A Joint Temperature , Humidity , Ozone , and SST Retrieval Processing System for IASI Sensor Data : Properties and Retrieval Performance Analysis

We show the improved performance of a joint retrieval algorithm of temperature, humidity, ozone, and SST (more precisely, the latter is the surface skin temp erature of the ocean) compared to more specific retrieval setups. The joint algorithm was de veloped based on optimal estimation methodology and carefully tested under quasi-realistic c onditions (using high resolution ECMWF analysis fields). The algorithm contains in a first step aneffective and fast channel selection method based on information content theory, w hich leads to a reduction of the total number of IASI channels (>8400) to about 3.5% only (∼300), which are subsequently used in the retrieval processing. We show that this reduct ion is possible without retrieval performance decrease compared to using many more ( ∼2000) channels. Additionally, it is shown that using standard climatology fields in the channel select ion process does also not decrease performance while significantly increasing computatio nal efficiency. Finally, the application and real-data-test of the algorithm with AIRS (Advanced Infrared Sounder) data, a next step planned, is addressed.