Global climate change has accelerated regional water circulation and caused asymmetry of precipitation distribution and high flood frequencies. In arid and semiarid regions, a small climate fluctuation may cause large environmental variation when human activity overwhelms the natural carrying capacity. For this reason, addressing the impact of regional climate change on runoff volume will support scientific and technological sustainable development of local water resources.

Tarim River Basin in Xinjiang is a typical inland watershed in the inland area of Northwest China, where abundant natural resources coexist with an ecologically fragile environment. Because of relatively scarce precipitation in this area, the main water resource is runoff from a mountainous drainage basin. It is very important to ascertain variations of regular hydrologic and meteorological time series data. However, there are few reports of periodicity and abrupt change of drought–flood variation, and multiple time-scale correlation between drought–flood variation and climate change in the river headstreams.

Basing on the monthly precipitation and hydrologic data in the three headstream mountain areas of the Tarim River over the past 50 years, BAI Yuan et al. analyzed the trends of annual runoff and drought–flood indices by a non-parametric test, the periodic significance of runoff and drought–flood variation through wavelet analysis, and the multiple time-scale correlation between runoff and drought–flood indices via a crosswavelet spectrum.

The results showed that the drought–flood indices showed increasing trends for the Aksu and Yarkand rivers, and rose non-significantly for Hotan River. The indices of the three headstreams changed remarkably in 1986. The curves of wavelet variance showed that significant periods of the indices are 4 and 8 years for Aksu and Hotan rivers, and 8 and 10 years for Yarkand River. Runoff of the Aksu and Hotan rivers had significant periods of 6 and 8 years, plus 3 and 9 years for Hotan River. There was significant correlation between the drought–flood indices and annual runoff volume in the three headstreams.

The results provided important information toward achieving predictability of flood and drought in Northwest China. The study was published in Environmental Earth Sciences in February 2014.