Reservoirs regulate the relationship between hydrological drought recovery water and drought characteristics

Abstract

Regulation of streamflow by large reservoirs alters the characteristics of hydrological drought. However, few studies have focused on the impacts of regulation on the water required (WR) for hydrological drought recovery, nor on the response relationship between WR and drought characteristics (i.e., duration and severity). Therefore, we proposed a comparative approach that integrates reconstructing unregulated streamflow and a drought identification method using variable threshold levels (VTLs) to evaluate the impacts. Long-term (>40 years) observational datasets for streamflow at a hydrometric station and inflow and outflow at multiple large reservoirs in the Dongjiang River Basin were used. Using VTLs that accounted for seasonal differences in hydrological processes, our method performed well in identifying historical hydrological droughts and their characteristics in the study area. The WR mainly depended on drought severity rather than drought duration and there was a clear nonlinear response relationship (i.e., power function) between WR and drought severity. The performance indices (i.e., R2 = 0.92, NSE = 0.97, PBIAS = 13.57%) indicated that the optimal nonlinear function model could accurately simulate the WR of hydrological drought events. Large reservoirs reduce the frequency of hydrological droughts, shorten drought duration, and reduce drought severity by storing water in the flood season and releasing it in the dry season. The presence of reservoirs shifted the relationship between WR and drought severity from linear to nonlinear. These findings showed that the proposed methodology can help us to optimize the management of water resources for drought prevention and disaster reduction under changing environments.