Significant progress has been made during recent years in applying information technologies to water resource management from a multidisciplinary field called hydroinformatics; this can change the focus of environmental impact assessment and decision-making according to available environmental resources. This article presents an image retrieval system which uses the Universidad de la Amazonia’s (Colombia) Aquatic Ecosystem Quality and Preservation Research Group’s (CAPREA) digital image collection. Automatic picture classification (having attributes relative to water color, edges and flow type) is proposed which has been based on Dinnertein's classification, adapted by the Universidad Nacional de Colombia's Hydrology and Meteorology Institute and Environment Studies in 2001.