Remote Observations of Ice Sheet Surface Temperature: Toward Multi-Proxy Reconstruction of Antarctic Climate Variability

This project will develop spatially extended and statistically reliable
estimates of Antarctic surface temperature variations over the past several
centuries, using a multi-proxy calibration/verification approach that combines
the climate signal in ice core, satellite remote sensing, and weather station
data. Antarctica has been problematic from the point of view of large-scale
paleoclimate reconstruction because of the paucity of long-term instrumental
data, and the relatively low resolution of most ice cores.

Several new developments, particularly the network of shallow (~100 meter) ice
cores from the ongoing International Trans-Antarctic Scientific Expedition
(ITASE) project will yield broad spatial coverage of annually resolved ice core
physical properties, chemistry, and stable isotope data over more than a
hundred years. Second, there are now over twenty years of microwave and
infrared brightness temperatures available from satellites covering virtually
all of Antarctica with seasonally resolved information that has been
demonstrated to record the ice surface/near surface temperature with very
reasonable precision and accuracy. Finally, higher quality microwave emission
data from Advanced Microwave Scanning Radiometers (AMSR) with much finer
spatial resolution and radiometric fidelity than those from previous sources
will offer an improved view of longer term mean temperatures in Antarctica.

The 40-year instrumental record and the shorter but spatially more
comprehensive Automatic Weather Station network will be combined with
seasonally-resolved 37-gigahertz satellite-based ice surface temperature
estimates to reconstruct Antarctic-wide temperature patterns during the past
forty years. The sparser Antarctic instrumental surface temperature data
available back nearly to the beginning of the century will be added for
longer-term, though quite spatially-restricted, cross-validation of these
reconstructions. This cross-validation procedure has been used successfully
with roughly century-long instrumental records at locations primarily in the
Northern Hemisphere. The longer time scale will be approached through a
cross-validation of the proxy-based pre-20th century surface temperature
reconstructions using information on thermal emission from deeper in the firn
that is contained in low-frequency passive microwave satellite measurements.
The low-frequency estimates, supplemented by borehole thermometry, will provide
important independent verification of the long-term averages of the annual
surface.