南水北调中线水源区堵河流域产流产沙对土地利用变化的响应

发布时间：2018-04-04 18:52

本文选题：堵河上游流域　切入点：基流　出处：《华中农业大学》2016年博士论文

【摘要】：水土流失是全球化的重大环境问题。水土流失与水环境恶化、河道功能的破坏和洪涝灾害的发生具有联动效应,严重影响人类生存环境和生态安全。流域产流产沙是衡量水土流失的重要指标,受多种因素影响,包括气候、地形、土壤、土地利用等,其中,气候因素能够在较长时间尺度上对流域水沙演变产生重要影响,而土地利用变化则是短期内人类活动影响水沙最直接的表现形式。研究流域产流产沙对土地利用变化的响应并定量分析其耦合关系,是当前环境背景下流域管理的基础。为了实现水资源的宏观调配,我国启动了南水北调中线工程,其水源区生态系统相对脆弱,面临潜在的水土流失、洪涝频发等环境问题的困扰。本文以水源区典型流域——堵河流域为研究对象,在国内外水沙响应研究的基础上,根据地形地貌、气象水文、土地利用和土壤等数据,借助地理信息等技术手段并结合野外调查建立流域空间和属性数据库,选取Soil and Water Assessment Tools(SWAT)模型进行无资料区域水沙模拟,选择双参数对数正态分布函数进行极端水文非稳定频度分析,综合采用偏最小二乘回归等多元回归工具量化了流域土地利用对产流产沙变化的相对贡献。论文取得的主要成果有:(1)以南水北调中线水源区堵河上游流域为研究对象,在子流域尺度上基于GIS平台求得土地利用面积转移矩阵和空间转移图并分析了土地利用空间变化特征,结果表明,1990年代林地的砍伐受到林分与道路和航道距离的影响,耕地和城镇用地多沿河道增加,2000年代国家退耕还林政策实施后,部分坡度大于25°的耕地转换为林地;与基线期(1980年代)相比,耕作期(1990年代)子流域土地利用空间变异程度增加,而生态恢复期(2000年代)土地利用空间变异程度在此基础上进一步增加,同时导致了土地利用格局的空间变化。(2)以堵河上游流域卡口站——竹山站和内部的新洲水文站长时间逐日径流数据为基础,借助“Bflow.exe”数字滤波程序进行基流分割,将自动校准和手动校准相结合进行月尺度径流和基流深度的SWAT模型校正和验证,进而预测107个子流域基流深度,结果表明,在流域尺度,堵河上游流域基流系数为0.358(竹山站)和0.364(新洲站),竹山站和新洲站在1980年代、1990年代、2000年代的月均基流深度分别为19.0和17.9、14.7和13.3、16.6和15.7mm,月间差异明显;在子流域尺度,swat径流和基流模型有较好的模拟效能,基线期(1980年代的土地利用)、耕作期(1990年代的土地利用)和生态恢复期(2000年代的土地利用)下的子流域月均基流深度分别为16.9、18.4和15.9mm,基线期基流深度空间变异程度最低,生态恢复期最高。(3)结合偏最小二乘回归在子流域尺度上量化土地利用对基流的相对重要性,结果表明,土地利用是基流变化的重要影响因素,林地、耕地和城镇用地对基流的影响显著且均呈现负相关关系,其中林地的相对贡献最大,灌丛、草地和未利用地对基流的作用不显著;对于林地被农田所取代的区域,蒸散量的强度和持续时间的降低导致基流增加,对于林地转化为城镇用地的区域,城镇用地对基流的负面作用被无贡献的不透水表面所削弱,基流增加;耕地和城镇用地对基流的相对重要性因季节而异,在夏季,耕地蒸散量较大,对基流产生的负面作用大于城镇用地,在其他季节,耕地的作用小于城镇用地。(4)在swat月尺度径流模型的基础上校验日尺度径流,并获取107个子流域逐日径流数据,选用年最大径流深度序列并借助双参数对数正态分布函数和极大似然估计进行非稳定性极端径流分析,在不同重现期下定量分析土地利用对极端径流事件频次的影响,结果表明,径流极值分位数分布规律在不同重现期下差异明显,2年重现期下,径流极值分位数从西南向东北部降低,在20年、100年和1000年重现期,较高的径流极值分位数出现在西北和东南部;各个年代的极端径流事件频次在较小或较大重现期内呈现相似的变化规律,而较小和较大重现期之间则表现出明显差异,1980年代极端径流风险最高,2000年代最低,洪水季明显高于平水季;土地利用对极端径流事件频次的影响随重现期的增加而减弱,当重现期大于100年,土地利用类型和极端径流事件不存在显著相关关系,在较低重现期下,林地能显著降低极端径流事件频次,耕地和城镇用地则能够增加极端径流风险,草地、灌丛、未利用地对极端径流事件的影响不显著,在较大重现期下,水库等水利设施能够对极端径流事件产生重要影响。(5)在swat径流模型的基础上校验了产沙和泥沙输移比模型,借助偏最小二乘回归探讨了堵河上游流域土地利用组成和格局对特定产沙的相对重要性。基于土地利用和自然地理特征选择了19种诠释性指数在子流域尺度下进行分析,借助变量对投影的影响(vip)值识别和量化影响因子对特定产沙变化的相对贡献,结果表明,堵河上游流域土地利用组成和格局对特定的土壤侵蚀量的影响最大,解释了65.2%的特定产沙量的变化,VIP值表明,子流域尺度下对特定产沙起主控作用的第一顺序因子为:耕地和林地所占的面积百分比、斑块密度、香农多样性指数、土壤可蚀性、蔓延度、高程积分值、土壤饱和导水率,流域面积对泥沙输移比影响显著,香农多样性指数、蔓延度和高程积分值同样对泥沙输移比有重要影响。
[Abstract]:Soil erosion is a major environmental problem of globalization. Soil erosion and deterioration of water environment, river function damage and occurrence of flood disaster has linkage effects, serious impact on the living environment of mankind and ecological security. Runoff and sediment is an important indicator of soil erosion, affected by many factors, including climate, topography, soil, land use, the climate factors can be in a long time scale of basin water and sediment evolution have an important impact, and the land use change is the short-term impact of human activities is the most direct form of water and sediment. The research of runoff and sediment in response to land use changes and quantitative analysis of the coupling relationship is the watershed management foundation under the background of the current environment. In order to realize the macro allocation of water resources, our country started the South-to-North Water Transfer Project, the water area ecosystem is relatively fragile, facing the potential of soil Flood loss, frequent environmental problems. Based on the typical