Evaluation of flooded areas with satellite imagery using an objective hydrologic criterion

Abstract

A method for computing flooded areas in relation to precedent climatic conditions in a plain area is evaluated. The studied area includes the most important production area for cattle breeding in Argentina; therefore, waterlogging periods create significant economic losses. In order to conduct this study, Landsat images that represent different hydrological conditions were used. The method is based on the frequency analysis of rainfall records of the 30, 60 and 120 days prior each image, thus obtaining a seasonally weighted mean frequency. In order to minimize biases of individual images, the images were combined linearly so as to obtain composed images with the desired antecedent precipitation frequency. Then, the flooded areas were associated with the frequency of the antecedent rainfall. Therefore, frequencies equal to or higher than 50% are associated with semi-permanent or permanent lentic waterbodies. Consequently, lower frequencies are associated with waterlogged areas reaching 35% of flooded area for a 20% precipitation frequency. © 2012 CIWEM.