Validation of Noaa-15 AMSU-A Rain Rate Algorithms

The recently launched NOAA-15 satellite (May 13, 1998) has on board a new series of microwave radiometers called AMSU (Advance Microwave Sounding Unit). These long-awaited instruments provide a new opportunity to produce precipitation estimates from NOAA polar satellites. Despite several critical sensor differences, e.g., scan geometry, polarization, etc., previous knowledge and experiences from SSM/I were vital in helping to develop the first generation of AMSU-A rain retrieval algorithms. As more data are collected, a detailed validation effort for different products is being conducted to test how well the pre-launch retrieval algorithms perform. Also this effort is being used to identify problems and search for further improvements before the product system becomes operational. Herein presented are some preliminary results from AMSU-A rain rate algorithm validation.

The algorithms being validated utilize the AMSU-A sensor and are a 23.8/31.4 GHz emission algorithm over ocean and a 23.8/89.0 GHz scattering algorithm over land. Validation is limited by the availability of ground truth data. Over ocean, surface data are severely limited, and validation was attempted only in the Pacific Atoll region. The Comprehensive Pacific Rainfall Data Base (CPRDB), which is generated from rain gauge measurements in the Pacific Ocean by Morrissey (1993), is used. Over land, a good analysis of radar and gauge data set, namely NCEP/EMC Stage IV hourly rainfall analysis, is used. The retrieved AMSU-A rainfall data are projected into NCEP/EMC precipitation analysis domain, and a grid-to-grid match-up between NCEP/EMC hourly rainfall data and AMSU-A rainfall estimates is generated. Then the AMSU-A rainfall estimates are compared with the closest observation of the NCEP/EMC hourly rainfall. In attempting to validate the algorithms, various products were generated including images, validation statistics, scatter plots, etc. The preliminary result shows that AMSU-A rainfall estimates agree well with NCEP/EMC hourly rainfall data and Pacific Ocean Atoll data. More detailed analysis is ongoing, and results will be presented in the conference.