Large-scale Hydrological Modelling for Flow Prediction in the Weihe River Basin

摘要

The objective of this study is to develop a capable large-scale hydrological model (LSHM) for flow prediction in Weihe River basin, which is the largest tributary of the Yellow River with an area of 136,000km2. The geological formation, hydro- meteorological conditions and human activities are complex and variable in this area. This study was performed based on a distributed catchment modelling system developed at Hydrology core, UNESCO-IHE. The required data include two categories: one is basic data, consisting of digital elevation map, land use data, soil type data, river network and river geometry, the other is time-depending data, consisting of daily rainfall, runoff and actual evapotranspiration. Data preparation for this study mainly includes catchment schematization, daily areal rainfall and actual evaporation calculation. Catchment sehematization consists of DEM derivation, land use and soil type map digitization, cross section schematization and sub-basin delineation. Daily areal rainfall and actual evaporation were calculated based on 1980~1983 4 years station-measured daily rainfall and monthly pan evaporation data. Specifically, daily rainfall in each 4.5km×5.5km grid cell of the basin was generated by using a spatial interpolation toolkit (Hykit) based on inverse distance weighting method. The daily actual evaporation in each grid cell was derived following sequential steps. Firstly, the potential evaporation was derived from the pan evaporation using proper pan coefficients and seasonally varied crop factors based on the land use map. The monthly actual evaporation was estimated using Thornthwaite-type monthly water-balance model method. Finally, the daily variation of the actual evaporation is achieved by consulting the variation of open water evaporation on sunny day and rainy day based on available data from the year 1980～1983. Model parameterization was done by establishing river segment geometry, Manning's roughness, diffusivity and initial water content deficit. Following, 1980~1983 4 years flow simulation was performed combining with model calibration, performance evaluation shows the present model has high capability for hydrological modelling in large- scale basin. Afterwards, flow prediction for flood season of the year 1983 was implemented on the calibrated model and validated by observed flow. The result indicates that the model is reasonably good for one day flow prediction by combining measured discharge at upstream stations.