Semi-distributed, physically based, hydrologic modeling of the Paddle River Basin, Alberta, using remotely sensed data

Abstract

A hydrologic model (DPHM-RS) is designed to assimilate remotely sensed data in a semi-distributed approach. Energy and water flux exchanges at the land–atmosphere interface is assumed to be one-dimensional (vertical), ET is modeled separately for land and vegetation, and surface and sub-surface runoff from each sub-basin are routed to the stream by a kinematic response function and then via a network of stream channel to the basin outlet. Using the Digital Terrain Elevation Data (DTED), the Paddle River Basin (265 km 2) of Central Alberta is divided into five sub-basins. DPHM-RS was calibrated with hourly hydroclimatic and remotely sensed data collected in 1996 (July 24–August 29, 1996). Land use classes were derived from Landsat-TM data while topographic information was derived from DTED data. The calibrated model was separately validated with similar data of 1997 (July 16–September 26, 1997), and 1998 (May 1–June 30, 1998). Simulated runoff at the basin outlet showed good agreement with the observed counterparts at both calibration and validation stages, with a coefficient of efficiency, E f of 85% in calibration (1996), 60% (1997) and 61% (1998) in validation stages. DPHM-RS also simulated surface temperature, net radiation and soil moisture at the active layer that compared favorably with their counterparts retrieved from space platforms. Encouraging results from the above multi-criteria assessment demonstrate the feasibility of semi-distributed modeling and the usefulness of remotely sensed data in basin hydrology.