Application of Distributed Hydrology-Soil-Vegetation Model in Pang Khum Experimental Watershed, Chiang Mai Province, Thailand
Abstract
Distributed hydrology-soil-vegetation model (DHSVM) uses a digital elevation model (DEM) and accounts for topographic effects on soil moisture, groundwater, and surface water redistribution in a complex terrain. In this study, DHSVM is used to simulate soil moisture, net radiation and stream flow in a 1-km2 tropical mountainous watershed in Pang Khum, Chang Mai, Thailand. Pang Khum Experimental Watershed (PKEW) has two meteorological stations, four soil moisture stations, and one stream flow station at basin outlet. Meteorological measurements are used as forcing data for DHSVM. The grid resolution for this simulation is 50 m. Initial soil and vegetation parameter settings based on field measurements and literature review are adjusted through model calibration. The model is run for a six month warm-up period, followed by a calibration period of approximately one year. Validation is done for two periods totaling 18 months. At the forested site, net radiation is reasonably well simulated, although underestimated in the dry season, and overestimated in the wet season. At the agricultural site, net radiation is consistently overestimated. Soil moisture is well simulated at the forest site. In the simulation, the water table rises into the soil zone during the wet season, saturating all three soil layers at the agricultural site; measured values remained at unsaturated levels. Baseflow is significantly underestimated in calibration and validation periods. Difficulty in simulating streamflow may be caused by road-related effects in the basin. Our prior field work has shown that the road significantly alters runoff in PKEW. The principal mechanism of road-induced effects is Horton Overland Flow (HOF) generated on the road surface. In its present form DHSVM can account for interception of subsurface flow by roads, but not HOF generated on the road. We do not think subsurface flow interception is important in PKEW, and have therefore not implemented the road in our simulation. In future applications, we hope to include road effects, including HOF generation. As far as we know, this is the first application of DHSVM in a tropical location. Also, we apply the model over a very small watershed using a smaller grid-cell size than prior applications. Overall, we find the model to perform reasonably well despite being applied in a region and at a scale that contrast strongly with those in which it was developed. >http://webdata.soc.hawaii.edu/climate/Roads1/Roads.html</a>
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2001
- Bibcode:
- 2001AGUFM.H21C0320C
- Keywords:
-
- 1860 Runoff and streamflow;
- 1866 Soil moisture;
- 1878 Water/energy interactions;
- 3322 Land/atmosphere interactions