The Effects of Amazon Deforestation on Rainfall

Abstract

This study begins with the hypothesis that heavily deforested regions will experience increased surface heating, leading to local circulations that will ultimately enhance the rainfall, or at least, change the pattern of diurnal evolution of rainfall. This would be an important finding because several modeling studies have concluded that widespread deforestation would lead to decreased rainfall. Towards that end rain estimates from a combined GOES infrared/TRMM microwave technique were analyzed with respect to percent forest cover from Landsat data (courtesy of TRFIC at Michigan State University) and GOES visible channel data over a deforested area in Rondonia (southwest Brazil). Five 1" x 1" areas of varying forest cover were examined during the onset of the wet season in Amazonia (Aug-Sept), when the effects of the surface would not be dominated by large-scale synoptic weather patterns. Preliminary results revealed that: maximum rainfall fell in most deforested area; heavily forested areas received the least rainfall; cumulus cloud development initiated at borders; the amplitude of the diurnal cycle of precipitation was a function of th surface cover. Further work will be presented detailing effects of land surface cover on the GOES infrared-measured surface heating, GOES visible observed cumulus development, thunderstorm initiation based on the location of temperature minima in the infrared data, and estimated rainfall and its diurnal cycle from a combined GOES/TRMM technique. Rainfall estimates derived from non-geosynchronous microwave observations (i.e. Goddard Profiling Algorithm, GPROF) will also be examined.