Recent studies have shown that the Arctic climate has changed significantly over the past 20 years, particularly in terms of ice extent, cloud cover, and surface temperature. Changes in these properties are a function of large-scale circulation patterns that affect surface-atmosphere interactions and feedback mechanisms. Perhaps the best tools for studying these interactions are the National Centers for Environmental Prediction (NCEP)/National Center for Atmospheric Research (NCAR) and the European Center for Medium Range Weather Forecasts (ECMWF) reanalysis products. Unfortunately, both of these products have been shown to have relatively large errors in the wind field over the Arctic where there is little or no radiosonde data available for assimilation.

Can satellites be used to fill the spatial gaps? While geostationary satellites provide useful wind information at lower latitudes, they are of little use in the polar regions due to poor spatial resolution and viewing geometry. Polar orbiting satellites provide excellent spatial resolution and have recently been shown to be useful for estimating high-latitude tropospheric winds. In fact, at least nine numerical weather prediction centers worldwide have demonstrated that satellite-derived polar winds have a positive impact on global weather forecasts.

The impact on climate reanalyses should be similar. Therefore, a polar wind data set spanning 23 years has been generated using Advanced Very High Resolution (AVHRR) Global Area Coverage (GAC) data from NOAA satellites. Wind speed, direction, and height are estimated for the Arctic and Antarctic, poleward of approximately 65 degrees latitude, by tracking the movement of clouds. Comparisons to ECMWF reanalysis winds have shown that the largest differences between the model and satellite-derived winds occur in areas with strong divergence gradients. It is recommended that the historical AVHRR polar winds be assimilated in future versions of the reanalysis products.