Downscaling & super-resolution
Uncertainty in Rainfall Intensity Duration Frequency Curves of Peninsular Malaysia under Changing Climate Scenarios
In one sentence. Combines MOS downscaling of GCM rainfall with ANN temporal disaggregation to project IDF curves for Peninsular Malaysia under RCP4.5 and RCP8.5 with uncertainty bounds.
Abstract
This study developed a methodological framework to update the rainfall intensity-duration-frequency (IDF) curves under climate change scenarios. A model output statistics (MOS) method is used to downscale the daily rainfall of general circulation models (GCMs), and an artificial neural network (ANN) is employed for the disaggregation of projected daily rainfall to hourly maximum rainfall, which is then used for the development of IDF curves. Finally, the 1st quartiles, medians, and 3rd quartiles of projected rainfall intensities are estimated for developing IDF curves with uncertainty level. Eight GCM simulations under two radiative concentration pathways (RCP) scenarios, namely, RCP 4.5 and RCP 8.5, are used in the proposed framework for the projection of IDF curves with related uncertainties for peninsular Malaysia. The projection of rainfall revealed an increase in the annual average rainfall throughout the present century. The comparison of the projected IDF curves for the period 2006–2099 with that obtained using GCM hindcasts for the based period (1971–2005) revealed an increase in rainfall intensity for shorter durations and a decrease for longer durations. The uncertainty in rainfall intensity for different return periods for shorter duration is found to be 2 to 6 times more compared to longer duration rainfall, which indicates that a large increase in rainfall intensity for short durations projected by GCMs is highly uncertain for peninsular Malaysia. The IDF curves developed in this study can be used for the planning of climate resilient urban water storm water management infrastructure in Peninsular Malaysia.
Labels
- Topic
- Downscaling & super-resolution, Extremes & hazards
- Type
- Application study
- Method
- Classical ML (trees, GP, linear, kernels)
- Time scale
- Climate projections / centennial
- Domain
- Regional
- Data
- CMIP / climate model output
- Status
- Research
- Identifiers
- DOI:10.3390/w10121750 OpenAlex:W2902731828
- Added
- 2026-10-22
Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.