basin water district, blocking the river basin as the research object, based on the domestic and foreign research on the response of sand, according to topography, meteorology and hydrology, land use and soil data, with geographical information technology and combining the field investigation of the watershed spatial and attribute database select the Soil and Water Assessment, Tools (SWAT) model was presented to simulate the water and sediment regional data, choose two parameter lognormal distribution function for unsteady extreme hydrological frequency analysis, using partial least square regression regression tool to quantify the relative contribution of the land use change of runoff and sediment. The main achievements of this paper are: (1) in the water source area blocking river basin as the research object, in the sub basin scale based on the GIS platform to obtain land use area transfer The transfer matrix and the spatial map and analysis of land use spatial variation, the results show that the forest felling is affected by the 1990s stand with road and waterway distance, cultivated land and urban land along the river, in 2000s the national policy of returning farmland to forest after the implementation of conversion part of slope greater than 25 degrees and cultivated land; baseline period (1980s) compared with the cultivation period (1990s) increase the spatial variability of land use sub basin, and ecological recovery period (2000s) further increases the spatial variation of land use degree, which leads to the land use spatial pattern changes. (2) to block the river basin bayonet - Xinzhou daily runoff hydrological data Zhushan station and station time inside the basis of base flow separation by means of "Bflow.exe" digital filtering process, the automatic calibration and manual calibration combined monthly scale size SWAT model flow and base flow depth calibration and validation, and to predict the 107 sub basins of base flow depth, results show that the basin scale blocking river basin base flow coefficient is 0.358 (Zhushan station) and 0.364 (Xinzhou station), Zhushan station and Xinzhou station in 1980s, 1990s, 2000s the average monthly basic flow the depth were 19 and 17.9,14.7 and 13.3,16.6 and 15.7mm, the obvious differences among months; scale in the sub watershed, runoff and base flow SWAT model has better simulation performance, baseline (land use, cultivation period (1980s) of land use in 1990s) and ecological restoration period (land use in 2000s) under the sub Basin the average monthly flow depth were 16.9,18.4 and 15.9mm, baseline based flow depth spatial variability and the lowest level of ecological restoration stage is the highest. (3) combined with partial least squares regression were used to quantify the relative importance of the basic flow of land in the sub basin scale, the Show that the land use is an important influence factor, the change of the base flow of woodland, arable land and urban land on the base flow significantly and the relationship is negative, the relative contribution of forest, shrub, grassland and unused land on base flow is not significant; for forest farmland replaced region. Reduce the evapotranspiration intensity and duration of the resulting increase in base flow, for the conversion of forest land into urban land area, impervious surface weakened by the negative effects