The California Current System is expected to experience the ecological impacts of ocean acidification earlier than most other ocean regions because marine waters in the North Pacific are among the oldest in the global oceans and natural upwelling processes in this eastern boundary current system bring CO2-rich water masses to the surface in coastal oceans during late spring-early fall months. We used a multiple linear regression (MLR) approach to generate predictive models using oxygen and temperature as proxy variables to reconstruct pH and carbonate saturation states in the Southern California Bight. The calibration data set included high-quality measurements of dissolved inorganic carbon, alkalinity, oxygen, temperature, salinity, and nutrients and was collected during a cruise from British Columbia to Baja California in May-June 2007. The resulting relationships predicting pH and aragonite and calcite saturation states (Ω) from oxygen and temperature data were robust, with r2 values >0.98 and root mean square errors of 0.020 (pH), 0.048 (Ωarag), and 0.075 (Ωcalc). Predicted vs. measured ocean acidification conditions (i.e. pH, Ωarag, and Ωcalc) matched very well for seven verification data sets collected between 2008 and 2010 during quarterly CalCOFI cruises in the Southern California Bight and during several sampling dates on an Ensenada transect occupied several times between 2006 and 2010. Over sub-decadal time scales, these predictive models provide a valuable tool for reconstructing historical time-series of ocean acidification conditions in the California Current Ecosystem where historical inorganic carbon measurements are scarce. Reconstructed pH and saturation state values based on CalCOFI oxygen and temperature data for all cruises between 2005 and 2010 reveal a seasonal cycle in the upper water column, with higher pH and Ω values present during the winter cruises, and stronger gradients including much lower pH and Ω values during spring through fall cruises. Deeper in the water column (~300 m), conditions are more stable throughout the annual cycle, with consistently low pH, undersaturation with respect to aragonite, and calcite saturation values <1.5. These predictive relationships can also be used to improve the performance of models used to “nowcast” and forecast ocean acidification in eastern boundary current systems like the California Current System.
Alin, S. R., Feely, R. A., Dickson, A. G., Hernandez-Ayon, J. M., Juranek, L. W., Ohman, M. D. & Goericke, R., 2010. Predictive relationships for pH and carbonate saturation in the Southern California Current System using oxygen and temperature data. American Geophysical Union, Fall Meeting 2010, abstract #OS23A-1570. Abstract.
