植被变化的流域生态水文响应是水文学研究前沿。开展植被变化驱动下的流域生态水文效应研究，会促进水文学与植物学、生态学、泥沙动力学、水环境科学等方面的融合与交叉，从而推动水文科学研究的深入发展。黄河的水沙变化事关流域水资源配置、洪涝灾害预警、水生态保护等民生与环境问题，受到历代管理者的高度关注。自1980s尤其是2000s以来，伴随黄土高原水保建设的成效逐年显现，黄河的来水来沙量显著减少，给黄河流域综合开发带来重要影响。分析黄河水沙情势变化特点，研究水沙变化的内在机理，揭示水沙变化成因和发展趋势，对制定治黄战略和措施至关重要。针对流域尺度次降雨产沙分布式估算还有待研究、黄河水沙变化成因还不甚明晰的现状，本文以黄河中游河口镇——潼关区间为研究区，运用分布式流域暴雨产沙模型对流域单元次降雨产沙开展模拟，分析1980s～现状年流域单元植被变化对次降雨侵蚀产沙的影响，揭示黄河中游区次降雨侵蚀产沙对植被变化的响应特征，藉此为分析黄河水沙变化成因提供基础。通过研究，形成的主要成果和结论如下：(1)分析黄河中游区的林草植被变化和次降雨水沙变化，结果表明，黄河中游区林草数量有所增加，林草质量明显改善；水沙变化显著，大部分支流产流输沙量下降。构建土地利用变化重要性指数、土地利用变化面积比重、林草植被变化指数、植被盖度变化指数等指标，分析黄河中游区1980～2010的土地利用与林草数量变化、植被盖度与林草质量变化。结果显示，与1980年相比，2010年土地利用结构有所调整，林草数量有所增加，林草质量有所改善。土地利用变化面积55022km2，占总面积的21%。高覆盖度林草面积比重从37.2%上升到45.64%，平均植被盖度从28.17%增加到43.7%。分析流域次降雨的降雨—径流、径流—输沙关系，对比流域1980s和2000s的平均雨强、平均径流系数和平均产沙系数。结果显示，雨强变化与水沙变化的关系较为复杂，有3条支流的雨强变化与水沙变化呈负相关关系，有7条支流的雨强变化与水沙变化呈正相关关系。计算径流系数、产沙系数与雨强和植被盖度变化的相关性，发现径流系数、产沙系数与植被盖度增加的相关性大于其与雨强减小的相关性。表明植被变化对流域次降雨径流产沙有所影响，流域次降雨侵蚀产沙对植被变化的响应有待深入研究。(2)基于LCM和MUSLE构建了流域分布式暴雨产沙模型，开展了流域次降雨产沙分布式模拟，结果表明，该模型在黄土高原具有更强的适用性。开展下渗率r值的空间离散化后，暴雨径流模型LCM实现了从集总式向分布式的发展，较好地刻画了次降雨的产流特征，可为MUSLE分布式模拟提供必备的径流因子。运用暴雨径流模型LCM估算MUSLE所必须的径流因子，并增加坡面汇沙和河道汇沙过程，可获取流域出口产沙量，克服了以往MUSLE模型获取流域出口产沙量时大多通过栅格产沙简单加总的不足，实现流域单元次降雨产沙分布式模拟。针对研究区灌草面积大、水保措施密集、地形崎岖的自然特点，选用当地研究成果，修正MUSLE中C、P和LS因子，并对修正后MUSLE模型的适应性加以分析。结果显示，运用蔡崇法等(2000)提出的C因子估算方法、王万忠和焦菊英(1996)提出的P因子赋值方法，以及Liu等(1994，2000)提出的LS估算方法对MUSLE中的C、P和LS三个因子同时修正后，MUSLE估算黄土丘陵沟壑区的流域次降雨产沙量效果更好。 (3)估算了林草植被变化的减沙效果，结果表明，林草植被变化后，流域的减沙比例为16.73%～80.25%，林草数量增加是减沙的主控因素。通过情景设置的方法，计算相同降雨不同植被条件下的次降雨产沙变化。结果显示，林草植被变化的减沙效果显著，减沙比例为16.73%～80.25%。其中，除偏关河偏关以上支流减沙减沙较少外，其余流域的减沙比例均在50%以上。通过相关分析方法，计算出次降雨的平均减沙比例与林草用地面积占比变化指数的相关系数为0.858，与植被盖度变化指数相关系数为-0.255。结果表明，流域次降雨产沙减少量与林草用地数量增加有显著的相关性，与林草植被质量变化相关性不显著。当林草用地面积占流域比重约85%时，次降雨产沙系数趋于稳定，次降雨产沙很少。(4)分析降雨特征和流域性状对林草植被减沙效果的影响，结果表明，降雨历时是降雨特征中影响林草植被减沙效果的主要因素，流域地表粗糙度是流域性状中影响林草植被减沙效果的主要原因。分析林草植被—减沙弹性系数分别与雨强、雨量、降雨历时的偏相关性，发现林草植被—减沙弹性系数与雨强、雨量、降雨历时的偏相关系数分别为0.536、-0.740、0.834。可看出，在流域尺度的次降雨事件中，降雨历时对林草植被的减沙敏感性影响较大，流域次降雨的历时越长，次降雨的产沙变化对林草植被数量的变化越敏感，植被增加后产生的减沙效果越明显。分析林草植被—减沙敏感性分别与流域地表粗糙度、流域形状指数和河网密度的偏相关性，发现林草植被—减沙弹性系数与流域地表粗糙度、流域形状指数和河网密度的偏相关系数分别为-0.862、-0.213、-0.476。可看出，在流域次降雨事件中，流域地表粗糙度对林草植被的减沙敏感性影响较大。流域地表粗糙度越小，次降雨的产沙变化对林草植被数量的变化越敏感，植被增加后产生的减沙效果越明显。(5)分析林草植被变化的减沙效果的区域分异规律，结果表明，受流域地表粗糙度的影响，次降雨侵蚀产沙对林草植被变化的响应敏感性由北向南存在区域差异。分析林草植被—减沙弹性系数分别与降雨历时、地表粗糙度的偏相关性，发现林草植被—减沙弹性系数与降雨历时、流域地表粗糙度的偏相关系数分别为-0.764和-0.962。可看出，林草植被—减沙弹性系数与流域地表粗糙度的偏相关性大于其与降雨历时的偏向关性，流域地表粗糙度较降雨历时对植被减沙效果的影响更明显。随着流域地表粗糙度的变化，次降雨产沙对流域林草植被变化的响应敏感性在黄土丘陵沟壑区存在区域差异，呈现由北向南减小的趋势。位于研究区北部的黄土平岗丘陵沟壑区，次降雨产沙对植被变化的响应最敏感，植被增加后产生的减沙效果最好。位于中部的黄土峁状、梁状丘陵沟壑区次之。位于南部的干旱黄土丘陵沟壑区，次降雨产沙对植被变化的响应敏感性最小，植被增加后产生的减沙效果不显著。总之，论文分析了黄河中游林草植被数量与质量变化，在此基础上，基于LCM模型与MUSLE模型，开展了流域次降雨产沙分布式模拟，估算了相同降雨不同植被条件下的次降雨产沙变化，揭示了次降雨产沙对流域植被变化的响应特征，可为研究黄河水沙成因提供基础。

Eco-hydrological responses on the forest-grass vegetation change of watershed is a frontier research of hydrology. To study on eco-hydrological effects drived by forest-grass vegetation changes of watershed, will promote the interdisciplinary research between hydrology and botany, ecology, sediment dynamics, and water environmental science, so as to promote the further development of hydrological science research.Water and sediment variation of the Yellow River closely related to regional water resources allocation, Flood forecasting and warning, Water environment protection, is highly concerned by the water resources management department. Since the 1980s