The Validation of the Operational AMSU Snowfall Identification Product

The Advanced Microwave Sounding Unit (AMSU) which is on board the NOAA Polar Orbiting Environmental Satellites (POES) and the EUMETSAT MetOp-A satellite has been extremely useful in precipitation retrievals. In particular, the combination of high frequency/water vapor channels on AMSU-B/MHS and the oxygen channels on AMSU-A, have been used to retrieve areas of falling snow and distinguish them from falling rain. This retrieval is part of the operational AMSU product suite generated at NOAA/NESDIS. There are limitations to the application of this algorithm to snowfall retrievals. In particular, it has been demonstrated that the algorithm will only work well when the precipitation system is rather robust in nature (i.e., large vertical and horizontal extent) and that there is sufficient atmospheric moisture content to greatly mask out the underlying surface. However, to date, only limited validation has been performed, mostly due to the difficulty in finding sufficient ground data that matches the attributes of the satellite retrievals. In this study, we validate the areal coverage and frequency of occurrence of the AMSU snowfall detection for the winter seasons for the period 2000 – 2008 against daily surface observations over North America to better understand the accuracy of the satellite retrieval. The statistics will be stratified according to different climate regimes (e.g., dry and moist), surface temperatures and precipitation amount. In addition, detailed analysis of some specific snowstorm “events” will be presented to better describe the synoptic conditions of instances where the retrieval works well and also fails.