A DISTRIBUTED HYDROLOGICAL MODELLING APPROACH TO FLOW FORECASTING OF A LARGE RIVER BASIN

摘要

A spatially distributed hydrological modelling approach for a large river basin with a lumped vertical process representation is presented. The model comprises a 2-dimensional distributed land component and a relatively dense representation of the river network. The flow in the land component is based on a diffusion process and the river flow routing is based on the Muskingum-Cunge method. The diffusivity for the land component is defined as a fuzzy parameter and is determined from spatial information on soil type, vegetation and land use. The model uses rainfall, evapotranspiration and snowmelt data derived from remote sensing information as inputs. The modelling approach is applied to the upper reach of the Yellow River Basin, approximately 125 000 km~2, in China for flow forecasting. The preliminary validation results for the years 2000 and 2001 show reasonable performance of the model.