An observation-based method to estimate carbonate system variations in the Labrador Sea

Abstract

Read online

The ocean carbonate system consists of pH, alkalinity, inorganic carbon and the partial pressure of carbon dioxide, and during the current era of anthropogenic change, its dynamics are key for understanding changes in the ocean and its ecosystem over time. The focus of this study is to estimate the carbonate system in the Labrador Sea with time series methods, using direct observations from the ocean surface and interior, and chemical relationships between variables. Interior ocean observations are minimal for some of these variables, however, connections between the variables rooted in chemistry were used to create pseudo-observations using CO2SYS, increasing the information available. A state space model was designed that combined GLODAP and SOCAT observations along with pseudo-observations in a time series estimate of the carbonate system. The Labrador Sea between 1993 and 2016 shows increasing rates for DIC (0.57-1.16 µmol kg−1 year−1) and fCO2 (0.70-2.45 µatm year−1), as well as acidification via pH trends (0.0007-0.0018 year−1). These ranges describe the scale of rates that are occurring at various depths through the water column, though they do not change linearly with depth. Largest rates are found at the surface for DIC, 500-1500 m for fCO2, and 500-1500 m for pH. Total alkalinity also decreased and is correlated with the freshening of salinity. With the core carbonate variables estimated, other aspects of the carbonate system are calculated using CO2SYS, such as the aragonite and calcite saturation states, the Revelle factor, and the carbonate species. Our method also calculates uncertainties that vary over time and depth based on the availability of observations and their variance, which has lowered the uncertainty for pH by 71% and for fCO2 by 64% compared to time-independent methods.

Keywords

• carbonate system
• acidification
• North Atlantic
• time series
• Kalman smoother
• GLODAP