Comparison of spectrophotometric and electrochemical pH measurements for calculating freshwater pCO(2)

Inland waters have an important role in the global carbon cycle, contributing significantly to terrestrial carbon fluxes through downstream export and exchange of CO2 with the atmosphere. However, large uncertainties in freshwater inorganic carbon fluxes remain. One contributing factor is uncertainty in carbonate system calculations for estimating the partial pressure of CO2 (pCO(2)) from pH and alkalinity in freshwater systems. The uncertainty stems largely from inaccurate pH values caused by glass pH electrode measurements in low ionic strength systems. This study compares indicator-based spectrophotometric and electrochemical pH measurements and their application for calculating freshwater pCO(2). Our study found that, compared to a pCO(2) reference method, pH electrode-based estimates of pCO(2) were overestimated by 230 +/- 200 mu atm (n = 54) where indicator-based spectrophotometric pH estimates of pCO(2) were 58 +/- 33 mu atm (n = 34) over the range of 1001600 mu atm. Furthermore, we found that when ionic strength was assumed to be zero, calculated pCO(2) error was similar to 20% of the reference pCO(2). A 19-d field study using autonomous spectrophotometric pH and pCO(2) sensors found an average error in calculated pCO(2) of -70 +/- 57 mu atm (n = 1685). Although, our focus is on riverine CO2, these findings and subsequent conclusions apply to all freshwater systems. Spectrophotometric pH measurements will improve future freshwater pCO(2) calculations and better quantify inland waters' role in the global carbon budget.