A COMPREHENSIVE AEROLOGICAL REFERENCE DATA SET (CARDS) - ROUGH AND SYSTEMATIC-ERRORS

The possibility of anthropogenic climate change and the possible probl ems associated with it are of great interest. However, one cannot stud y climate change without climate data. The Comprehensive Aerological R eference Data Set (CARDS) project will produce high-quality, daily upp er-air data for the research community and for policy makers. CARDS in tends to produce a dataset consisting of radiosonde and pibal data tha t is easy to use, as complete as possible, and as free of errors as po ssible. An attempt will be made to identify and correct biases in uppe r-air data whenever possible. This paper presents the progress made to date in achieving this goal. An advanced quality control procedure ha s been tested and implemented. It is capable of detecting and often co rrecting errors in geopotential height, temperature, humidity, and win d. This unique quality control method uses simultaneous vertical and h orizontal checks of several meteorological variables. It can detect er rors that other methods cannot. Research is being supported in the sta tistical detection of sudden changes in time series data. The resultin g statistical technique has detected a known humidity bias in the U.S. data. The methods should detect unknown changes in instrumentation, s tation location, and data-reduction techniques. Software has been deve loped that corrects radiosonde temperatures, using a physical model of the temperature sensor and its changing environment. An algorithm for determining cloud cover for this physical model has been developed. A numerical check for station elevation based on the hydrostatic equati ons has been developed, which has identified documented and undocument ed station moves. Considerable progress has been made toward the devel opment of algorithms to eliminate a known bias in the U.S. humidity da ta.