Evaluation of upper ocean processes in the northern Indian Ocean simulated by CMIP5 models

Abstract

CMIP5 (Coupled Model Intercomparison Project Phase 5) coupled models provide up-to-date scientific information on global climate change, contributing to the recent fifth assessment report of the Intergovernmental Panel on Climate Change. In this study, we aimed to evaluate the upper ocean processes in the northern Indian Ocean simulated by 19 CMIP5 models by comparing their historical run simulation with observed climatology. The comparison analysis shows that CMIP5 models still have significant biases in the upper ocean in the northern Indian Ocean: deep winter MLD biases in the Arabian Sea and the Bay of Bengal, saltier biases in the Bay of Bengal in all seasons, and cold biases in most of the northern Indian Ocean except near the western Arabian Sea in summer showing a warm bias. In addition, the inter-model standard deviation of the biases suggests a significant difference in CMIP5 simulation in winter MLD in the Arabian Sea, in the SST in the Arabian Sea and in SSS in the Bay of Bengal. The deep MLD biases and the cold SST biases in winter appear to be related with a stronger wind bias in CMIP5 models. This relationship among the biases in CMIP5 models suggests importance of proper wind representation for realistic upper ocean simulation in the northern Indian Ocean. this study, we aimed to evaluate the upper ocean processes in the northern Indian Ocean simulated by 19 CMIP5 models by comparing their historical run simulation with observed climatology. The comparison analysis shows that CMIP5 models still have significant biases in the upper ocean in the northern Indian Ocean: deep winter MLD biases in the Arabian Sea and the Bay of Bengal, saltier biases in the Bay of Bengal in all seasons, and cold biases in most of the northern Indian Ocean except near the western Arabian Sea in summer showing a warm bias. In addition, the inter-model standard deviation of the biases suggests a significant difference in CMIP5 simulation in winter MLD in the Arabian Sea, in the SST in the Arabian Sea and in SSS in the Bay of Bengal. The deep MLD biases and the cold SST biases in winter appear to be related with a stronger wind bias in CMIP5 models. This relationship among the biases in CMIP5 models suggests importance of proper wind representation for realistic upper ocean simulation in the northern Indian Ocean.