To assess the usability of satellite rainfall estimates for erosion studies in Cape Verde, the 3B42 and 3B43 products
fromTRMM, and theMulti-Sensor Precipitation EstimateMPE fromMeteosat are compared to daily and monthly
ground rainfall measured between 1998 and 2010. TRMMestimates from1998 andMPE from2009 were studied
and itwas found that they detect the general trend and direction of rainy clouds but underestimate the amount of
rainfall compared to rain gauges. To obtain an improved rainfall estimate fromMeteosat data, the cloud top temperature
derived from the 10.8 μm infrared channel was correlated to rainfall intensity measured at a 3-minute
interval between 2008 and 2010 by a Parsivel optical disdrometer. Amodified exponential equation of daily rainfall
intensity as a function of cloud top temperature was obtained. The equation was applied to 15-minute intervals
from the infrared channel to obtain 15‐minute rainfall and summed up as daily rainfall. The daily rainfall
depths estimated from cloud top temperature showed a better correspondence with the amounts measured
by ground gauges on the island (r = 0.75), and are a first approach to cope with lack of short duration rainfall
data for erosion studies in island states such as Cape Verde.