Publication Date



water balance, Devils Lake, remote sensing, GIS, evapotranspiration, runoff, land use/land cover


In this part of the study, grid-based spatial water balance approach was used to estimate the annual water balance of Devils Lake basin, hydrologically closed lake located in the Red River of the North basin, northeastern North Dakota. Landsat images from 1991 to 2003 were used in the study. Using spatial precipitation, land-cover and soils data, grid-based surface runoff was estimated based on the Curve Number method. The calibrated upstream runoff inflow for each grid cell was computed using a 10-m digital elevation model. Spatial evapotranspiration was estimated for the study area from remotely-sensed data using a surface energy flux model. The spatial water balance for each grid was constructed using grid geographic information system (GIS). The modeled average change in storage depth was compared to observed values of the lake stage. The grid GIS-based spatial surface water balance predicted the observed values with an average error of prediction of 0.12m. With better understanding of the groundwater contribution to the water balance, the prediction accuracy can be improved. The study ensures the applicability of the technique for surface water budget computation using GIS and remote sensing.