Quantifying coastal ocean anthropogenic carbon changes and their uncertainties

Abstract Despite recent progress in quantifying anthropogenic carbon (C anth ) in global oceans, coastal regions remain understudied due to methodological and observational challenges. Here, we develop and generalize a Regional rEgression–based Coastal Anthropogenic carbon estimation algorithm (RECA) designed specifically for coastal environments. Regional rEgression–based Coastal Anthropogenic carbon estimation algorithm builds upon the extended multiple linear regression framework, incorporating adaptations such as improved regression construction strategies, optimized data selection, and reduced subjectivity in decision‐making, creating a unified approach across diverse coastal settings. Using synthetic datasets with known ∆C anth from global ocean biogeochemical model simulations across multiple North American coastal regions, we estimate that the uncertainty of RECA is about 3.25 μ mol kg −1 . We also compare RECA with two other regression‐based algorithms, originally developed for the open ocean, to evaluate their performance in coastal environments after some adaptations. All three approaches have similar biases but different strengths, underscoring the importance of using a consistent method. An ensemble approach combining these three algorithms modestly improves reconstruction fidelity, but at the cost of increased complexity. We also evaluate the contributions of non‐steady‐state variations of both natural and C anth components because uncertainty about these components has long challenged the interpretation of regression‐based ∆C anth results. We find that the algorithms indeed remove most, but not all, of the natural variations and appear to capture some, but not all, of the non‐steady‐state C anth . This methodological framework provides a robust, broadly applicable, and easy‐to‐implement tool for estimating coastal ∆C anth with uncertainty quantification, allowing us to more confidently track C anth accumulation in a changing dynamic coastal ocean.