Skillful Multiyear Predictions of pH Variability in the California Current System

The California Current System (CCS) is an eastern boundary upwelling system characterized by high primary productivity supported by upwelling of nutrient-rich waters from depth. These waters also are naturally corrosive, with relatively high pCO² and relatively low pH. The compounded effects of corrosive upwelling and the diffusion of anthropogenic CO² into surface waters has already begun to cause shell degradation in pteropods along the continental shelf. Thus, skillful seasonal to decadal forecasts of carbonate chemistry parameters, such as pH, are crucial for the proper management of living marine resources in the CCS. Here we utilize output from the Community Earth System Model Decadal Prediction Large Ensemble (CESM-DPLE), a forecasting system that initializes 40 ensemble members annually from a forced ocean sea-ice reconstruction, to consider the potential to predict pH in the CCS. We show that CESM-DPLE has statistically significant predictability in annual pH anomalies out to five years over a persistence forecast, with some regions exhibiting predictability out to eight years. Further, we find that CESM-DPLE can skillfully predict past pH anomalies that occurred in the observational record. Predictability in CCS pH is driven by the multiyear persistence of dissolved inorganic carbon anomalies, with some contribution from the forecasting system's ability to predict sea surface temperature and alkalinity anomalies. Our results demonstrate the utility of an initialized decadal prediction ensemble for managing the onset and impacts of ocean acidification in the CCS.