Comparison of two techniques to separate physical- and biological-mediated pCO(2) in seawater

In this methodical work, we investigate the applicability of two approaches to determine the biologically induced variations of marine pCO(2). The first method was proposed by Takahashi et al. (1993), who used data of temperature and pCO(2) in the North Atlantic to linearize the temperature dependency of pCO(2). We compare the Takahashi method with a superposition approach by using model results. The superposition approach assumes that biological, chemical and physical influences on the partial pressure of carbon dioxide are adding up. Thus, the biologically induced pCO(2) variations are calculated as the difference between a standard run with an ecosystem model and a run in which only physical and chemical processes are taken into account. Both methods agree reasonably for a station in the northern North Sea, characterized by weak dissolved inorganic carbon (DIC) gradients and moderate biological production. The deviations between both approaches are larger for a station in the southern North Sea, characterized by stronger DIC gradients and higher biological activities. In this area, the Takahashi approach has not only the problem that sinks and sources of DIC mask the biologically induced pCO(2) variations but also that the constant for the linearization (lp = 0.0423) has to be changed. We show that during strong events which increase the pH (> 8.3) the pCO(2) could not be determined correctly with this method. In all other cases, the use of the Takahashi constant is valid. Locally and temporally calculated constants, which substitute lp, lead to deviations in the calculated pCO(2) smaller than 2 mu atm