Release 4.0
August 2000 2
CERES GEOLOCATE AND CALIBRATE EARTH RADIANCES
LEVEL 1 INSTRUMENT VALIDATION PLAN
1.1 INTRODUCTION
This plan is designed to (1) trace the absolute calibrations of the Clouds and the Earths Radiant
Energy System (CERES) spacecraft sensors radiance and geometric measurements from ground to
ight, (2) dene short-term and long-term shifts or drifts in the measurements caused by sensor
response variations, and (3) determine CERES measurement consistency among the same types of
CERES sensors on the same and different spacecraft platforms and with other similar spacecraft
radiance/ux measuring sensors. The plan and on-orbit results are described in detail by Lee et al.
(1996a, 1998) and Priestley et al. (2000)
1.1.1 Measurement and Science Objectives
The CERES sensors are scanning thermistor bolometers which measure Earth-reected and Earth-
emitted ltered radiances in the broadband shortwave (0.3
�
m - 5.0
�
m), broadband total-wave (0.3
�
m - >100
�
m), and narrow-band water vapor window (8
�
m-12
�
m) spectral regions. Broadband
longwave radiances (5
�
m - >100
�
m) are derived from the differences between the total-wave and
shortwave radiances. These radiance measurements, along with imager measurements, dene the
impacts of clouds and of certain cloud properties upon the Earths radiation budget and climate
(Wielicki and Barkstrom 1991, Wielicki et al. 1996, Wielicki et al. 1998).
1.1.2 Missions
On November 27, 1997, the rst set of the CERES bolometers was placed into orbit aboard the
Tropical Rainfall Measuring Mission (TRMM) Spacecraft platform. The TRMM spacecraft was
launched into a low-inclination 35
o
, 350-km altitude orbit using a National Space Development
Agency (NASDA) H-II expendable launch vehicle from the Tanegashima Space Center, Japan.
During December 1999, the second and third sets of bolometers are scheduled for launch on the
EOS Terra spacecraft platform. December 2000, the fourth and fifth sets of bolometers will be
launched on the EOS Aqu spacecraft. The EOS spacecraft platforms will be launched into Sun-
synchronous polar, 705-km orbits using NASA Atlas IIC launch vehicles.
1.1.3 Science Data Product
The level 1, CERES instrument data product is geolocated ltered broadband Earth radiances, in
Wm
-2
sr
-1
, at the top-of-the-atmosphere (~30 km). For average target scenes less than 100
Wm
-2
sr
-1
, the broadband shortwave and longwave instrument measurement accuracy
requirements are 0.8 Wm
-2
sr
-1
and 0.6 Wm
-2
sr
-1
, respectively, as indicated in Table 1. Earth
Radiation Budget Experiment (ERBE) Spacecraft (Barkstrom 1984, Barkstrom and Smith 1986)
in-ight calibration systems, earlier versions of the CERES systems, were used to verify shortwave
and longwave radiance measurement precisions at the + 0.3% measurement precision levels (Lee
et al. 1993). During the ERBE missions, a coastline validation technique (Hoffmann et al. 1987)
was used to verify geolocation calculations at the + 6 km uncertainty level. The coastline technique
was improved and used to validate the TRMM CERES geolocation calculations at uncertainty CERES VAL Subsystem 1.0
- Geolocate and Calibrate Release 4.0
August 2000 3
levels approaching + 0.5 km near the nadir (Currey et al. 1998).
