Theories, Methods And Applications Of A Physically-based Distributed Hydrological Model--GHM Model

Spatial heterogeneity of physiogeographical components leads to spatial and chronological changes of hydrological processes in watersheds. These effects were not taken into consideration by the traditional black box models and lumped models. In order to simulate the hydrological processes and calculate hydrological effects of forests and other vegetations more precisely, it is essential to establish a physically-based distributed hydrological model and upscale it from small watershed to medium and large watershed based on upscaling methods . Guansihe river basin, a small-scale watershed in the upper and middle reaches of the Yangtze River, is selected as a case study area in this dissertation. Guansihe Model (GHM), a physically-based distributed hydrological model, is established and run. Software package of this model is compiled. The possibilities of applying this model to medium or large watersheds are discussed.Guansihe River Basin is typical in physiogeographical conditions in the shelter forest area of upper and middle reaches of Yangtze River. Parameters for the GHM Model are obtained by measurements in the fields. In the GHM Model, a watershedmModel are obtained by measurements in the fields. In the GHM Model, a watershed is first divided into square-form cells. The size of cells represents spatial resolution. When it is applied to the Guansihe River Basin, the entire watershed is divided into 7378 cells with a size of 50 m X 50m. Each cell is horizontally homogeneous in elevation, soil and vegetation conditions. The physiogeographical features of a cell determine its hydrological processes. A cell is divided into layers vertically. Four layers (plant community canopy layer, litter layer, surface layer and soil layer) from top to down, are divided. Values of topographic, soil and vegetation conditions are given under the supports of GIS tools Arc/ Info and Idrisi. Interception, infiltration and formation of runoffs are then simulated for each cell. Water movement in the watershed is simulated based on water routes among cells and layers.Tests by 4 rainfall-runoff processes in the Guansihe River Basin have demonstrated that the GHM Model is applicable. This model can be further applied to jthe assessment of ecological effects of watershed management. A simulation of land cover change to runoff shows that an increase of forested area is not the only factor in the flood peaks. The changes of hydrological processes have also relations with soil conditions.Upscaling is widely discussed by both forest hydrology and landscape ecology. The GHM Model makes upscaling from site to area, from small area to large area, possible. A comparison with physically-based distributed hydrological models established overseas shows that the GHM Model has the potential to use in other watersheds with larger scales.