1244 / 2024-09-20 19:43:54

GOCI-estimated air-sea CO2 fluxes in the East China Sea: Patterns and variations during summer 2011–2020

Air-sea CO2 flux,sea surface pCO2,Remote sensing (RS),east china sea,east china sea,Semi-analytical algorithm (MeSAA),Machine learning,Geostationary Ocean Color Imager(GOCI)

Inadequate spatiotemporal representativeness of observational data can lead to the oversight of significant short-term fluctuations in sea-atmosphere carbon dioxide (CO₂) flux. This oversight risks misrepresenting the ocean's CO₂ source-sink dynamics, particularly in marginal seas where these fluxes can vary greatly. To address this challenge, this study combines data-driven machine learning techniques with process analysis to develop a remote sensing model for sea surface pCO₂ using geostationary satellite data. This model enables large-scale, daily observations of CO₂ exchange processes in the East China Sea during summer.The model's performance has been validated, showing strong results with a root mean square error (RMSE) of 27 μatm and an R² of 0.96. Additionally, SHAP values provide valuable insights into the underlying mechanisms influencing pCO₂. While no significant upward trend in summer sea surface pCO₂ was identified from 2011 to 2020, the carbon sink capacity of the East China Sea has gradually increased in response to rising atmospheric pCO₂ levels, aligning with estimates from polar orbit satellites.However, the lack of polar satellite data in coastal regions (9.86%) has led to a 71.13% overestimation of cumulative carbon sink estimates for the East China Sea during the summers of 2011 to 2019. This study contributes important new insights into the daily variations of CO₂ exchange at the sea-atmosphere interface.