Application of Wavelet Transform to Meteosat-Derived Cold Cloud Index Data over South America

Abstract

Cold cloud index (CCI) data derived from Meteosat infrared imagery are used to detect periodicities in convective activity in South America. The generally used Fourier transform (FT) cannot provide time-localized information but gives information on the average periodicity of oscillations over the entire time domain. As many events in the atmosphere are intermittent, wavelet transform (WT) is used to identify periodic events in CCI data. First, the Morlet WT is applied to different combinations of time series data of known periodicities to demonstrate the advantage of WT over FT. Later it is applied to CCI data over four 9° square areas between the latitudes 4.5°N and 31.5°S, and longitudes 54°?45°W. Near the equator periodic convective activities are observed to be more prominent in the boreal summer than in the austral summer. Between the latitudes 4.5° and 22.5°S, 1-, 2?3-, approximately 5-, and 8?10-day oscillations are seen in the austral summer and seldom is any convective activity seen in the winter. In January semidiurnal variation of cloudiness is also observed for a few days. Farther south in the extratropics, approximately 10- and approximately 20-day periodic events, which refer to the baroclinic waves, are seen more prominently in the austral autumn and winter, and 1- and approximately 5-day oscillations are seen in the summer, perhaps due to convective cloudiness.