especially 2000s, with the effect of soil and water conservation engineering construction appeared in the Loess Plateau, the incoming water and sediment of the Yellow River had a significant reduction, bring important influence to the comprehensive development of the Yellow River River basin. Analysing the characteristics of water and sediment change, studing on the intrinsic mechanism of water and sediment variation, and revealing the causes of changes and development trend of water and sediment, are very important to establish the governance strategies and measures of Yellow River.Aiming at the situation that the rainfall-sediment estimation at the basin scale remains to be studied and the causes of incoming water and sediment changes of Yellow River water is still not clear, this study takes the reach from Hekouzhen to Tongguan of the Yellow River as study area, simulates the rainfall-erosion-sediment at watershed scale using a distributed rainfall-erosion-sediment model, analyses the effect of forest-grass vegetation changes of watershed on rainfall-erosion-sediment during the period from 1980s to present, reveals the characteristics of the rainfall-erosion-sediment responses on the forest-grass vegetation changes in the middle reaches of the Yellow River, and then provides a basis for analyzing the causes of the water and sediment variations of the Yellow River.The results are shown bellow:(1) To analyze the forest-grass vegetation change and rainfall-erosion-sediment variation in the middle reaches of the Yellow River. The results indicate that the quality and quantity of forest-grass vegetation changed remarkably in the period 1980–2010, and the runoff and sediment of most tributaries are decreased.This study constructs five indicators, including (a) a land use change significance index, (b) a land use change intensity index, (c) a vegetation change index, (d) vegetation coverage change index, and (e) vegetation coverage grades change index, to analyze the quality and quantity change of vegetation in the period 1980–2010. The results indicate that: Land use changed remarkably in the period 1980–2010. The area of land use change was 55,022 km2, accounting for 21% of the study area. The proportion of high coverage vegetation increased from 37.2% to 45.64%. The average vegetation coverage improved from 28.17% to 43.7%. The study also analyses the rainfall-runoff, runoff-sediment relationships in watershed scale, compares their average rainfall intensity, runoff coefficient and sediment coefficient. Results show that rainfall intensity change negatively correlated runoff and sediment change in three watersheds, while that rainfall intensity change positively correlated runoff and sediment change in the other seven watersheds. By analyzing the correlation of runoff and sediment change with rainfall intensity and vegetation coverage change, and the results indicate that the correlation of runoff and sediment change with forest-grass vegetation increase was more obvious than the rainfall intensity reduction, the rainfall-erosion and sediment yield responses on the forest-grass vegetation changes of watersheds need to be further studied. (2) Based on the LCM model and MULE model, to establish the rainfall-sediment model to distributed simulate the sediment yield of watershed outlet during rainfall event. The results