Near-surface retrieval of air temperature and specific humidity using multisensor 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 Advanced Microwave Sounding Unit-A (AMSU-A), Special Sensor Microwave Imager (SSM/I), and Special Sensor Microwave Temperature Sounder (SSM/T-2). Several retrieval methods based on linear regression are presented that include both single-sensor and multisensor approaches. The most accurate linear regression algorithms combined AMSU-A and SSM/I observations. The Q(a) and T-a retrieval methods had RMS differences of 0.83 g/kg and 1.53 degrees C, respectively, relative to direct surface observations. These differences were found to be significantly lower than those computed from previously published algorithms applied to daily values. An independent validation data set showed that the current Q(a) retrieval method led to a reduction in RMS error of 0.5 g/kg and reduced bias at the lowest and highest Q(a) values. Likewise, the current T-a retrieval method had a reduction in RMS errors of 1.0 degrees C from previously published methods. Improvements in both retrievals can be attributed to inclusion of the AMSU-A 52.8 GHz channel, which helps distinguish variations near the surface from those at higher levels of the atmosphere. Selection of a retrieval approach represents a trade-off between accuracy, coverage, and temporal extent. Global coverage every 6 hours is approximately 80% using two NOAA satellites; however, such coverage would rely on use of a single-sensor AMSU-A approach with somewhat degraded accuracy. A multisensor approach using AMSU-A and SSM/I on the current NOAA and DMSP satellites will at best provide global coverage over 1 day.