1.2 VALIDATION CRITERION
1.2.1 Overall approach
The CERES level 1 radiometric data product validation plan includes assessments of (1) the degree
to which the absolute longwave and shortwave radiometric scales are transferred from the ground
to space by the CERES sensors and in-ight calibration sources; (2) the in-ight long-term
stabilities of the CERES sensor responses; (3), in the case of sensor response drifts or shifts,
revisions to the sensor in-ight count conversion coefcients (gains and offsets); and (4)
validations of the geolocation calculations from analyses of the calculated locations of geometric
registration sites and their corresponding measured ltered radiances. Ground-calibrated
instrument count conversion coefcients are used to convert the sensor output signals into
radiances. The Ground coefcients were derived in the TRW Radiometric Calibration Facility (Lee
et al. 1996b, Lee et al. 1997, Lee et al. 1998) and tied radiometrically to the International
Temperature Scale of 1990 (ITS-90). The ground-derived coefcients are revised only if the sensor
response drifts or shifts more than 0.5 Wm
-2
sr
-1
in the longwave spectral region or more than 0.8
Wm
-2
sr
-1
in the shortwave region. If the sensor response drifts or shifts above these levels, the ight
coefcients will be revised off-line in science computing facilities (SCF) and applied only after the
approval of the CERES Science Team. The TRMM/CERES ground-derived gains and offsets were
used to process the December 1997 thru June 2000 ight measurements. Analyses of ground and
ight calibrations indicated that the CERES bolometers responses (gains) did not change between
the ground and the on-orbit calibrations and that they were stable at uncertainty levels better than
0.2 Wm
-2
sr
-1
. Therefore, the ground-derived gains were not revised. Radiometric measurements of
deep space were used to dene the nal sensor zero-radiance offsets. Therefore, the archived
January thru June 2000 radiances were processed using the ground gains and on-orbit offsets.
The elements of the overall validation plan for the level 1 geolocated ltered radiances include (a)
radiometric analyses (Lee et al. 1992, Lee et al. 1993, Lee et al. 1998) of ground and on-orbit/in-
ight calibration measurements; (b) geometric registration sites/coastline detection analyses
(Hoffmann et al. 1987, Currey et al. 1998) to estimate upper limits in geolocation errors; and (c)
single and multi-spacecraft intercomparisons of Earth radiance measurements from the same type
of broadband sensors (Avis et al. 1994, Green and Avis 1996) in the cross-track and rotating
azimuth plane (RAP) Earth radiance scanning modes.
The CERES instrument radiometric performance is validated/checked using multi-sensor
comparisons of Earth radiance measurements from the same spacecraft (Green and Avis 1996) and
from other past and current spacecraft sensors [Earth Radiation Budget Satellite (ERBS)
nonscanners, ERBS scanners, Scanner for Radiation Budget (ScaRaB) scanners, etc.). Using in-
ight calibration sources, the CERES sensors were checked for instrumental drifts or shifts. These
activities, may not guarantee that the CERES data are consistent with the historical ERBS data
(Barkstrom 1984, Barkstrom and Smith 1986), or that the in-ight calibration systems are stable.
For example, if on the ground, the total channel is tested against the on-board blackbodies. In ight,
if the same procedure is performed, and if the two readings are different, it will not be clear whether
the difference is caused by the radiometer or by the blackbody. The same is true for any detected CERES VAL Subsystem 1.0
- Geolocate and Calibrate Release 4.0
August 2000 4
drift. For these reasons, the CERES radiances are validated against Earth validation targets (Staylor
1986, 1993, Priestley 2000). The statistics of the radiance for these targets were established with 5
years of validated ERBS scanner data (Green and Avis 1996). If statistically signicant differences
between the CERES measurements and the Earth validation targets are found, and if these
differences are consistent with the in-ight calibration systems and results are within the
uncertainties of the instrument calibration and validation, then the radiances will be revised with
the approval of the CERES Science Team. If the differences are outside the instrument
uncertainties, then all analyses will be re-examined and additional sources of verication will be
sought. It is probable that superior CERES calibration systems and consistency between the
CERES ground and in-ight calibration will make CERES the standard and assess possible biases
in the ERBS data.
1.2.2 Sampling Requirements and Trade-offs
For a minimum of the rst 30 days in orbit, the sensor contamination doors are closed. During this
period, the sensors are calibrated daily using measurements from the internal calibration module
(ICM) which are referenced to the radiances from the contamination doors emitted and reected
radiances. After the doors are opened, measurements of the ICM sources are performed daily
during the rst week, every other day during the second week, once a week during the third and
fourth weeks, and thereafter every 14 days (Lee et al. 1998).
Validation measurements of the Earth radiances are used to verify sensor response changes, indi-
cated by in-ight calibrations. The CERES science team will conduct detailed analyses of at least
the rst 6 months of in-ight calibrations and validations before the sensor gains or offsets are
revised. Validation of the CERES radiances against Earth validation targets require a minimum of
two months of data. Additional months will strengthen the statistical hypothesis tests and increase
the probability of detecting errors if they exist. After the 90th day and during the remainder of the
mission, the in-ight calibrati