- 1Division of Environment and Sustainability, The Hong Kong University of Science and Technology, Hong Kong, China
- 2University of Science and Technology of Hanoi, Vietnam Academy of Science and Technology, Hanoi, Vietnam
- 3Department of Civil and Environmental Engineering, The Hong Kong University of Science and Technology, Hong Kong, China
- 4Center for Climate/Environment Change Prediction Research, Ewha Womans University, Seoul, Republic of Korea
- 5Severe Storm Research Center, Ewha Womans University, Seoul, Republic of Korea
- 6Department of Climate and Energy System Engineering, Ewha Womans University, Seoul, Republic of Korea
Extreme precipitation presents a significant environmental challenge that threatens the economic and social stability of Southeast Asian countries, highlighting the critical need for reliable model simulations for early warnings and impact mitigations. Recently, the fifth version of the regional climate model (RegCM5) has been released, featuring updates in multiple model components including the physical parameterizations, which is expected to advance the simulation capability for extreme precipitation events. However, optimizing the model parameterization remains challenging due to the vast array of parameters that require fine-tuning. Traditional approaches that use random-based sensitivity tests to identify optimal scheme combinations are constrained by computing power and often fail to explore the complete range of possible combinations needed for accurate regional climate representation. Moreover, parameters within each scheme exist on a continuous spectrum rather than as discrete options, exponentially increasing model optimization's complexity and computational demands.
To overcome these limitations, advanced optimization techniques have emerged to efficiently explore the complete range of possible combinations, without relying solely on random-based sensitivity tests. In this study, we employ a micro-genetic algorithm (micro-GA) for combinatorial optimization of key parameters within the cumulus convection schemes in RegCM5. The model, driven by ECMWF Reanalysis version 5 (ERA5), covers most of Southeast Asia at a 0.22-degree resolution. This study aims to:
(1) validate the capability and efficiency of the coupled RegCM5-micro-GA interface in improving the simulation of extreme precipitation events in Southeast Asia
(2) investigate the sensitivity of the RegCM5-micro-GA algorithm to different fitness functions and different physical parameters
(3) reveal the mechanism of model optimization by examining physical processes improved by the tuned cumulus convection parameters.
The findings will provide valuable insights to facilitate the wider use of RegCM5 and benefit the broad community in model optimization, fostering more accurate and timely predictions of extreme weather events.
[Acknowledgements]
This research was supported by project GRF16308722, which was funded by the Research Grants Council (RGC) of Hong Kong. This study was also supported by the “Research Program for Agricultural Science & Technology Development (Project No. RS-2024-00399847)”, National Institute of Agricultural Sciences, Rural Development Administration, Republic of Korea.
How to cite: Zhou, Z., Nguyen-Xuan, T., Im, E.-S., Yoon, J. W., and Park, S. K.: Combinatorial Optimization of Cumulus Convection Scheme Parameters in RegCM5 Using a Micro-Genetic Algorithm for Extreme Precipitation Event Simulations in Southeast Asia, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-4222, https://doi.org/10.5194/egusphere-egu25-4222, 2025.
