Flow routing method and its application in distributed hydrological modeling
分布式水文模拟汇流方法及应用

Flow routing model (FRM) plays important roles in the distributed hydrological modeling. Whether a simple or complex FRM is chosen depends on the relative relationship between the time resolution and the lagging time of routing. Generally, the FRM of a distributed hydrological model(DHM) consists of three components: the first is subdivision method of the basin; the second is the routing path; and the last is the routing method based on the routing path. In order to decrease the runtime of model, an efficient routing method, the so-called Ranked-Grid-based Kinematic Wave Routing model (RGKWR), is developed in this paper. This model applies the kinematic wave method to successive routing on ranked raster grids. Raster grids can be divided into different ranks due to each grid's flow direction computed by the single direction algorithm. We define the basin's outlet grids as the first rank, and the grids from which water flows into the first ranked grids as the second rank, the rest can be deduced by analogy, so the most upstream grids are defined as the highest rank. Flow routing is undertaken from those grids having higher rank onto the grids having lower rank.This paper puts forward the classifying method for raster grids, discusses the kinematic wave model and its numerical solutions, and carries out a case study on Chaobaihe River Basin in North China. The basin, with a drainage area of 13,846 km 2, has two outlets to Miyun Reservoir: one is Xiahui hydrologic station on Chaohe River, the other is Zhangjiafen station on Baihe River. The whole basin is divided into 55,444 grid elements at a 500-m resolution, with the Chaohe River basin partitioned into 298 ranks, and Baihe River basin 398 ranks. The application shows RGKWR based on GIS/DEM is reasonable and efficient.