A Reconstructed Coastal Acidification Database (ReCAD) pCO2 data product for the North American Atlantic Coastal Ocean Margins

<h1><strong>Reconstructed Coastal Acidification Database (ReCAD)</strong></h1> <p>Insufficient spatiotemporal coverage of the partial pressure of CO₂ (<em>p</em>CO₂) observations has hindered precise carbon cycle studies in coastal oceans and justifies the development of spatially and temporally continuous <em>p</em>CO₂ data products. Earlier <em>p</em>CO₂ products have difficulties in capturing the heterogeneity of regional variations and decadal trends of <em>p</em>CO₂ in the North American Atlantic Coastal Ocean Margin (NAACOM). This study developed a regional reconstructed <em>p</em>CO₂ product for the NAACOM (Reconstructed Coastal Acidification Database-<em>p</em>CO₂, or ReCAD-NAACOM-<em>p</em>CO₂) using a two-step approach combining random forest regression and linear regression. The product provides monthly <em>p</em>CO₂ data at 0.25&deg; spatial resolution from 1993 to 2021, enabling investigation of regional spatial differences, seasonal cycles, and decadal changes in <em>p</em>CO₂. The observation-based reconstruction was trained using Surface Ocean CO₂ Atlas (SOCAT) observations as observational values, with various satellite-derived and reanalysis environmental variables known to control sea surface <em>p</em>CO₂ as model inputs. The product shows high accuracy during the model training, validation, and independent test phases, demonstrating robustness and capability to accurately reconstruct <em>p</em>CO₂ in regions or periods lacking direct observational data. Compared with all the observation samples from SOCAT, the <em>p</em>CO₂ product yields a determination coefficient of 0.92, a root-mean-square error of 12.70 &micro;atm, and an accumulative uncertainty of 23.25 &micro;atm. The ReCAD-NAACOM-<em>p</em>CO₂ product demonstrates its capability to resolve seasonal cycles, regional-scale variations, and decadal trends of <em>p</em>CO₂ along the NAACOM. This new product provides reliable <em>p</em>CO₂ data for more precise studies of coastal carbon dynamics in the NAACOM region. The dataset is publicly accessible at <a href="https://doi.org/10.5281/zenodo.11500974">https://doi.org/10.5281/zenodo.11500974</a> (Wu et al., 2024a) and will be updated regularly.</p> <div> <div>&nbsp;</div> </div> <p><strong>Version 1.1: </strong>Update the training set output as the direct model outputs instead of being the 10-fold cross-validation output in v1.0.</p> <p><strong>Data description paper</strong><em><strong>: </strong></em>Wu, Z., Lu, W., Roobaert, A., Song, L., Yan, X.-H., and Cai, W.-J.: A machine-learning reconstruction of sea surface&nbsp;<em>p</em>CO₂ in the North American Atlantic Coastal Ocean Margin from 1993 to 2021, Earth Syst. Sci. Data, 17, 43&ndash;63, <a href="https://doi.org/10.5194/essd-17-43-2025">https://doi.org/10.5194/essd-17-43-2025</a>, 2025.</p>