Modeling the sea-surface pCO2 of the central Bay of Bengal region using machine learning algorithms. (October 2022)

• Record Type: Journal Article
• Authors: Joshi, A.P.; Kumar, V.; Warrior, H.V.
• Is Part Of: Ocean modelling. Volume 178(2022)
• Publication Date: 2022-10
• Subjects: Bay of Bengal (BoB) -- Partial pressure of carbon dioxide (pCO2) -- pCO2 trends -- ANN -- XGBoost; Oceanography -- Periodicals; Océanographie -- Périodiques; Oceanography; Periodicals; 551.46
• DOI: 10.1016/j.ocemod.2022.102094
• Languages: English
• ISSNs: 1463-5003

Abstract

Abstract: The present study explores the capabilities of advanced machine learning algorithms in predicting the sea-surface p CO2 (partial pressure of carbon dioxide) in the open oceans of the Bay of Bengal (BoB). We collect the available observations (outside EEZ (Exclusive Economic Zone)) from the cruise tracks and the mooring stations. Due to the paucity of data in the BoB, we attempt to predict p CO2 based on the Sea Surface Temperature (SST) and the Sea Surface Salinity (SSS). Comparing the MLR, the ANN, and the XGBoost algorithm against a common dataset reveals that the XGBoost performs the best for predicting the sea-surface p CO2 in the BoB. Using the satellite-derived SST and SSS, we predict the sea-surface p CO2 using the XGBoost model and compare the same with the in-situ observations. The model performs satisfactorily, having a correlation of 0.75 and the RMSE of ± 12 . 23 μ atm. Further using this model, we emulate the monthly variations in the sea-surface p CO2 for the central BoB between 2010–2019. Using the satellite data, we show that the central BoB is warming at a rate of 0.0175 °C per year, whereas the SSS decreases at a rate of -0.0088 PSU per year. The modeled p CO2 shows a declination at a rate of −0.4852 μ atm per year. We perform sensitivity experiments to find that the variations in SST and SSS contribute ≈ 41% and ≈ 37% to the declining trends of the p CO2 for the last decade. Seasonal analysis shows that the pre-monsoon season has the highest rate of decrease of the sea-surface p CO2 . Highlights: Performance of MLR, ANN, and XGboost for emulating sea-surface p CO2 is evaluated. XGBoost outperforms MLR and ANN in reproducing sea-surface p CO2 . The p CO2 reproducibility using satellite-derived SST and SSS is best using XGBoost. The central BoB has been warming at a rate of 0.0175°C per year during 2010–2019. The sea-surface p CO2 decreases at a rate of −0.4852 μ atm per year in the BoB.