of urban land use on the base flow is no contribution, base flow increased; the relative importance of cultivated land and urban land based on flow varies seasonally in the summer. A large amount of cultivated land, evapotranspiration, negative effect produced on base flow is greater than the urban land, in other seasons, the role of land less than urban land. (4) based on SWAT monthly runoff model calibration on scale runoff, and obtain 107 sub basins daily The annual runoff data, the maximum runoff depth sequence and with two parameter lognormal distribution function and maximum likelihood estimation of non stability analysis in the period of extreme runoff, quantitative analysis of the effects of land use, the results show that the frequency of extreme runoff events in different return, runoff extreme quantile distribution in different return period differences, 2 year return period, runoff extreme quantile decreased from southwest to northeast, in 20 years, 100 years and 1000 return period, the higher the runoff extreme quantile in the northwest and Southeast; extreme runoff events in each frequency in smaller or larger reproduction period showed a similar variation, and between the small and the larger the return period exhibited significant differences in 1980s, the highest risk of extreme runoff, the lowest in 2000s, the flood season was significantly higher than that of normal water season; land use frequency of extreme runoff events With the increase of the return period of the influence weakened when the return period of more than 100 years, the land use types and extreme runoff events have no significant correlation, at low return periods, forest land can significantly reduce the frequency of extreme runoff events, cultivated land and urban land can increase the risk of extreme runoff, shrub, grassland, unused land the impact of extreme runoff events is not significant in a large return period, reservoirs and other water conservancy facilities can have an important impact on the extreme runoff events. (5) based on the SWAT Model on the calibration of runoff sediment production and sediment delivery ratio model by partial least squares regression of blocking river basin land use pattern and composition the specific sediment yield. The relative importance of land use and natural geographical feature selection 19 interpretive index analysis in sub basin scale based on the impact of variables on the projection (VIP) and the amount of value identification The relative contribution of specific factors on sediment production results show that the effect of blocking river basin land use composition and pattern on soil erosion in particular the maximum, to explain the variation of the specific sediment yield of 65.2%, VIP values show that the sub basin scale of specific sediment yield from the first order factor controlling effect as the percentage of area of cultivated land and forest land accounted for the patch density, Shannon diversity index, soil erodibility, spreading degree, elevation integral value, soil saturated hydraulic conductivity, the basin area of significant sediment transport ratio influence, Shannon's diversity index, contagion and the integral value of the same height on sediment delivery ratio has the important influence.

【学位授予单位】：华中农业大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：P333

【参考文献】