Scholarlink

<p>Accurate estimation of surface ocean <i>p</i>CO<sub>2</sub> is crucial for understanding the ocean’s role in the global carbon cycle and its response to climate change. In this study, we employ a machine learning algorithm to correct the deviations in high-resolution (1/12°) model simulations of surface <i>p</i>CO<sub>2</sub> from the INCOIS-BIO-ROMS model (<i>p</i>CO<sub>2</sub><sup>model</sup>) for the period 1980–2019, using available observations (<i>p</i>CO<sub>2</sub><sup>obs</sup>). We train the XGBoost model to generate spatio-temporal deviations (<i>p</i>CO<sub>2</sub><sup>obs</sup> − <i>p</i>CO<sub>2</sub><sup>model</sup>) of <i>p</i>CO<sub>2</sub><sup>model</sup>. The interannually and climatologically varying deviations are then added back to the original model separately, which results in an improved surface <i>p</i>CO<sub>2</sub> data product. A comparison of our surface <i>p</i>CO<sub>2</sub> data product with moored observations, gridded SOCAT, CMEMS-LSCE-FFNN, and OceanSODA demonstrates an improvement by approximately 40% ± 3.31% in RMSE. Further analysis reveals that adding climatological deviations to <i>p</i>CO<sub>2</sub><sup>model</sup> results in greater improvements than adding interannual deviations. This analysis underscores the ability of machine learning algorithms to enhance the accuracy of model-simulated surface <i>p</i>CO<sub>2</sub> outputs.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

An improved long-term high-resolution surface pCO2 data product for the Indian Ocean using machine learning

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年4月5日

查看全文
  • Prasanna Kanti Ghoshal,
  • A.P. Joshi,
  • Kunal Chakraborty

摘要

Accurate estimation of surface ocean pCO2 is crucial for understanding the ocean’s role in the global carbon cycle and its response to climate change. In this study, we employ a machine learning algorithm to correct the deviations in high-resolution (1/12°) model simulations of surface pCO2 from the INCOIS-BIO-ROMS model (pCO2model) for the period 1980–2019, using available observations (pCO2obs). We train the XGBoost model to generate spatio-temporal deviations (pCO2obs − pCO2model) of pCO2model. The interannually and climatologically varying deviations are then added back to the original model separately, which results in an improved surface pCO2 data product. A comparison of our surface pCO2 data product with moored observations, gridded SOCAT, CMEMS-LSCE-FFNN, and OceanSODA demonstrates an improvement by approximately 40% ± 3.31% in RMSE. Further analysis reveals that adding climatological deviations to pCO2model results in greater improvements than adding interannual deviations. This analysis underscores the ability of machine learning algorithms to enhance the accuracy of model-simulated surface pCO2 outputs.