Evaluation of Global Climate Models (GCMs) in Representing Climate Phenomena Signatures Over Maritime Continent

Abstract

The improvements of resolution and physical parameterization in the GCMs from Coupled Model Intercomparison Project Phase 5 (CMIP5) to Phase 6 (CMIP6) do not always guarantee the better representation of climate in several regions. The Maritime Continent (MC) region is influenced by dominant modes of climate phenomena such as the Madden Julian Oscillation (MJO), Indian Ocean Dipole (IOD), and El Niño-Southern Oscillation (ENSO). However, the representation of those phenomena from CMIP5 to CMIP6 models over MC has not been assessed yet. This study evaluates the signature of those phenomena on daily precipitation data of twelve models from each CMIP5 and CMIP6 against ERA-5 (as a reference) in the historical period (1981-2005) over MC. We perform Multiple Linear Regression (MLR) to assess the spatial signatures on daily precipitation anomalies and quantify the spatial similarity between models and reference using Cosine Similarity (CS). Then, we rank the CS values to obtain the best model. Result shows that only half of 12 models in CMIP6 improve from the models in CMIP5. Overall, MIROC5 (CMIP5) is the best model in representing the signature of those phenomena. Therefore, this study recommends the best models for climate change projection and adaptation strategy over the MC.