Using time continuity for sounding retrieval from GOES data

A new retrieval algorithm, which is based on the concept of time continuity, is developed and applied to Geostationary Operational Environmental Satellite-12 (GOES-12) sounder radiance measurements. One of the most important advantages of geostationary satellites is high temporal resolution (1 hour for GOES-12 sounder). Obviously, the atmospheric status of one hour ago also provides information for current hour retrieval. For optimal sounding retrieval using GOES data, time continuity needs to be accounted for so that more known parameters/information are introduced in the inversion procedure. Results show better consistence with microwave radiometer measured total precipitation water (TPW). Using the match-up files from CIMSS/University of Wisconsin-Madison, a new regression method is used to provide the first guess, which was usually the forecast data, for physical retrieval. By combining the forecast data and GOES observations, the regression provides better first guess than forecast comparing with radiosonde observations (RAOB). Also, a new filtering method, Fourier transfer filtering in frequency domain, is used to filter noise in Channel 1, 2, 3, and 15. Results show the filtering has positive impact on retrieval.