分布式水文-土壤-植被模式的改进及气候水文Off-line模拟试验

IMPROVEMENTS OF A DISTRIBUTED HYDROLOGY MODEL DHSVM AND ITS CLIMATOLOGICAL-HYDROLOGICAL OFF-LINE SIMULATION EXPERIMENTS

中文关键词:  水文模式改进  Off-line模拟试验  模拟效率系数

英文关键词:Improviments of hydrology model  Off-line simulation experiments  Hydrologic modeling efficiency

中文摘要:

      文中对引进的高分辨率(942.5m×942.5m)分布式水文-土壤-植被模式(DHSVM),针对海、滦河流域的特点进行了改进,主要包括:(1)改变蒸散发模拟方法,用改进的Penman-Monteith模式模拟海滦河流域的蒸发,较好地模拟出1a内的两个峰值,最大值出现在4～6月,次大值出现在7～8月;(2)改变水文模式结构,用多站点气候观测资料内插到模式网格点,充分实现了分布式水文模拟;(3)发展新的水文、植被、土壤参数化方案,对33个参数分区计算和确定,并重点对土壤孔隙度φ、土壤蓄水能力θ_fc、叶面指数LAI、随机阻抗γ_s等11个参数进行调试和修订,提高了水文模拟精度。用改进的DHSVM模式分别对滦河、桑干河流域蒸散发、地下水位、土壤湿度、土壤水下渗、产流、汇流与径流等水文过程进行Offline模拟试验,模拟结果与实测值一致性较好,滦河流域1979～1991年、桑干河流域1979～1987年水文模拟效率系数分别为0.89和0.82,均高于国内其他相关研究。

英文摘要:

      The advanced distributed hydrology-soil-vegetat ion model(DHSVM), developed by Wigmosta et al. (1994), is introduced from US Pacific Northwest National Laboratory. To apply DHSVM in China for the first time, some improvements have been implemented in terms of the basin characteristics: (1) To change evapotranspiration model, using the improved Penman-Monteith approach in replace of the original one; (2) To change the model structure, inserting datasets from 4 stations to grid cells for each river basin, instead of datasets from one or two stations; (3) To develop new hydrology, vegetation and soil parameterization schemes for improving the simulated results, with focus on calculation and adjustment of 11 parameters, such as soil porosity φ, field capacity θ_fc, leaf area index (I_LA), stochastic resistance γ_s, among the total 33 parameters. Then the improved DHSVM is driven by observed datasets for Luanhe River Basin and Sanggan River Basin, respectively. The simulated evapotranspiration (ET), runoff, snow water equivalent, water table, soil moisture and percolation are then gained as DHSVM outputs. The simulated ET shows that the highest peak appears in May or June instead of July or August. This is consistent with the real situations, owing to the improvement of ET model. The simulated runoff process and flood peak are quite consistent with the observed ones. The model efficiency values for Luanhe River and Sanggan River basins are 0.89 and 0.82, respectively, which shows high modeling ability of the model system for both relatively humid and dry basins.