Abstract: Intensification in climate change and frequent human activities makes it unclear whether or not the precipitation-temperature dependent structure in the Xijiang River basin has changed and what are the driving forces behind it. Using the maximum likelihood ratio (CLR) method of Copula functions, this paper presents a diagnosis of the structure changes and explores the driving forces behind the change through considering the factors of local meteorology and remote correlation. The results show that a sudden change in this structure occurred in the upper Xijiang basin in February 2002 and also in the Yujiang basin in April 1985. Wind speed has significant positive correlation with the dependent structure of the whole time series, but relative humidity is correlated negatively. We found that arctic oscillations are the main driving force that impacts the East Asian monsoon and then on the local wind speed, and variations in the latter can modify the precipitation-temperature relationship, thereby causing significant variations in the dependent structure.

Key words: Xijiang River basin, precipitation, temperature, Copula, variability diagnosis, cross-wavelet analysis, remote correlation factor