Abstract: Rainfall-runoff relationship is crucial in characterizing a watershed's capability to generate runoff. Links between rainfall and runoff have changed inconsistently as a result of the coupling effects of climate change, underlying surface change, and human activities; and the effects of these elements need to be assessed. This paper presents a method for identifying and estimating the driving factors of the rainfall-runoff inconsistency relationship, and quantifies the contribution of meteorological factors, underlying surface conditions, and human activities. This method integrates detrended cross-correlation analysis and detrended partial cross-correlation analysis. The findings demonstrate it has strong application to inconsistent series. In the Wei River basin, human activities, with an average contribution rate of 58.86%, have the greatest influence on the rainfall-runoff relationship, followed by the meteorological factors, while the underlying surface conditions have the least impact. The impact of human activities on the rainfall-runoff relationship in this basin generally takes a continuous growing trend in the time domain, while spatially its degree can be listed in a decreasing order: the upper and middle reaches of the mainstream of the Wei River basin > the Jing River basin > the Beiluo River basin.

Key words: cause analysis, rainfall-runoff relationship, inconsistency, contribution rate, Wei River basin