Path Analysis on Impacts of Meterological Factors on Runoff from Tianshan Mountains:A Case Study on Manas River and Kaidu River Watersheds
天山南北坡气象因子对出山口径流影响通径分析—以开都河和玛纳斯河流域为例

The correlation coefficients between independent variables and dependent variables can reflect the impact of the former on the latter. However, the direct and indirect impacts of independent variables cannot be reflected by the correlation coefficients because multiple paths from the independent variables to the dependent variables might exist. The path analysis derived from correlation analysis provides an approach to evaluate both direct and indirect impacts. Based on the meteorological and hydrological data of Manas River and Kaidu River watersheds since late 1950, this paper has analyzed the variation trends of accumulated temperature (AT), accumulated precipitation (AP) and accumulated runoff depth (ARD, on both annual and seasonal scales). Both direct and indirect impacts of daily accumulated temperature (DAT) and precipitation (DAP) on daily runoff depth (DRD) were also investigated by using the path analysis. The results show that there are significant increases in annual accumulated temperature (AAT) and annual accumulated runoff depth (AARD) in both watersheds, but no significant changes in annual accumulated precipitation (AAP). In the Kaidu River watershed, the seasonal accumulated temperature (SAT) increases markedly in all seasons except winter, but in the Manas River watershed it increases markedly only in autumn. The seasonal accumulated precipitation (SAP) in the Kaidu River watershed has a tendency to increase remarkably only in winter. The seasonal accumulated runoff depth (SARD) in the Kaidu River watershed in spring is much larger than that in the Manas River watershed. In the summer and winter, the Manas River watershed shows significant increase in SARD while the Kaidu River watershed only does so in winter. Correlation and path analyses show that DAT and DAP have remarkable direct impacts on DRD in all seasons except winter. The direct effect of DAT on DRD in the Manas River watershed is significantly larger than that of DAP on DRD, but it is reverse in the Kaidu River watershed. DAT is more important than precipitation as the driving factor for daily runoff generation bacause of the critical impact of temperature on snow/glacier-melting. Other factors may include topography, vegetation cover, soil and human activities.