Arctic Surface Temperature: A Comparison among Satellite Retrievals and Conventional Observations

Abstract

Surface temperature is an essential variable for studying the heat budget of the Arctic. Many efforts have been made to get accurate observations over this region. Satellites are the only practical source of basin-wide information over the Arctic. In this study, we compare several surface temperature data sets, including recently available satellite retrievals from the NASA/NOAA TOVS (TIROS Operational Vertical Sounder) Polar Pathfinder Data Set (Path-P); the IABP/POLES (International Arctic Buoy Program/Polar Exchange at the Sea Surface) gridded temperatures interpolated from Russian drifting NP (North Pole) station data, Arctic Rawinsonde Data from Arctic land stations north of 65°N (HARA), and ship observations in the GIN Sea from the Comprehensive Ocean-Atmosphere Data Set (COADS). The comparisons show that there are apparent discrepancies between TOVS Path-P and IABP/POLES, especially in summer, when the difference is larger than 4 to 6 K in some areas. Some possible reasons for those discrepancies are given. Besides the limitations inherent in each data set, which introduce errors into the comparison, the effect of clouds in the satellite retrieval process and the large view angle over the central Arctic Ocean account for some of the observed differences. Additional validation is needed before either satellite retrievals or the IABP/POLES gridded dataset can be used to study the whole Arctic region.