Projection of spatial and temporal changes of rainfall in Sarawak of Borneo Island using statistical downscaling of CMIP5 models

Abstract

This study assesses the possible changes in rainfall patterns of Sarawak in Borneo Island due to climate change through statistical downscaling of General Circulation Models (GCM) projections. Available in-situ observed rainfall data were used to downscale the future rainfall from ensembles of 20 GCMs of Coupled Model Intercomparison Project phase 5 (CMIP5) for four Representative Concentration Pathways (RCP) scenarios, namely, RCP2.6, RCP4.5, RCP6.0 and RCP8.5. Model Output Statistics (MOS) based downscaling models were developed using two data mining approaches known as Random Forest (RF) and Support Vector Machine (SVM). The SVM was found to downscale all GCMs with normalized mean square error (NMSE) of 48.2–75.2 and skill score (SS) of 0.94–0.98 during validation. The results show that the future projection of the annual rainfalls is increasing and decreasing on the region-based and catchment-based basis due to the influence of the monsoon season affecting the coast of Sarawak. The ensemble mean of GCMs projections reveals the increased and decreased mean of annual precipitations at 33 stations with the rate of 0.1% to 19.6% and one station with the rate of − 7.9% to − 3.1%, respectively under all RCP scenarios. The remaining 15 stations showed inconsistency neither increasing nor decreasing at the rate of − 5.6% to 5.2%, but mainly showing a trend of decreasing rainfall during the first period (2010–2039) followed by increasing rainfall for the period of 2070–2099.