基于AnnAGNPS模型的罗李村流域水文模拟与评价

发布时间：2018-06-18 23:40

本文选题：AnnAGNPS + 罗李村流域　；参考：《陕西科技大学》2017年硕士论文

【摘要】：随着城市化的发展,世界范围内的供水危机日益严重,河流的水环境问题越来越受到人们的关注。分布式水文模型考虑了气候和下垫面的空间异质性,根据地形和水文特征将流域划分为若干计算单元,基于此模拟的流域水文物理过程更接近实际。科学合理的模拟流域水文过程为研究流域水环境的可持续管理提供科学依据。灞河属黄河支流渭河的一级支流,是西安市典型的城市河流。灞河流域地处西北半干旱区,降雨是流域径流的主要补给方式。AnnAGNPS分布式水文模型在灞河流域的应用未见报道。本研究以灞河上游罗李村流域为研究区,基于AnnAGNPS模型构建流域日尺度径流泥沙数据库,探讨流域最优分室划分,分割流域基流,分析径流泥沙的敏感性参数。在此基础上,运行模型模拟流域的径流过程和泥沙负荷并分析其模拟效果,评价Ann AGNPS模型在罗李村流域的适用性,为充分了解流域水文特征及合理管理流域环境提供理论参考。本研究取得的主要成果如下:(1)通过实验发现流域离散化参数初始源面积(CSA)为100 hm2,最小初始沟道长度(MSCL)为70m,模型将流域划分为1030个分室(Cell)和420个沟道(Reach),模型能够较好的反映流域地形、土地利用、土壤类型等下垫面特征。(2)采用加里宁算法对流域基流进行分割,得出流域基流指数为47.48%。(3)通过DSA进行模型参数敏感性分析发现:流域径流敏感性参数从高到低依次为径流曲线数CN、田间持水量、饱和导水率、沟道曼宁系数。泥沙敏感性参数依次为耕作后地表残留物覆盖率、地表覆被与作物管理因子、水土保持措施因子、土壤可蚀性因子、初始随机粗糙度、田间持水量、年根系生物量、耕作扰动面积、地表残留物覆盖率。此外,沟道曼宁系数对泥沙的输出无影响。(4)通过对模型进行率定和验证,模型在率定期(2001-2005年)和验证期(2006-2010年),年、月、日尺度径流模拟效率系数Ens均大于0.51,决定系数R2均大于0.55,模拟值低于实测值但相对误差Re均控制在10%以内。月尺度径流模拟效果最佳,Ens和R2均大于0.78。模型对大中等强度降雨事件的模拟效果优于暴雨,降雨事件的模拟偏差随降雨强度的增大而增大。径流的空间分布显示,流域内耕地产流能力最强,乔木林产流能力最弱。AnnAGNPS模型适用于罗李村流域的径流过程模拟。流域河段水环境容量的计算结果表明,灞河干流COD和NH3-N的水环境容量分别为3814t/a和223t/a,干流对流域水体的面源污染有一定的调节作用。(5)由于输沙过程的复杂性,与径流相比,模型对泥沙的模拟精度偏低,Re在-30%左右。月尺度率定期Ens为0.35,R2为0.36,验证期分别为0.27和0.35。总体上模型可以反映流域输沙量的变化。流域土壤侵蚀以微度为主。
[Abstract]:With the development of urbanization, the water supply crisis in the world is becoming more and more serious. The distributed hydrological model takes into account the spatial heterogeneity of climate and underlying surface and divides the watershed into several computing units according to the topography and hydrological characteristics. The hydrophysical process of the watershed based on the simulation is closer to the actual situation. Scientific and reasonable simulation of watershed hydrological process provides scientific basis for the study of sustainable management of watershed water environment. Bahe is a first-class tributary of the Weihe River, a tributary of the Yellow River, which is a typical urban river in Xi'an. The Bahe River basin is located in the semi-arid region of northwest China. Rainfall is the main recharge mode of runoff. The application of Ann AGNPS distributed hydrological model in Bahe River basin has not been reported. Based on AnnAGNPS model, the daily scale runoff and sediment database of the upper reaches of Bahe River was constructed, and the optimal partition of watershed basic flow was discussed, and the sensitivity parameters of runoff and sediment were analyzed. On this basis, the operation model simulates the runoff process and sediment load of the watershed and analyzes its simulation effect, and evaluates the applicability of the Ann AGNPS model in the Luoli village basin, which provides a theoretical reference for fully understanding the hydrological characteristics of the basin and for reasonably managing the basin environment. The main results of this study are as follows: (1) the experimental results show that the initial source area and minimum initial channel length are 100 mm ~ 2 and 70 m, respectively. The model divides the watershed into 1030 compartments (Cell1) and 420 channels (CSAs). The model can be compared with that of the model. A good reflection of the watershed topography, Land use, soil type and other underlying surface characteristics. The sensitivity analysis of model parameters by DSA showed that the runoff sensitivity parameters from high to low are runoff curve number CNN, field water holding capacity, saturated water conductivity and channel Manning coefficient. The parameters of sediment sensitivity were surface residue coverage, land cover and crop management factors, soil and water conservation measures, soil erodibility, initial random roughness, field water holding capacity and annual root biomass. Tillage disturbance area, surface residue coverage. In addition, the channel Manning coefficient has no effect on the sediment output. (4) through the rate determination and verification of the model, the model is at regular intervals from 2001 to 2005) and the verification period is from 2006 to 2010. The daily runoff simulation efficiency coefficients Ens and R ~ 2 are all greater than 0.51and 0.55 respectively. The simulated values are lower than the measured values, but the relative error re is controlled within 10%. The results of monthly scale runoff simulation showed that both Ens and R2 were more than 0.78. The simulation effect of the model is better than that of rainstorm, and the simulation deviation of rainfall event increases with the increase of rainfall intensity. The spatial distribution of runoff shows that the runoff production capacity of cultivated land in the watershed is the strongest, and the runoff production capacity of Arbor forest is the weakest. Ann AGNPS model is suitable for the runoff process simulation of Luoli village watershed. The results show that the water environmental capacity of 3814t/a and NH3-N are 3814t/a and 223t / a, respectively. The main stream has a certain regulating effect on the non-point source pollution of the basin water body. (5) because of the complexity of sediment transport process, compared with runoff, the water environmental capacity of the main stream of Bahe River is more complex than that of the runoff. The simulation accuracy of the model for sediment is about -30%. The monthly scale rate of the periodic Ens is 0.35 ~ 2 is 0.36, and the verification period is 0.27 and 0.35, respectively. On the whole, the model can reflect the change of sediment discharge. Soil erosion in watershed is dominated by microdegree.
【学位授予单位】：陕西科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P333

【相似文献】