Annual maximum flow regionalization of Colorado River System, Argentina

Abstract

The prediction of the maximum annual flow is necessary for flood management. Large amounts of hydrological information are required to make meaningful estimates. The Colorado River System basins have a topography that makes it difficult to maintain hydrometric stations, so there is a lack of continuity in records and in several cases there are ungauged basins. Regionalization methods consist of transferring information from gauged to ungauged sites in order to make predictions. The objective is to find regional regression models that relate the climate and morphometric characteristics of the basins with the maximum annual flow. For this purpose simple linear regression models were used. From this relationship and the regional frequency curve it will be possible to predict the maximum annual flows for different return periods in ungauged basins of the Colorado River System, Argentina. Regionalization models show that the best estimates occur when the predictor variable is the area and perimeter of the basin. Errors in the regionalization models of various sites in the system resulted between 6% and 67%. The models found are a tool for flood management in central-western Argentina.

Keywords

• regionalización
• crecida
• modelos de regresión
• período de retorno