中文摘要：

为明确黄河中游地区水土流失综合治理、退耕还林（草）政策的水沙效应问题，选择黄河中游风沙-丘陵过渡区、典型丘陵区以及土石山林区5条典型流域，用降雨及面积对水文站实测径流量、输沙量数据进行处理，探讨了近50 a来流域水沙及其关系的演变过程。结果表明：5条流域标准化年径流量、标准化年输沙量均表现出显著减少趋势，且两者发生跃变时间基本一致。时段频率上的河流输沙量减少程度较径流量大，零输沙天数逐时段增加。植被覆盖度低的风沙-黄土丘陵过渡区流域，较植被覆盖度高的土石山林区流域，径流量和输沙量年均变化斜率大，跃变时间早。5条流域汛期径流量和输沙量的关系表明，植被覆盖度低的风沙-黄土丘陵过渡区流域的输沙量分布域较植被覆盖度高的土石山林区流域大一个数量级，其产沙动力也最强。近50 a来，5条流域的水沙动力系数均呈现逐时段降低趋势。20世纪70～80年代水土流失综合治理使5条流域水沙动力系数平均降低24.5%，1999年后的退耕还林（草）工程则使其平均降低70.6%。流域水沙拟合线性关系常数项绝对值，以及流域冲淤平衡时的径流量呈现随时段减小趋势，进一步说明水土流失治理及退耕还林（草）措施对流域水沙动力关系的削弱效应。

英文摘要：

To reduce the severe soil erosion on the Loess Plateau, a great number of soil and water conservation measures was implemented in 1970s-1980s, and after that, Grain for Green project was applied since 1999. The great anthropogenic forces resulted in the decreasing trend of stream flow and sediment load on the Loess Plateau. However, the relationship between streamflow and sediment load wasn＇t clear. Understanding the change of streamflow, sediment load, and their relation was helpful to enable sustainable catchment management in the region. The monthly streamflow and sediment load data were used in the study and summed annually to check the trends in the past five decades. To compare the impacts between catchments, the data were standardized with precipitation at a corresponding temporal scale and catchment area. The results showed that all the 5 catchments in the study had a statistically significant negative trend in annual streamflow and sediment load given by the M-K trend test. Change points given by the Pettitt test in streamflow of the catchments occurred were basically consistent with that of sediment load. The extent of reduction in the sediment load was much greater than that of streamflow with the periods altered, and an increasing number of zero-sediment load days was observed. The average annual reduction both in streamflow and sediment load was much greater in the low-vegetation-coverage catchments than that in the high-vegetation-coverage catchments, and the change point was also earlier. The scattered distribution for monthly streamflow and sediment load for the wet season indicated that the low-vegetation-coverage catchments have wider domain by one order of magnitude in sediment load than that of the high-vegetation-coverage catchments. The sediment load was simply regressed with the streamflow, and the regression coefficient was considered as the sediment production coefficient here. The sediment production coefficient of the 5 catchments was observed to decrease by periods during the past 50