Characterization of Monthly Precipitation over Indochina Region to Evaluate CMIP5 Historical Runs

Abstract

Toward the evaluation of climate simulations, the exploration of local climate contexts is an important step to understand how well the simulation can reproduce that regional climate. In this study, we characterize climatology of precipitation over Indochina Region (ICR) to evaluate CMIP5 historical simulations; such characterization is not much done for this region. After an intercomparison of four gridded precipitation datasets, we choose APHRODITE for further characterization of the climatology of precipitation for this region. The EOF analysis of APHRODITE dataset for the region shows two dominant precipitation patterns corresponding to the first and second EOFs; the first mode significantly connects with the annual variation of monsoon circulation (monotonic structure), and the second mode more connects with ENSO phenomena (north-south seesaw structure). IOD phenomenon seems to attribute to the first mode rather than the second mode. The time variation of the first two principal components (PCs) from EOF analysis conveys some climate information which is meaningful to evaluate GCMs using those modes. 14 metrics are obtained from this characterization to evaluate CMIP5's GCMs for this region. The initial results based on these metrics suggest that HadGEM2-CC, HadGEM2-ES, MPI-ESM-LR, IPSLCM5A- LR, and HadGEM2-AO models are the five highest ranking in simulating precipitation over ICR.