Calibrating a System of Satellite Instruments - Satellite-based Applications on Climate Change - page 14

2.1 Introduction

Climate change detection and numerical weather prediction (NWP) critically depend
on accurate, reliable, and consistent radiance data from multiple sensors onboard a
constellation of satellites. The calibration consistency across international satellites is
further challenged in the implementation of GEOSS (Global Earth Observation
System of Systems). As a result, establishing on-orbit calibration links among
operational satellite radiometers is receiving increasing attention. According to the
CEOS (Committee on Earth Observation Satellites), which consists of the majority of
the space agencies worldwide, its member agencies are operating or planning about
240 satellites with Earth observation missions over the next 15 years, carrying over
385 different instruments ( http://www.eohandbook.com ). Unfortunately, not all sat-
ellite observations can be put into productive use because the data quality varies from
one instrument to another significantly and the measurements may not be consistent
even for satellites within the same agency due to different calibration methodologies
and techniques used. This chapter presents the fundamental issues in satellite instru-
ment calibration for climate applications and provides an overview of the current
status, challenges, and latest developments in the calibration of operational satellite
radiometers with specific examples.
For climate change detection, calibration accuracy and consistency are essen-
tial since the signal of climate change can be very small, such as 0.1 K per decade
in sea surface temperature. Given a satellite radiometer's typical design life of
about 5 years, the detection of decadal climate change relies on observations from
a series of satellites. It is well known that despite the best effort in prelaunch and
postlaunch calibration, the same series of radiometers on different satellites, such
as the AVHRR (Advanced Very High Resolution Radiometer), HIRS (High
Resolution Infrared Radiation Sounder), and AMSU (Advanced Microwave
Sounding Unit)/MSU(Microwave Sounding Unit) on NOAA satellites, do not
necessarily produce consistent measurements. This leads to the inter-satellite
biases which have become major concerns in constructing time series for climate
change detection. Current satellite systems are significantly challenged to meet
these application requirements. Sensors and onboard calibration sources may
degrade in orbit; measurements may not be traceable to international standards;
long-term data sets must be stitched together from a series of overlapping satellite
observations; and orbital drift can introduce artifacts into long-term time series.
As demonstrated in the tropospheric temperature trend study using MSU channel
2 observations, the inter-satellite biases can become so critical that depending on
how the biases are treated may lead to different conclusions about the tropo-
spheric warming (e.g., compare results from Zou et al. 2006 ; Vinnikov and Grody
2003 ; Christy et al. 1998 ).
Clearly, satellite detection of the global climate change signals critically
depends on consistent and accurately calibrated Level 1b data or FCDRs (funda-
mental climate data records). Without dependable FCDRs, the derived TCDRs
(thematic climate data records) may produce false trends with questionable scien-
tific value. This has been increasingly recognized by more and more users of