The accuracy of satellite polar cloud cover change from satellite retrievals by use of orbital lidar comparison

Arctic and Antarctic Peninsula amplification of near surface warming is not well understood. Corresponding changes in polar cloud cover are a possible important factor, but the accuracy of polar cloud cover observations is know to be a problem for existing passive satellite retrievals. The Geoscience Laser Altimeter System (GLAS) measurements provide satellite lidar data of clouds and aerosol starting in 2003, and in 2006 Caliop data became available. From these orbital lidar data the presence and height of all clouds and significant haze are detected with much higher sensitivity, and unambiguous accuracy. In the case of transmissive layers, the spatial and optical thickness is retrieved. Comparisons have been made between satellite lidar data products and TOVS, ISCCP, MODIS and associated CERES cloud products for the Arctic and Antarctic. Disagreements of the passive results to lidar results in monthly averaged parameters are, in important measures, very large. Differences between the zonal average GLAS and ISCCP and MODIS cloud fractions are as much as 40% over Antarctica and in the Arctic. In most regions the TOVS cloud fraction is much closer to the lidar results. From the lidar data there are good monthly statistics on cloud height distribution and optical thickness distribution. For the cloud height there is little correspondence between the cloud top height reported by lidar and any of the passive retrievals, the lidar height being an accurate measure of the true cloud top and the other an effective radiated average. Reports of increasing Arctic cloudiness in the current decade have been found from MODIS and other satellite retrievals. The comparisons to satellite lidar results are applied to study to the accuracy and meaning to the retrieved cloud cover variations.