis
could
be
pointed
off
to
the
side
of
the
aircraft
to
be
parallel
to
the
line
between
the
satellite
scan
and
the
target
area.
This
was
necessary
because
in
some
cases
the
subsatellite
track
of
the
spacecraft
was
so
far
off
the
Atlantic
coast
that
it
would
have
been
impractical
to
fly
the
airplane
out
directly
under
the
subsatellite
track
which
was
beyond
the
round
trip
flight
range
of
the
aircraft.
The
scan
of
the
408
APPLIED
OPTICS
/
Vol.
24,
No.
3
/
1
February
1985
D
ig
ita
l
Coun
ts
Primary
Mirrors
Battle
Intermediate
Slit
Diagonal
Mirror
Exit
Slit
Grating
Entrance
Slit
Fig.
3.
One-eighth-meter
zero
dispersion
monochromator.
Ebert
double
LEG
5
DATE
05/13
1983
CZCS
OVERPASS
TIME
1613.0
Z
AIRCRAFT
SCAN
TIME
1613.0
Z
AIRCRAFT
ALTITUDE
12.46
Km
UPWELL
RADIANCE
in
MW/CM
-
2
STER.uM
CZCS
CHANNEL
1-
5.47
2
3
8
2-
3.24
3-
2.67
4—1.21
U)
4
U
2
0
1
0
0
0
0
r
,
0
0
co
NANOMETERS
Fig.
4.
Aircraft
data.
260
240
Coastal
Zone
Color
Scanner
220
-
Channel
1
433-463
nm
Lear
Jet
Calibration,
May
1983
200
-
I80
Pre
Launch
May
4,
83
160
May
W
140
May
7,83
120
-
May13,83
100
-
80
60
40
-
20
0
1
0
2
3
4
5
6
7
8
9
10
N
mw/cm
°
star
fc
Fig.
5.
CZCS
channel
1
Lear
jet
calibration.
CZCS
of
40°
to
either
side
of
nadir
allowed
the
aircraft
insrument
to
be
adjusted
in
elevation
to
the
off-nadir
CZCS
view
vector
at
a
point
close
enough
to
shore
to
be
within
the
flight
range
of
the
aircraft.
The
Lear
jet
operated
at
a
pressure
altitude
of
—200
mbars
with
the
exact
pressure
altitude
determined
by
utilizing
both
the
aircraft's
pressure
altimeter
and
measurements
of
surface
pressure
at
the
National
Weather
Service
Sta-
tion
nearest
to
the
aircraft
track.
The
contribution
by
the
atmospheric
backscatter
above
the
aircraft
was
calculated
by
determining
the
Rayleigh
phase
function
from
the
known
solar
elevation
and
phase
angle
at
the
point
of
measurement.
Data
were
recorded
on
board
the
aircraft
during
flight
utilizing
a
digital
system
and
a
cassette
recording
device
which
could
be
immediately
entered
into
a
Hewlett-
Packard
9836
computer
on
return
to
the
laboratory.
The
computer
was
programmed
to
produce
a
calibrated
spectrum,
in
mW/cm
2
sr
Am
vs
wavelength
and
to
in-
tegrate
under
the
CZCS
spectral
response
functions
to
determine
the
radiance
within
the
four
bands
of
interest
of
the
CZCS.
A
sample
of
the
computer
printout
showing
all
the
various
information
that
is
produced
is
shown
in
Fig.
4.
Numerous
spectra
were
taken
on
each
flight
of
the
Lear
jet
before
and
after
coincident
mea-
surement
with
the
spacecraft.
These
spectra
will
not
be
shown
in
this
paper
for
the
sake
of
brevity;
however,
investigators
who
may
have
use
for
such
data
may
ac-
quire
copies
of
these
spectra
by
contacting
the
au-
thors.
Digital
magnetic
tapes
of
the
CZCS
imagery
for
the
days
in
question
were
produced
by
NASA
Goddard
Space
Flight
Center,
Nimbus
Operations
Control
Center,
immediately
after
the
flights.
They
were
ex-
amined
to
determine
the
digital
output
count
of
the
CZCS
while
viewing
the
same
area
as
viewed
by
the
aircraft
instrument.
Calculations
were
made
of
the
Rayleigh
backscattered
radiation
along
the
path
from
the
aircraft
to
the
spacecraft,
not
included
in
the
mea-
surement.
This
backscatter
was
added
to
the
measured
upwelled
radiance
as
seen
by
the
aircraft.
Figure
5
compares
the
digital
output
count
vs
radiance
for
the
433-453-nm
channel
of
the
CZCS
as
measured
prior
to
launch
and
as
measured
during
the
period
of
Lear
jet
flights
between
4
May
and
13
May.
This
figure
shows
that
the
433-453-nm
band
has
degraded
—25%
in
re-
sponsivity,
over
a
period
of
4
years
and
7
months,
be-
tween
prelaunch
calibration
and
this
calibration.
Figure
6
shows
the
results
for
the
channel
from
510
to
530
nm,
indicating
a
degradation
of
—9.5%
as
compared
to
prelaunch
measurements.
The
spectral
band
from
540
to
560
nm
was
found
to
be
degraded
by
—3%,
and
the
band
from
660
to
680
nm
of
the
CZCS
showed
no
measurable
degradation
within
the
accuracy
of
this
experiment.
Figure
7
summarizes
the
measured
degradation
after
4
years
and
7
months
in
orbit
and
shows
a
rather
star-
tling
spectral
dependence
of
the
degradation.
This
spectral
characteristic
of
degradation
has
implication
far
beyond
the
CZCS
program.
