A monthly surface pCO2 product for the California Current Large Marine Ecosystem

A common strategy for calculating the direction and rate of carbon dioxide gas (CO2) exchange between the ocean and atmosphere relies on knowledge of the partial pressure of CO2 in surface seawater (PCO2(sw), a quantity that is frequently observed by autonomous sensors on ships and moored buoys, albeit with significant spatial and temporal gaps. Here we present a monthly gridded data product of pCO(2(sw)) at 0.25 degrees latitude by 0.25 degrees longitude resolution in the northeastern Pacific Ocean, centered on the California Current System (CCS) and spanning all months from January 1998 to December 2020. The data product (RFR-CCS; Sharp et al., 2022; https://doi.org/10.5281/zenodo.5523389) was created using observations from the most recent (2021) version of the Surface Ocean CO2 Atlas (Bakker et al., 2016). These observations were fit against a variety of collocated and contemporaneous satellite- and model-derived surface variables using a random forest regression (RFR) model. We validate RFR-CCS in multiple ways, including direct comparisons with observations from sensors on moored buoys, and find that the data product effectively captures seasonal pCO(2(sw)) cycles at nearshore sites. This result is notable because global gridded pCO(2(sw)) products do not capture local variability effectively in this region, suggesting that RFR-CCS is a better option than regional extractions from global products to represent pCO(2(sw)) in the CCS over the last 2 decades. Lessons learned from the construction of RFR-CCS provide insight into how global pCO(2(sw)) products could effectively characterize seasonal variability in nearshore coastal environments. We briefly review the physical and biological processes - acting across a variety of spatial and temporal scales - that are responsible for the latitudinal and nearshore-to-offshore pCO(2(sw)) gradients seen in the RFR-CCS reconstruction of pCO(2(sw)). RFR-CCS will be valuable for the validation of high-resolution models, the attribution of spatiotemporal carbonate system variability to physical and biological drivers, and the quantification of multiyear trends and interannual variability of ocean acidification.

Automated summary

Researchers have provided a pCO(2(sw)) data product for the northeastern Pacific Ocean, specifically the California Current System, which effectively captures seasonal CO2 cycles nearshore. The product, RFR-CCS, outperforms global gridded products in representing local variability. It can be used for model validation, quantifying multiyear trends and interannual variability of ocean acidification.