indicate that the rainfall-sediment model has a stronger applicability in the Loess Plateau of China.The study improves the spatial discrete method of infiltration rate ‘r’, develops the LCM model from lumped model to distributed model, characterizes the spatial heterogeneity of rainfall and runoff, can help provide the runoff factor necessary for MUSLE. Using the rainstorm-runoff model LCM to simulate the runoff factor necessary for MUSLE, adding the overland and channel sediment concentration module, after doing these works, the coupling of LCM and MUSLE can not only reflect the process of runoff and sediment generated during rainfall event, but also overcome the previous shortcoming of MUSLE obtaining outlet sediment yield mostly through sum of the sediment yields generated in each grid, and makes it is possible to distributed simulate the sediment yield of watershed outlet during rainfall event. Nowadays, most watersheds of study area are covered with shrub and grass, appeared massively by various water conservation measures, and covered by steep and fragmentary topography. Aiming at this, the study selects and uses common local methods to improve the MUSLE. Results show that the combination of revising C, P, and LS simultaneously may be applicable in loess hilly and gully region based on the C factor suggested by Cai et al.(2000), the P factor suggested by Wang and Jiao(1996), and LS factor suggested by Liu et al.(1994，2000).(3) To estimate the effect of forest-grass vegetation change on sediment reduction. The results indicate that the effect of forest-grass vegetation change on sediment reduction is remarkable during rainfall event with the reduction ratio of 16.73% ~ 80.25%. The increase in forest-grass vegetation quantity is key factor affecting sediment reduction. The study uses scenario method, evaluates the sediment change during rainfall event in the scenario of same rainfall and different forest-grass vegetation conditions. Results show that sediment reduction ratio during rainfall event caused by forest-grass vegetation change at watershed scale is 16.73% ~ 80.25%.The study also calculates the correlation coefficient between average sediment reduction ratio and vegetation area proportion change index, vegetation coverage change index. Results show that there is obvious correlation between sediment reduction and forest-grass vegetation area increase, and no obvious correlation between sediment reduction and vegetation quality change during rainfall event. When the forest-grass vegetation area accounts for 85% of watershed, sediment coefficie of rainfall tends to be stable, and sediment yield is little.(4) To analyze the effect of rainfall and watershed characteristics on forest-grass vegetaion reducing sediment. The results indicate that rainfall duration (one of the rainfall characteristics) and watershed surface roughness (one of the watershed characteristics) are the major factors affecting the forest-grass vegetation reducing sediment. By analyzing the partial correlation of vegetation-sediment reduction elasticity to rainfall intensity, rainfall amount and rainfall duration, it is found out that the partial correlation coefficient of vegetation-sediment reduction elasticity to rainfall intensity, rainfall amount and rainfall duration are 0.536、-0.740、0.834,respectively. Results show that the