Trivariate copula functions for constructing a comprehensive atmosphere-land surface-hydrology drought index: A case study in the Yellow River basin

Abstract

The current univariate and multivariate drought indices cannot fully reflect the multiple processes of meteorological-hydrological-agricultural drought. So it is important to construct a comprehensive drought index for drought monitoring in a changing environment from the perspective of the whole process of the natural water cycle “atmosphere-land surface-hydrology”. In this study, based on the standardized indices (SCI) framework and trivariate Copula functions, a multi-scale comprehensive drought index (SPERSI) integrating rainfall, temperature, runoff and soil moisture is proposed, which can synthetically reflect meteorological, hydrological and agricultural droughts. The spatial and temporal variations of drought in the Yellow River Basin (YRB) from 1961 to 2019 were investigated, and then the links between SPERSI and some remotely correlated factors were revealed by using the multiple cross wavelet transform technique. Results show that: (1) the SPERSI constructed by fusing rainfall-temperature-runoff-soil moisture information can sensitively and effectively capture the onset, persistence and termination of drought; (2) the SPERSI exhibits good performance in detecting integrated drought events in the YRB with probability of detection (POD) > 0.75, false alarm rate (FAR) < 0.04 and critical success index (CSI) > 0.73; (3) From 1961 to 2019, drought in different time scales of above Longyangxia (AL) and Longyangxia to Lanzhou (LL) zones in the YRB both show a slight mitigation trend, while the drought in remaining subzones show a trend of intensification; (4) Multiple cross-wavelet results indicate that AO is the dominant factor of drought in the Yellow River Basin, followed by ENSO. In general, SPERSI results are reliable and can provide basis for drought decision-making.