ABSTRACT
Asok, A.B.; Sumangala, D.; Joshi, A., and Warrior, H.V., 2025. Enhancing mixed-layer depth predictions in the Bay of Bengal: A comparative study using CMIP6 models and artificial neural networks.
This paper presents a comprehensive evaluation of the mixed-layer depth (MLD) variations in the Bay of Bengal using data from 30 coupled model intercomparison project phase 6 (CMIP6) models. The study aims to assess the performance of these models in capturing MLD dynamics in the region and identifies the CAMS-CSM1-0 model as having superior accuracy compared with others. Additionally, artificial neural networks (ANNs) are used to refine MLD estimations, enhancing predictions derived from the CMIP6 model. The MLD predictions are made with ANN on the CAMS-CSM1-0 model alone and with an ensemble of the top 10 CMIP6 models, resulting in a correlation of up to 0.98 in the latter case. The paper also investigates the relationship between pH variabilities in the Bay of Bengal and MLD. The study offers insights into regional MLD variability and climate resilience efforts, providing valuable information for researchers and policymakers. Potential methods to further improve MLD predictions are also discussed.
