The Bay of Bengal is a strongly stratified and climate-sensitive region of the Indian Ocean, where monsoon forcing, intense river discharge, precipitation, tropical cyclones, and air–sea interactions regulate upper-ocean variability. Accurate representation of sea surface temperature, sea surface salinity, mixed-layer depth, and sea surface height is therefore essential for understanding regional ocean dynamics and improving predictions of monsoon variability and climate-related extreme events. However, the ability of global ocean reanalyses products to reproduce these interconnected upper-ocean processes remains insufficiently evaluated over the Bay of Bengal.
This study assesses upper-ocean variability and model skill in three high-resolution global ocean reanalysis products—FIO-COM, GLORYS4, and BRAN—over the Bay of Bengal during 2013–2022. The evaluation uses multiple observational datasets, including OISST for sea surface temperature, EN4 for salinity and upper-ocean hydrography, and satellite altimetry for sea surface height. Model performance is quantified using bias, root-mean-square error, mean absolute error, correlation, seasonal cycles, annual variability, and spatial error distributions.
All three products reproduce the broad seasonal evolution of the Bay of Bengal upper ocean, but substantial differences occur in regions affected by freshwater input and strong stratification. FIO-COM provides the most consistent overall performance, particularly for sea surface salinity, with a domain-averaged RMSE of 0.48, compared with 0.56 for BRAN and 1.38 for GLORYS4. For mixed-layer depth, FIO-COM produces the lowest mean RMSE and mean absolute error, at 8.85 m and 6.45 m, respectively, whereas GLORYS4 shows the smallest mean bias. The largest discrepancies occur during monsoon and post-monsoon periods, when river discharge, precipitation, and shallow stratification exert strong control on upper-ocean structure. These results demonstrate that model skill in the Bay of Bengal is strongly dependent on the representation of freshwater forcing, vertical mixing, and stratification. The intercomparison provides a scientific basis for selecting suitable ocean products for regional modeling and highlights the importance of accurately resolving upper-ocean processes for understanding Indian Ocean dynamics under increasing climate stress.

keywords: Bay of Bengal; Indian Ocean; ocean reanalyses; freshwater forcing; upper-ocean stratification; mixed-layer depth; monsoon variability