FY 2012
Robust empirical relationships for estimating the carbonate system in the southern California Current System and application to CalCOFI hydrographic cruise data (2005-2011)
Alin, S.R., R.A. Feely, A.G. Dickson, J.M. Hernández-Ayón, L.W. Juranek, M.D. Ohman, and R. Goericke
J. Geophys. Res., 117(C5), C05033, doi: 10.1029/2011JC007511 (2012) |
| The California Current System (CCS) is expected to experience the ecological impacts of ocean acidification (OA) earlier than most other ocean regions because coastal upwelling brings old, CO2-rich water relatively close to the surface ocean. Historical inorganic carbon measurements are scarce, so the progression of OA in the CCS is unknown. We used a multiple linear regression approach to generate empirical models using oxygen (O2), temperature (T), salinity (S), and sigma theta (σθ) as proxy variables to reconstruct pH, carbonate saturation states, carbonate ion concentration ([CO32−]), dissolved inorganic carbon (DIC) concentration, and total alkalinity (TA) in the southern CCS. The calibration data included high-quality measurements of carbon, oxygen, and other hydrographic variables, collected during a cruise from British Columbia to Baja California in May–June 2007. All resulting empirical relationships were robust, with r2 values >0.92 and low root mean square errors. Estimated and measured carbon chemistry matched very well for independent data sets from the CalCOFI and IMECOCAL programs. Reconstructed CCS pH and saturation states for 2005–2011 reveal a pronounced seasonal cycle and inter-annual variability in the upper water column. Deeper in the water column, conditions are stable throughout the annual cycle, with perennially low pH and saturation states. Over sub-decadal time scales, these empirical models provide a valuable tool for reconstructing carbonate chemistry related to ocean acidification where direct observations are limited. However, progressive increases in anthropogenic CO2 content of southern CCS water masses must be carefully addressed to apply the models over longer time scales. |
