Staged optimization of urban drainage systems considering climate change and hydrological model uncertainty

Abstract

Flood risk increases with climate change and rapid urbanization, which urgently needs to improve the capacity of urban drainage systems. In the study, a staged optimization model considering climate change and hydrological model uncertainty (SOCU) was proposed for urban drainage system design. The SOCU model refers to the construction of drainage system by stages instead of traditional implement-once plans, which was established based on an integration of staged optimization policy, urban hydrological model, generalized likelihood uncertainty estimation (GLUE) method and chance-constrained programming (CCP) model. The staged optimization policy was employed to deal with the uncertainty of climate change. The GLUE method was adopted to analyze the uncertainty of urban hydrological model established by PCSWMM. Considering the uncertainty from urban hydrological simulation, the CCP model (one of the main methods of stochastic mathematical programming) was used to deal with the uncertainty of the optimization model. Subsequently, a case study of the Haidian Island of Hainan Province in China was used to demonstrate the proposed model. The result shows that the optimal investment is 183 million Yuan with a pumping capacity of 28.3 m3/s for the first stage and 38.4 m3/s for the next stage. The SOCU model is not only more flexible to adapt to climate change, but also is economically efficient (10% lower than implement-once plans). The urban drainage system design obtained from the SOCU model is more reliable and robust than traditional implement-once plans since the climate change and hydrological model uncertainty are simultaneously taken into account. Furthermore, the investment increases from 166 million Yuan to 195 million Yuan when the flood constraint satisfaction probability increases from 0.75 to 0.95. Therefore, the model could provide richer decision-making information than traditional implement-once plans and help decision makers seek a trade-off between system investment and acceptable flood damage. The study outcomes provide a reliable optimization model for urban drainage design and may have profound implications and contributions for urban flood management.