Analysis of seasonal variation in cumulus cloud frequency of occurrence

The GOES 08 and GOES 12 data were used to estimate the cumulus cloud frequency of occurrence as a means of determining the affects of deforestation on boundary layer processes, area climate, and especially rainfall (e.g., rainy season onset) across portions of Central America. The hypothesis is that because cumulus clouds occur when turbulent boundary layers develop, due sensible heat flux, deeper and more widespread clouds are likely over de-forested regions versus virgin forested. The cloud classifier developed by /Berendes et al/., [2008] is used to identify and classify the various types convective cloud cover, with cloud classes being "fair weather" cumulus, towering cumulus, cumulonimbus, etc. The classifier was applied to data from March, May, June and September of 2000 to 2008 to observe the dry and wet season cumulus cloud frequency so to discern trends in dry-season length. We see that the frequency of cumulus clouds are higher during the dry season and gradually reduces towards the wet season. This agrees with the results by /Durieux et al/., [2003], where they see more low level clouds during the dry season and high convective clouds during the wet season. Also among the different habitats, the deforested locations show highest frequency of cumulus clouds than the forested and partially forested regions. This study therefore has implications to regional hydrology across Central America, and other Tropical regions, as forest character and extent change over time due to human influences.