Item

Abstract

Undesirable disturbances (e.g., extraordinary drought) are prone to cause cascading failures of water supply-hydropower generation-environment conversation (WHE) nexus system. As the response of the WHE nexus system to the cascading failures is a time-varying process and is defined as a dynamic resilience, simulating the dynamic resilience process can provide opportunities for pre-disturbance effective mitigation/prevention strategy and post-disturbance optimal resource allocation. The previous dynamic resilience simulating is often based on the linear assumption which is not applicable for the WHE nexus system. To improve simulation accuracy of the dynamic resilience process of the WHE nexus system, a novel nonlinear dynamic resilience model that incorporates both a WHE nexus model and a Second-order Volterra (SOV) model is proposed. The WHE nexus model simulates amount of impact propagated by a disturbance on each subsystem, while the SOV model captures the nonlinear relationship between the amount of impact and a corresponding dynamic resilience process. The proposed model was applied to the Liuchong River Basin, China. The results show that the dynamic resilience process simulated through the SOV model performs better than that through the model with linear assumption, with Nash-Sutcliffe efficiencies of water supply, hydropower generation, and environment conservation subsystems increasing by 36.8%, 70.9%, and 17% respectively. Moreover, the SOV model are more suitable for simulating and predicting dynamic resilience when a severer water resources drought occurs. The proposed model helps to guide resilience management of water resource system by combining nexus and resilience thinking.