Calibrating a System of Satellite Instruments - Satellite-based Applications on Climate Change - page 20

Fig. 2.4 Dome C time series comparisons between MODIS, MetOp/AVHRR, and SeaWiFS

JPSS (Joint Polar Satellite Systems), has onboard solar diffusers with improved
calibration, but it is still desirable to verify their consistency relative to other satellites.
Calibration traceability issues often cause discrepancies between different
instruments. For example, since MODIS calibration uses a solar diffuser, while
AVHRR calibration is traceable to the Libyan Desert, a discrepancy is observed
when comparing AVHRR and MODIS solar bands. Multiple independent studies
suggest that a ~10% difference in measured reflectance values exists betweenMODIS
and AVHRR 0.63 um band, and this difference cannot be accounted for by the
differences in their spectral response functions (Cao et al. 2008a ). This discrepancy
is also observed at vicarious sites such as the Dome C site in Antarctica (Fig. 2.4 ), and
it was found to be due to inaccurate characterization of the desert target in the early
1980s. Clearly on-orbit calibration traceability is causing uncertainties which will
undoubtedly affect time series analysis using both AVHRR and MODIS. Similar
issues exist when comparing data from international satellites for which not all
calibration information is readily available to the international research community.
For the infrared and microwave instruments calibrated with blackbody and space
view, prelaunch thermal vacuum tests are typically performed in the factory to
ensure that the instrument meets specifications, but more work needs to be done to
improve the calibration traceability. For example, in the case of AVHRR and HIRS
on NOAA satellites, the “NIST traceability” simply meant that the blackbody PRTs
(platinum resistance thermometers) are traceable to National Institute of Standards
and Technology (NIST) standards, while the radiometric traceability has not been
established. The postlaunch performance of the blackbodies for the legacy
radiometers is generally good enough for traditional weather applications with an
estimated accuracy of 0.5 K but may not be sufficiently accurate for climate

applications and today's numerical weather predictions. On-orbit uncertainties are
mainly caused by a combination of stray light in the scan-cavity and blackbody
emissivity, especially in terminator orbits.
Nonlinearity is another major source of uncertainty which can introduce scene-
dependent biases. For the silicon detectors in the visible/near-infrared, the relationship
between incoming radiation and output counts is generally linear, the MCT (mercury
cadmium telluride) detectors in the infrared in photoconductive mode, as well as the
square law detectors in the microwave, are nonlinear between incoming radiation and
output counts. The nonlinearity of these systems is tested in a thermal vacuum