以黄河中游多沙粗沙区代表性支流无定河为例，研究了降水量变化的背景下水土保持措施对无定河流域侵蚀产沙的影响。从总体上看，年输沙量和年径流量都表现出随时间而减小的趋势，梯田、造林、种草和淤地坝拦沙造地4项水土保持措施面积随时间增大的趋势极其显著。全流域面平均年降水量和汛期 (6~9月) 降水量则略有减少的趋势，但其减少趋势的显著性远低于水土保持措施随时间变化的显著性。产沙模数与梯田、坝地、造林和种草面积的关系，均表现出明显的负相关。对基准期 (1956~1969年) 和措施期 (1970~1996) 进行了比较，结果表明，实施水土保持措施以后，产沙模数、径流系数和汛期径流能够被降水所解释的百分比，分别由69%、80%和77%，下降为26%、31%和54%。说明在后一时期中，水土保持措施起了很大作用，因而降水变化对于产沙模数、径流系数、汛期径流量变化的贡献率大大减小了。以1956~1996年间41年的资料为基础，建立了全沙产沙模数与径流系数、汛期降雨量、最大一日雨量、最大30日累积雨量、梯田林草面积和坝地面积之间的多元回归方程，该方程表明，全沙产沙模数随汛期降雨量、最大一日雨量、最大30日累积雨量的增大而增大，随梯田林草面积和坝地面积的增大而减小， 随径流系数的增大而增大。为了区分人为因素与气候变化对侵蚀产沙变化的影响，采用对数据进行标准化处理之后再进行多元回归分析的方法，来确定相对贡献率。结果表明，降水变化和水土保持措施变化对无定河全沙产沙模数变化的贡献率大致相等，前者占50.3%，后者占49.7%。文中还建立了粗泥沙产沙模数与上述影响因子之间的多元回归方程。

Based on data from the Wudinghe River, a tributary of the Yellow River in the major coarse sediment source area of the middle Yellow River Basin, a study has been made of the effect of soil and water conservation measures on sediment yield in the condition of changing precipitation. Both annual sediment load and annual river flow show a decreasing trend with time, and the areas of soil and water conservation measures. The area-weighted annual and high-flow season precipitations over the drainage basin also show a mild trend of decreasing, less significant than the above variables. It has been found that specific sediment yield is negatively correlated with the areas of land terracing, tree planting and grass planting. The period between 1956 and 1969 and the period between 1970 and 1996 are taken as "baseline" period without measures and the compared period with measures, respectively. Comparison between these two periods indicates that, during the former period, the percentages of the variation in specific sediment yield, runoff to precipitation ratio and high-flow season water yield can be explained by the variation in precipitation were 69%, 80% and 77%, respectively. However, in the latter period, these percentages decreased to 26%, 31% and 54%, respectively. This change means that, in the latter period, soil and water conservation measures played a great role, and thus the contribution of the variation in precipitation to the variation in sediment yield was greatly reduced. Based on the 41-year data from 1956 to 1996, a multiple regression equation has been established.