DETECTING DISCONTINUITIES IN TIME-SERIES OF UPPER-AIR DATA - DEVELOPMENT AND DEMONSTRATION OF AN ADAPTIVE FILTER TECHNIQUE

Recognizing the need for a long-term database to address the problem o f global climate change, the National Climatic Data Center has embarke d on a project called the Comprehensive Aerological Reference Data Set to create an upper-air database consisting of radiosondes, pibals, su rface reports, and station histories for the Northern and Southern Hem ispheres. Unfortunately, these data contain systematic errors caused b y changes in instruments, data acquisition procedures, etc. It is esse ntial that systematic errors be identified and/or removed before these data can be used confidently in the context of greenhouse-gas-induced climate modification. The purpose of this paper is to illustrate the use of an adaptive moving average filter in detecting systematic biase s and to compare its performance with the Schwarz criterion, a paramet ric method. The advantage of the adaptive filter over traditional para metric methods is that it is less affected by seasonal patterns and tr ends. The filter has been applied to upper air relative humidity and t emperature data. The accuracy of locating the time at which a bias is introduced ranges from about 600 days for Changes of 0.1 standard devi ations to about 20 days for changes of 0.5 standard deviations.