The Problem of Extracting Precipitation Information in the Tropics from the UWM/COADS Data

Abstract

This study employs principal component analysis to correct tropical precipitation estimates in the University of Wisconsin?Milwaukee (UWM)/Comprehensive Ocean?Atmosphere Dataset (COADS) data. The idea was to use a matrix made up of the other variables in the set, to reduce the dimensionality of the matrix by considering a small number of principal components, and then to regress precipitation to these principal components. The results indicate that, although some information on precipitation could be restored by this method, overall the resulting precipitation estimates are not reliable. This result is traced to the intrinsic complexity of precipitation and possibly to a newly discovered bias in the UWM/COADS data.