Near-surface retrieval of air temperature and specific humidity using multi-sensor microwave satellite observations

A new method for instantaneous retrievals of near-surface specific humidity (Q_a) and air temperature (T_a) over the oceans was developed by combining satellite microwave observations from AMSU-A, SSM/I and SSM/T2. A suite of retrieval methods are presented that include both single and multi-sensor approaches. The most accurate Q_a retrieval method had an rms difference of 0.90 g/kg when using a combination of AMSU-A and SSM/I channels, while the most accurate T_a retrieval had a regression error of 1.53^oC when using all three instruments. These differences were found to be significantly lower than those computed from previously published algorithms applied to daily values. A direct comparison with previous methods using an independent validation data set indicates the multi-sensor approach is more accurate. It was found that using multiple satellites to provide global coverage of Q_a and T_a observations does have limitations. Global coverage every six hours can nearly be achieved using two NOAA satellites; however, such coverage would rely on many of the retrievals using a single AMSU-A instrument because the SSM/I satellites do not provide close time matches with the NOAA afternoon (2am/pm) satellite. As a result, a multi-sensor approach using AMSU-A, SSM/I and SSM/T2 on the NOAA and DMSP satellites at will at best provide daily mean global coverage.