Water Balance for the Ji-Paraná River Basin, Western Amazon, Using a Simple Method through Geographical Information Systems and Remote Sensing

Abstract

The Thornthwaite?Mather climatological model integrated into a Geographic Information System (GIS) was used to simulate the monthly water balance for the Ji-Paraná river basin, in the western Amazonian state of Rondônia (RO), from February 1995 through December 1996. Model inputs were Advanced Very High Resolution Radar (AVHRR) images, rainfall from rain gauges within and around the basin, soils, and land cover maps. Mean monthly temperature images, estimated from AVHRR, were used to calculate monthly potential evapotranspiration images. Rainfall measurements were interpolated to obtain monthly rainfall images using the spline method. Results were compared with annual water mass balance for 10 internal subbasins. Model results underestimated basin evapotranspiration in ?6% and water surplus was overestimated in +12%. Annual basin evapotranspiration was 1166 mm, or 3.2 mm day?1, with a minimum in the dry season (1.7 mm day?1 in July) and maximum in the beginning of the wet season (4.2 mm day?1 in October). Three distinct land cover scenarios were evaluated. Complete basin deforestation reduced basin evapotranspiration and increased water deficit and surplus in approximately 200 mm yr?1.