Pre-operational retrieval of radio occultation based climatologies

CHAMPCLIM is a joint project of WegCenter/UniGraz and GFZ Potsdam. The overall aim of the project is to exploit the CHAMP (CHAllenging Minisatellite Payload for geoscientific research) radio occultation (RO) data in the best possible manner for climate monitoring. This paper focuses on describing the pre-operational status and technical aspects of the CHAMPCLIM processing system at WegCenter/UniGraz. For creating RO based climatologies we ingest, on the one hand, the complete CHAMP RO dataset provided by GFZ at excess phase level (GFZ level 2, similar to 180 profiles/day), which is processed to obtain atmospheric profiles of refractivity, geopotential height, and temperature (in future also humidity). On the other hand, we use operational atmospheric analysis fields from the European Centre for Medium-Range Weather Forecasts (ECMWF), at T42L60 resolution, as reference for quality control and evaluation. For delivering climatologies operationally, which will be prepared at monthly, seasonal, and annual time scales, our aim is to provide them with a delay of at most two weeks after the last measurement (e.g., JJA 2003 seasonal climatology available by September 14, 2003, latest). The climatologies are set up in overlapping equal-area and non-overlapping almost equal-area grids. In order to monitor the error characteristics of the climatologies, various types of error statistics (vs. ECMWF analyses) are performed. The main emphasis of this paper lies on processing the complete 20022004 data - starting from March 2002 when the CHAMP data stream became stable and quasi continuous - and on creation of climatologies including error estimates. The spatial set up of the climatologies, exemplary seasonal climatologies (as far as processed) as well as preliminary climatological error estimates are presented.