Abstract: The evolution law and cause analysis of runoff in a climate change, principally the quantitative analysis of the impact of precipitation and soil and water conservation measures on runoff, are still hot issues in the current hydrology and water resources research. An improved understanding of the effect of rainfall and soil and water conserving practices on runoff is obligatory to manage runoff under a changing climate. The current study aims specifically to the evolution of precipitation, measures, and runoff under different periods and soil and water conservation measures. Here, the average precipitation using the Tyson polygon method was determined. The periodicity of precipitation and runoff was analyzed using the Morlet wavelet analysis. Moreover, in order to analyze the relationship between different periods of precipitation, a runoff regression analysis was performed in a controlled area. Through the construction of multivariate function, precipitation was determined, intensity was measured, and runoff evolution was predicted in Anding District, Dingxi City, Gansu Province, China in the past 60 a. The results depicted that (1) the annual runoff exhibits a decreasing trend from 1957 to 2016. The runoff decline reached a very significant level (P<0.001) except with the annual precipitation. On the time scale of 22- 24 a, 8 a, and 4 a, the annual precipitation and runoff have obvious oscillation periods and the average period is about 15 a, 5 a, and 3 a. In the first 10 a, the runoff responded strongly to the change of precipitation and the trend was highly synergistic, indicating that this period had a better correlation. In the last 50 a, the response of the runoff to the change of precipitation is attenuated rapidly, and the trend is synergistic alienation, illustrating that the runoff change of the study area after 1986 is greatly affected by other factors, especially after 2000. (2) With the advancement of soil and water conservation work, the intensity of soil and water conservation measures such as terraces, artificial afforestation, artificial grass planting, and enclosure gradually increase to 36.14 hm² ?km^{- 2}, 25.26  hm² ?km^{- 2}, 11.56  hm² ?km^{- 2}, and 3.22 hm² ?km^{- 2}, respectively. With the implementation of soil and water conservation measures, such as conversion of cropland to forest and grassland and terracing of sloping fields, the non-controlled area is gradually transformed into the controlled area. (3) With the increase of the period (3 a, 5 a, 15 a), the effect of the same precipitation on the runoff generation increases, and the impact of the same intensity of measures on the runoff generation decreases. The combination of precipitation and measured intensity explained 57.46%- 85.80% of the total variance of runoff modulus. The influence of precipitation on runoff modulus was about less than 40% and the impact of measure intensity was about 60% higher. This shows that the impact of the intensity of measures on runoff exceeds the driving force of the precipitation factor. Under the condition of no-significant changes in precipitation in the past 60 a from 1957 to 2016, the continuous decrease in runoff is due to the increase in the intensity of soil and water conservation measures. This study provides support for scientific and rational adjustment of land use structure, deployment of soil and water conservation measures, and construction of ecological environment.

Key words: evolution law, precipitation, soil and water conservation measures intensity, runoff, semi-arid area