Spatiotemporal variability of drought/flood and its teleconnection with large-scale climate indices based on standard precipitation index: a case study of Taihu Basin, China

With the intensification of global warming, frequency of floods and droughts has been increasing. Understanding their long-term characteristics and possible relationship with large-scale meteorological factors is essential. In this study, we apply signal denoising, dimensionality reduction technique, and wavelet transform to study the spatiotemporal distribution pattern of drought/flood and its teleconnection with large-scale climate indices. Based on the precipitation data of 63 hydrological stations in the Taihu Lake Basin (TLB) for 54 years from 1965 to 2018, the standard precipitation index (SPI) was used as an indicator. The ensemble empirical mode decomposition (EEMD) and empirical orthogonal function (EOF) methods were used to explore the spatiotemporal evolution characteristics of droughts and floods. In addition, the cross-wavelet transform (XWT) method was used for teleconnection analysis. The results indicated that during 1965-2018, the SPI of the TLB showed quasiperiodic oscillations dominated by interannual oscillations (52.5%). Except for the trend of drought in spring, the basin showed a wetter trend at annual, summer, autumn, and winter scales. There were two main spatial modes (total 78.48% contribution) in the TLB, consistent across the region and reverse distributed from south to north. The dry areas were mainly in southern Zhexi and the northern Huxi sub-regions; the Hangjiahu and Yangchengdianmao sub-regions were prone to flooding. In addition, SPI was correlated with various large-scale meteorological factors, but the strength of the correlation had specific temporal and spatial heterogeneity. The research results can provide TLB reference values for water resource management and flood/drought disaster control.

Automated summary

With the intensification of global warming, the frequency of floods and droughts is increasing. This study uses signal denoising, dimensionality reduction techniques, and wavelet transform to analyze the spatiotemporal distribution patterns of droughts/floods and their connection to large-scale climate indices. The research results provide valuable references for water resource management and flood/drought disaster control.