Evaluation of CMIP5 models on sea surface salinity in the Indian Ocean

Prior to future climate assessment of the 5th Coupled Model Intercomparison (CMIP5) experiments, how well CMIP5 models simulates present climate should be examined. Sea surface salinity (sss) play important role in ocean stratification and indirectly affects air sea interaction. However, few studies have been carried out to evaluate sss in CMIP5 models. In this study, performance of CMIP5 models in simulating sss in Indian Ocean was examined with respect to the observation. Our results showed that multi model ensemble (MME) mean of CMIP5 models displayed annual and seasonal salinity bias in three regions i.e. Western Indian Ocean (WIO), Bay of Bengal (BOB) and Southeastern Indian Ocean (SEIO). CMIP5 models overestimate sss in BOB about 1.5 psu and underestimated sss in WIO and SEIO about 0.4 psu. Biases in WIO and BOB were mainly attributed to bias in precipitation. CMIP5 models overestimated (underestimated) precipitation in WIO (BOB) with greater bias found during Boreal summer to winter. Meanwhile, advection process was responsible for negative SSS bias in SEIO.