Application of Bias Correction to CORDEX Southeast Asia Multi-Model Simulations over Southeast Asia / Dr Sheau Tieh Ngai [魏曉蝶 博士] (國家災防科技研究中心)

Abstract: Past studies have shown that most global climate models (GCM) underestimate the sensitivity of extreme precipitation variability and trends, particularly in tropical regions, leading to uncertainties in future projections. Dynamical downscaling with regional climate model (RCM) enhances resolution and provides valuable climate information for regional and local impact assessments. However, RCM outputs remain prone to biases, similar to GCMs, limiting their direct applicability in climate impact studies. To address these limitations, bias correction or adjustment is routinely applied to climate model outputs, ensuring greater reliability when used as inputs for impact assessment models. 

For the past few decades, the activities of regional climate downscaling have been coordinated under the Coordinated Regional Climate Downscaling Experiment (CORDEX), a program under the World Climate Research Programme. CORDEX facilitates the coordination of regional climate modeling efforts worldwide across different regional domains (Giorgi et al., 2009; www.cordex.org). It includes 14 simulation domains, with CORDEX Southeast Asia (CORDEX-SEA) being the most recently established. This study applies the quantile mapping (QM) bias correction method to adjust biases in regional climate model (RCM) simulations from CORDEX-SEA. Multi RCM simulations were bias-corrected to assess future changes in mean precipitation and extreme indices across the region. Validation against observations shows that QM improves the spatial representation of seasonal mean rainfall, enhancing correlation coefficients, standard deviation, and RMSD while reducing inter-model variability. The bias correction method was found to modify the climate change signal and amplifying the range of projected changes while addressing systematic biases. However, its impact varies across RCMs and extreme indices. These findings provide valuable insights for climate-hydrological impact studies in humid tropical regions, particularly Southeast Asia.