A machine-learning reconstruction of sea surface <i>p</i>CO<sub>2</sub> in the North American Atlantic Coastal Ocean Margin from 1993 to 2021
<p>Insufficient spatiotemporal coverage of observations of the surface partial pressure of CO<span class="inline-formula"><sub>2</sub></span> (<span class="inline-formula"><i>p</i></span>CO<span class="inline-formula&q...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2025-01-01
|
| Series: | Earth System Science Data |
| Online Access: | https://essd.copernicus.org/articles/17/43/2025/essd-17-43-2025.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | <p>Insufficient spatiotemporal coverage of observations of the surface partial pressure of CO<span class="inline-formula"><sub>2</sub></span> (<span class="inline-formula"><i>p</i></span>CO<span class="inline-formula"><sub>2</sub></span>) has hindered precise carbon cycle studies in coastal oceans and justifies the development of spatially and temporally continuous <span class="inline-formula"><i>p</i></span>CO<span class="inline-formula"><sub>2</sub></span> data products. Earlier <span class="inline-formula"><i>p</i></span>CO<span class="inline-formula"><sub>2</sub></span> products have difficulties in capturing the heterogeneity of regional variations and decadal trends of <span class="inline-formula"><i>p</i></span>CO<span class="inline-formula"><sub>2</sub></span> in the North American Atlantic Coastal Ocean Margin (NAACOM). This study developed a regional reconstructed <span class="inline-formula"><i>p</i></span>CO<span class="inline-formula"><sub>2</sub></span> product for the NAACOM (Reconstructed Coastal Acidification Database-<span class="inline-formula"><i>p</i></span>CO<span class="inline-formula"><sub>2</sub></span>, or ReCAD-NAACOM-<span class="inline-formula"><i>p</i></span>CO<span class="inline-formula"><sub>2</sub></span>) using a two-step approach combining random forest regression and linear regression. The product provides monthly <span class="inline-formula"><i>p</i></span>CO<span class="inline-formula"><sub>2</sub></span> data at 0.25° spatial resolution from 1993 to 2021, enabling investigation of regional spatial differences, seasonal cycles, and decadal changes in <span class="inline-formula"><i>p</i></span>CO<span class="inline-formula"><sub>2</sub></span>. The observation-based reconstruction was trained using Surface Ocean CO<span class="inline-formula"><sub>2</sub></span> Atlas (SOCAT) observations as observational values, with various satellite-derived and reanalysis environmental variables known to control sea surface <span class="inline-formula"><i>p</i></span>CO<span class="inline-formula"><sub>2</sub></span> as model inputs. The product shows high accuracy during the model training, validation, and independent test phases, demonstrating robustness and a capability to accurately reconstruct <span class="inline-formula"><i>p</i></span>CO<span class="inline-formula"><sub>2</sub></span> in regions or periods lacking direct observational data. Compared with all the observation samples from SOCAT, the <span class="inline-formula"><i>p</i></span>CO<span class="inline-formula"><sub>2</sub></span> product yields a determination coefficient of 0.92, a root-mean-square error of 12.70 <span class="inline-formula">µ</span>atm, and an accumulative uncertainty of 23.25 <span class="inline-formula">µ</span>atm. The ReCAD-NAACOM-<span class="inline-formula"><i>p</i></span>CO<span class="inline-formula"><sub>2</sub></span> product demonstrates its capability to resolve seasonal cycles, regional-scale variations, and decadal trends of <span class="inline-formula"><i>p</i></span>CO<span class="inline-formula"><sub>2</sub></span> along the NAACOM. This new product provides reliable <span class="inline-formula"><i>p</i></span>CO<span class="inline-formula"><sub>2</sub></span> data for more precise studies of coastal carbon dynamics in the NAACOM region. The dataset is publicly accessible at <span class="uri">https://doi.org/10.5281/zenodo.14038561</span> (Wu et al., 2024a) and will be updated regularly.</p> |
|---|---|
| ISSN: | 1866-3508 1866-3516 |