This paper presents an assessment of the changes in future floods. The ranked area-average heavy daily rainfall amounts simulated by a super-high-resolution (20 km mesh) global climate model output are corrected with consideration of the effects of the topography on heavy rainfall patterns and used as a basis to model design storm hyetographs. The rainfall data are then used as the input for a nearly calibration-free parameter rainfall-runoff model to simulate floods in the future climate (2075-2099) at the Upper Thu Bon River basin in Central Vietnam. The results show that although the future mean annual rainfall will not be considerably different compared to the present-day climate (1979-2003), extreme rainfall is projected to increase vigorously, leading to a similar order of intensification of future floods. It is very likely that the flood peak with a 25-year recurrence will increase approximately 42% relative to the present-day climate. The occurrence of floods with a 10-year recurrence may exceed those with a 25-year recurrence in the present-day climate. The projection results also exhibit insignificant uncertainties caused by an artificial neural network-based bias correction model. Additionally, the presented bias correction model shows advantages over a simple climatology scaling method.
ASJC Scopus subject areas
- Global and Planetary Change
- Water Science and Technology
- Atmospheric Science
- Management, Monitoring, Policy and Law