Comparison of Spatial Variability Techniques for Runoff Estimation from a Canadian Watershed

Accuracy in generation of surface contours from elevation points and digital elevation models assists in meaningful hydrological predictions and estimation of geomorphological parameters. Also, the accuracy in the generation of isochrones representing the spatial variability of surface water travel time over watersheds leads to a reliable estimation of the direct runoff hydrograph. In the present study, a deterministic method, inverse distance weighting, was compared with ordinary kriging, a geostatistical method for mapping the surface elevation and runoff travel time of the St.-Esprit watershed, Quebec, Canada. The contour map and isochrone map of the watershed were generated through exploratory data analysis, fitting of semivariogram models and kriging using ArcGIS® and GS+ software. The digital elevation model and surface elevation data acquired from Canadian Digital Elevation Database were subjected to geo-spatial analysis with reference to the delineated 2610 ha St.-Esprit watershed. The surface elevation points (488 numbers) and estimated travel times within the watershed were subjected to inverse distance weighting and ordinary kriging analysis. It was observed that the Gaussian semivariogram model fitted well (0·914