It
implies
that
any
in-
strument
with
a
spectral
band
extending
into
the
blue
1
February
1985
/
Vol.
24,
No.
3
/
APPLIED
OPTICS
409
DEGRADA
TIO
N
%
30
25
20
15
10
5
0
Coastal
Zone
Color
Scanner
Channel
2
610-630
nm
Lear
Jet
Calibation.
May
1983
Pre
Lowe
Lear
Jet
180
160
-
7
8
9
CZCS
DEGRADATION
4
YRS..
7
MOS.
IN
ORBIT
400
500
600
700
800
WAVELENGTH
NANOMETERS
Fig.
7.
CZCS
degradation.
1.10
1.00
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
o
'
400
500
600
WAVELENGTH,
X
(nm)
Fig.
8.
Thematic
Mapper
channel
1.
region,
such
as
the
Landsat
Thematic
Mapper
channel
1
illustrated
in
Fig.
8,
will
undergo
not
only
a
degrada-
tion
in
response,
but
a
change
in
the
shape
of
the
spec-
tral
response
function,
should
it
continue
to
function
for
a
period
of
time
equivalent
to
the
CZCS.
Present
schemes
and
those
being
planned
for
onboard
calibra-
tion
of
optical
sensors
such
as
the
CZCS,
the
Thematic
Mapper,
and
the
MSS
of
Landsat,
will
provide
only
total
responsivity
information
and
will
not
have
the
capability
of
determining
if
there
has
been
a
change
in
spectral
response.
If
spectral
response,
especially
in
the
blue
region
of
the
spectrum,
is
to
be
an
important
part
of
a
measurement,
it
would
appear
that
it
is
necessary
to
add
another
calibration
source,
namely,
a
source
that
can
measure
the
response
as
a
function
of
wavelength
to
determine
the
degree
of
alteration
with
time.
IV.
Future
Plans
The
instrument
will
be
flown
again
in
the
summer
of
1984
on
the
Lear
jet
and
then
be
transferred
to
the
NASA
ER-2
aircraft
which
can
reach
a
pressure
altitude
of
50
mbars.
This
will
reduce
the
amount
of
calculation
that
is
part
of
these
calibration
measurements
and
will
be
used
to
monitor
further
the
degradation
of
the
CZCS
optics
in
flight.
V.
Conclusion
Degradation
in
the
response
of
satellite
instruments,
sensing
in
the
visible
wavelength
region,
is
not
being
adequately
monitored
by
onboard
calibration
devices
since
these
devices
do
not
include
all
the
optical
ele-
ments
of
the
system.
The
results
from
the
CZCS
pro-
gram
and
from
the
aircraft
calibration
measurements
confirm
that
the
major
source
of
degradation
is
loss
of
reflectance
from
the
large
optics
that
are
exposed
to
the
spacecraft
environment,
namely,
the
scan
mirror
and
the
primary
and
secondary
of
the
telescope.
If
quan-
titative
measurements
are
to
be
made
with
such
in-
struments
in
the
future,
the
onboard
calibration
source
should
pass
a
beam
through
all
the
optical
elements
and
itself
must
be
protected
so
that
any
observed
degrada-
tion
is
indeed
in
the
instrument
and
not
in
the
optics
of
the
calibration
soruce.
Devices
such
as
white
diffuser
plates
would
have
to
be
enclosed
and
exposed
to
the
sunlight
only
for
very
brief
periods
of
time
to
avoid
solar
degradation.
The
French
HRV
instrument
scheduled
to
fly
on
the
SPOT
satellite
will
utilize
a
collimator
with
an
active
light
source.
The
lens
of
this
collimator
should
be
protected
so
that
it
does
not
undergo
degra-
dation
due
to
whatever
mechanism
is
degrading
the
CZCS,
and,
therefore,
if
any
degradation
is
observed,
it
can
truly
be
attributed
to
the
instrument
performance
and
not
to
the
performance
of
the
calibration
source
itself.
Such
additions
will
certainly
be
expensive,
but
if
quantitative
measurements
are
to
be
made
from
space
they
appear
to
be
essential.
The
authors
wish
to
thank
our
colleagues
in
the
Sat-
ellite
Experiment
Laboratory:
Lee
Johnson,
Francesco
Mignardi,
Kenneth
Hayes,
and
Robert
Levin,
for
their
contributions
to
the
instrument
electronic
design,
as-
sembly,
calibration,
and
data
gathering
in
the
field;
and
John
Bray
and
Robert
Koyanagi
for
their
contribution
to
instrument
mechanical
design,
fabrication,
and
as-
sembly.
References
1.
H.
R.
Gordon,
J.
W.
Brown,
0.
B.
Brown,
R.
H.
Evans,
and
D.
K.
Clark,
"Nimbus
7
CZCS:
Reduction
of
its
Radiometric
Sensitivity
with
Time,"
Appl.
Opt.
22,
3929
(1983).
2.
G.
R.
Smith
et
al.,
"The
NESDIS-SEL
Lear
Aircraft
Instruments
and
Data
Recording
System,"
NOAA
Tech.
Memo.,
NESDIS
9
(June
1984).
3.
W.
A.
Hovis
and
J.
S.
Knoll,
"Characteristics
of
an
Internally
Il-
luminated
Calibration
Sphere,"
Appl.
Opt.
22,
4004
(1983).
0
140
8
120
-
O
(")
too
o,
so
a
40
20
-
200
0
0
2
3
4
5
6
N
mw/cm`
ster
Fig.
6.
CZCS
channel
2
Lear
jet
calibration.
RELATIVE
S
PECTRAL
RESPONSE
410
APPLIED
OPTICS
/
Vol.
24,
No.
3
/
1
February
1985