Research papersCopula-based drought risk assessment combined with an integrated index in the Wei River Basin, China

- A reliable multivariate integrated drought index (MIDI) was constructed.
- A drought event was redefined with the MIDI using run theory.
- The drought risk with multi-dimensional and multivariate characteristics was fully assessed.
- The subzones thatwhere had high-risk occurrence of heavy-droughts were found.

It is critical to assess drought risk based on a reliably integrated drought index incorporating comprehensive information of meteorology, hydrology and agriculture drought indices, which is of great value for further understanding the future drought tendency, prevention and mitigation. Thus, the primary objective of this study was to focus on constructing a multivariate integrated drought index (MIDI) by coupling four drought indices (i.e., Precipitation Anomaly Percentage (PAP), Runoff Anomaly Percentage (RAP), Standardized Precipitation Index with 6-month aggregation time step (SPI6) and Modified Palmer Drought Severity Index (MPDSI)) to objectively and comprehensively investigate drought risk. The variable fuzzy set theory and entropy weight method are used during the MIDI construction process. Based on the MIDI, a drought event including drought duration and severity is redefined using run theory. Then copula-based drought risk is fully assessed through the joint probability distribution of drought duration and severity. Results indicate the following: (1) the constructed MIDI is consistent with the Standardized Precipitation Index (SPI) and Runoff Anomaly Percentage (RAP) series, and it is more sensitive and effective to capture historical drought events; (2) the drought characteristics present noticeable spatial variability among five subzones, and the entire basin has 49 droughts with the longest drought duration spanning 8.55 months; and (3) the mainstream, especially the middle and lower reaches, has higher occurrences of severe droughts for approximately every 10 years.