Significant portions of Central America experienced flooding in recent years from hurricanes Mitch (1998) and Keith (2000), making quantitative measurements of precipitation in this region very important for predicting and mitigating the effects of flooding. However, a number of the raingauges in the already sparse network have been lost in natural disasters or discontinued because of maintenance costs. In addition, large areas of the region are inaccessible, making direct rainfall measurements extremely difficult. Satellite precipitation techniques represent a valuable alternative for estimating precipitation over the region for flash flood forecasting, as well as for daily and monthly precipitation estimates for hydrological, climatological and agricultural studies.

The Central American weather services will have access to GOES digital data after July 2001, and efforts are underway to validate several satellite rainfall algorithms in order to determine their suitability for Central America. As part of this study, two algorithms have been used to estimate rainfall from hurricane Keith from 30 September to 2 October 2000, when the heaviest precipitation was observed. The estimates were compared to available raingauge data (radar data were not available) in four different ways in order to account for navigation errors and differences in representation between point and areal measurements. The results indicate that satellite techniques are a promising tool for estimating precipitation over Central America.