Evaluación de la efectividad de las imágenes del sensor GPM para la representación de la precipitación en el departamento del Meta

Abstract

The variability of precipitation in the Meta department is complex, not only due to the different subregions that compose it and the environmental climatic conditions that influence it, but also because its study is determined by the availability of traditional rainfall information, which for this case is scarce considering the area and distribution in a way that does not represent the reality of the behavior of the variable. The present investigation tries to evaluate the effectiveness of the satellite images of the GPM sensor to represent the precipitation in the department of Meta for the period 2014-2018, taking into account the observed data (meteorological stations) and the estimated ones (data from the IMERG V06 product) , and thus be able to validate the suitability of the remote sensor to estimate precipitation, taking into account performance measures such as the correlation coefficient, the determination coefficient and the covariance.The validation showed that the images correctly represent the precipitation behavior, but it cannot adequately represent extreme precipitation events. For the dry months the correlations were higher than for the rainier months, while the correlations of the plateau/high plain and foothills/piedmont were better than those of the Amazon plain, always showing how satellite images underestimated actual precipitation. In general, the correlation for the stations was 0,82, which is considered a good correlation.