longer the rainfall lasted, the more sensitive the forest-grass vegetation quantity on sediment change be, and then the more obvious sediment reduction generated by forest-grass vegetation increases. Using three indicators, include the watershed surface roughness, the watershed shape index, and the drainage density, to indicate topography condition, watershed shape and connectivity of watershed, and to study the effect of watershed features on vegetation reducing sediment. By analyzing the partial correlation of vegetation-sediment reduction elasticity to watershed surface roughness, shape coefficient and channel density, it is found out that the partial correlation coefficient of vegetation-sediment reduction elasticity to watershed surface roughness, shape coefficient and channel density are -0.862、-0.213、-0.476,respectively.The results show that: At the watershed scale, the effect of topography condition on vegetation reducing sediment is the most obvious. The lowest the watershed surface roughness is, the more sensitive the vegetation quantity on sediment reduction is, and then the more obvious sediment reduction generated by vegetation increases. (5) To analyze the regional differentiation rules of the effect of forest-grass vegetation change on sediment reduction. The results indicate that because of the major influence of watershed surface roughness, the sensitivity of rainfall-sediment responses on vegetation change decrease gradually from north to south in the middle reaches of the Yellow River.By analyzing the partial correlation of vegetation-sediment reduction elasticity to rainfall duration and watershed surface roughness, it is found out that the partial correlation coefficient of vegetation-sediment reduction elasticity to rainfall duration and watershed surface roughness are -0.764 and -0.962, respectively. Results show that the watershed surface roughness than the rainfall duration has a more obvious effect on vegetation reducing sediment, makes the sensitivity of rainfall-sediment responses on forest-grass vegetation change decrease gradually from north to south in the middle reaches of the Yellow River. In the Pinggang shape loess hilly and gully region located in the north of study area, the sediment yield responses on forest-grass vegetation change are the most sensitive, and the effect of vegetation increase on sediment reduction is most obvious. Secondary by the Mao shape loess hilly and gully region and Liang shape loess and gully region located in the middle of study area. In the droughty loess hilly and gully region located in the south of study area, the sediment yield responses on vegetation change are not sensitive, and the effect of vegetation increase on sediment reduction is not obvious.In short, the study analyses the change of vegetation quality and quantity in the middle reaches of the Yellow River, establishes the rainfall-sediment model to distributed simulate the sediment yield of watershed outlet during rainfall event based on the LCM model and MULE model, evaluates the sediment change during rainfall event in the scenario of same rainfall and different vegetation conditions，reveals the characteristics of sediment yield responses on vegetation change at watershed scale. It is can provide the basis for the analysis on causes of water and sediment change in the Yellow River basin.