A framework for projecting future intensity-duration-frequency (IDF) curves based on CORDEX Southeast Asia multi-model simulations: An application for two cities in Southern Vietnam

Abstract

To date, no previous studies based on the Coordinated Regional Climate Downscaling Experiment Southeast Asia (CORDEX-SEA), which represents the most comprehensive set of regional climate model simulations in Southeast Asia, have reported the climate change impacts on Intensity-duration-frequency (IDF) curves for designing urban drainage systems. In this study, a framework to project future IDF curves based on a temporal disaggregation of bias-corrected CORDEX-SEA simulations was developed and applied to two cities in Southern Vietnam, i.e. Ho Chi Minh City (HCMC) and Can Tho City (CTC). First, a new bias correction (BC) method, which we call normalized quantile mapping (NQM), was proposed and compared with two other BC methods, quantile mapping (QM) and quantile delta mapping (QDM), to select a suitable BC method for downscaling each dataset of CORDEX-SEA multi-model simulations. A temporal disaggregation model in an artificial neural network (ANN) framework was introduced to obtain future sub-daily rainfall extremes, and the results were compared with the widely used stochastic model, Hyetosminute. Selecting a suitable BC method for each dataset performs better than using one BC method for all datasets in decreasing the ensemble spread by 47.1% in HCMC and 61.4% in CTC. The temporal disaggregation by ANN performs much better than Hyetosminute in durations shorter than or equal to six hours. IDF curves (based on ensemble mean of seven climate models) projected for the far future period (2066–2085) reveal that the rainfall intensity for 25-yr return period is expected to increase by 27.4–32.6% in HCMC and 55.4–72.8% in CTC (depending on the rainfall duration) from those observed for the historical period (1986–2005), while expected to change by −1.1–11.3% for 2-yr return period.Generally, more increase of rainfall intensity for rarer rainfall events (especially for longer durations) urges us to update the existing drainage systems for higher return periods. The framework developed in this study can serve as an important reference for studying climate impacts on the urban drainage systems in other cities located in the same region.