Differences of Precipitation Characteristics among GCMs over Southeast Asia under AR6 Climate Change Scenarios

Southeast Asia is known globally as a highly vulnerable climate change region. Precipitation is the primary factor that impacts livelihood in this region due to recurring flood and drought incidents. Variables projections under climate change can be made using General Circulation Models (GCMs). An investigation of projected precipitation with the new phase of the model experiment, the Sixth Assessment Report (AR6), is worth to be considered. This study investigates the spatial distributions of variability, trend and conditions (wet/dry) of precipitations generated using 10 GCMs over the SEA under AR6 with four scenarios (SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8). Three statistical methods, coefficient of variation (CV), Mann-Kendall test (MK) and standardized anomaly index (SAI), were calculated by grid cells. Significant differences among GCMs could be seen in the results. High precipitation variation with CV was indicated around the southern part of Indonesia and the Philippines oceans for six models, whereas only one model (MRI-ESM2) returned strong variation for mainland countries. A decreasing precipitation trend during the historical period could be observed in mainland countries with four GCMs. However, the SSP3-7.0 and SSP5-8.5 of most models presented precipitation increment. The extremely wet and dry ratio to all other years was calculated. Highly wet years higher than 10% were indicated in SSP5-8.5 with MPI-ESM1 occurring in most areas of the region, whereas other models gave 6-10% of highly wet occurrence. Drought situation occurred higher than 10% and could be seen with only three models with small areas under all scenarios.  returned strong variation for mainland countries. A decreasing precipitation trend during the historical period could be observed in mainland countries with four GCMs. However, the SSP3-7.0 and SSP5-8.5 of most models presented precipitation increment. The extremely wet and dry ratio to all other years was calculated. Highly wet years higher than 10% were indicated in SSP5-8.5 withMPI-ESM1 occurring in most areas of the region, whereas other models gave 6-10% of highly wet occurrence. Drought situation occurred higher than 10% and could be seen with only three models with small areas under all scenarios.