中文摘要：

使用挠力河流域水文站1956年以来的径流资料，应用多种数值分析方法，系统分析了挠力河径流的季节及年际变化规律，对人类活动导致的径流变化进行了估算。结果表明，挠力河上游径流年内分配不均匀系数Cvy（0．77～1．97）大于下游径流（0．62～1．32），然而下游径流c竹有显著上升趋势，表明下游地区河流径流情势发生显著改变。径流的年际变化剧烈，丰水时段较短，通常只有1年～4年，枯水时段较长，最长可达9年。应用累积滤波器和Kendall秩次相关分析方法，对年径流变化趋势进行了定性和定量分析，结果表明，挠力河流域年径流量呈减少趋势，上下游径流Kendall秩次相关系数绝对值均大于1．96，减少趋势非常显著。年降水及径流双累积曲线及回归分析结果表明，宝清站以上上游地区径流自1976年后开始减少，菜嘴子站中下游地区径流自1964年后开始减少；宝清站上游地区和菜嘴子站中下游地区人类活动导致的径流的减少量分别占径流减少总量的71．19％和62．21％，人类活动是导致挠力河径流减少的主要因素，其次才是降水变化。

英文摘要：

Using runoff data of 1956 - 2001 from the two hydrometrie stations in the Naolihe River watershed, the seasonal and annual change in the runoff regime was analyzed, and the influence of human activities on the runoff was estimated. It was found that the seasonal distribution of runoff was quite uneven. The runoff from April to November accounted for 94.74% to 99.24% of the total. As to the long-term distribution in the runoff, the high-flow period was short （only 1 - 4 years）, while the low-flow period was long （9 years ）. Comparison of uneven coefficients of annual distribution for the two hydrometric stations showed that the Cvy for the upstream runoff was 0.77 to 1.97, which was higher than that for the middle-stream runoff （0.62- 1.32）, while there was obvious increase in the Cvy for the middle-stream runoff, which indicated that the alteration in the runoff regime in the middle-stream area was more remarkable. Result of statistical analysis of stochastic hydrology showed that there was obvious decline in the runoff for the whole river. Analyses of double mass curve of annual runoff and precipitation showed that the decline in the runoff occurred in 1975 for the upstream runoff, and 1964 for the middle-stream runoff. The decline in runoff resulted from human activities such as reclamation of wetland, agricultural water use and so on. To assess the influence of human activities on the decline in the runoff, the correlations between precipitation and runoff for period without influence of human activities were established and used to estimate the natural runoff for period affected by human activities. The simulated results, compared with the monitored runoff data, demonstrated that decreases in runoff from human activities accounted for 71.19% of the total for the upstream, and 62.21% of total decrease in runoff for the middle-stream runoff. This showed that human activities, not annual precipitation, were the main driving forces that resulting in the decline of runoff in the Naolihe River. Affecte