Long-term trends in Songhua River Basin streamflow and its multivariate relationships with meteorological factors.

Long-term streamflow trends are closely related to meteorological factors; understanding the relationships between them helps to improve water resources management in advance. In this study, we examined long-term annual and seasonal streamflow trends over 1961-2010 in 28 stations in the Songhua River Basin (SRB), China, using four kinds of trend detection methods and then determined the optimal meteorological predictors for SRB streamflow based on the multiple wavelet coherence. We found significant downward trends in annual streamflow in a large part of the study stations (varies from 10 to 18 for different methods), and fewer decreasing stations were detected when we consider the full autocorrelation and the long-term persistence in streamflow. In contrast to annual streamflow, fewer stations showed significant downward trends in summer and winter streamflow. Streamflow generally followed the pattern of precipitation (PRE); the largest streamflow changes occurred in summer and August monthly streamflow variation contributed the most to the annual streamflow variation. We found PRE and potential evapotranspiration (PET) combined was the optimal predictor for streamflow above Jiangqiao and on the Jiangqiao-Dalai section of the Songhua River; as for the Dalai-Harbin section and the Harbin-Jiamusi section, the optimal predictor combinations are PRE and number of rainy days (WET), and PRE and average monthly temperature (TMP) respectively.