Abstract. Inland waters have been recognized as a significant source of carbon dioxide (CO 2 ) to the atmosphere at the global scale. Fluxes of CO 2 between aquatic systems and the atmosphere are calculated from the gas transfer velocity and the water-air gradient of the partial pressure of CO 2 (pCO 2 ). Currently, direct measurements of water pCO 2 remain scarce in freshwaters, and most published pCO 2 data are calculated from temperature, pH and total alkalinity (TA). Here, we compare calculated (pH and TA) and measured (equilibrator and headspace) water pCO 2 in a large array of temperate and tropical freshwaters. The 761 data points cover a wide range of values for TA (0 to 14 200 µmol L −1 ), pH (3.94 to 9.17), measured pCO 2 (36 to 23 000 ppmv), and dissolved organic carbon (DOC) (29 to 3970 µmol L −1 ). Calculated pCO 2 were > 10 % higher than measured pCO 2 in 60 % of the samples (with a median overestimation of calculated pCO 2 compared to measured pCO 2 of 2560 ppmv) and were > 100 % higher in the 25 % most organic-rich and acidic samples (with a median overestimation of 9080 ppmv). We suggest these large overestimations of calculated pCO 2 with respect to measured pCO 2 are due to the combination of two cumulative effects: (1) a more significant contribution of organic acids anions to TA in waters with low carbonate alkalinity and high DOC concentrations; (2) a lower buffering capacity of the carbonate system at low pH, which increases the sensitivity of calculated pCO 2 to TA in acidic and organicrich waters. No empirical relationship could be derived from our data set in order to correct calculated pCO 2 for this bias.Owing to the widespread distribution of acidic, organic-rich freshwaters, we conclude that regional and global estimates of CO 2 outgassing from freshwaters based on pH and TA data only are most likely overestimated, although the magnitude of the overestimation needs further quantitative analysis. Direct measurements of pCO 2 are recommended in inland waters in general, and in particular in acidic, poorly buffered freshwaters.

Abstract. Inland waters have been recognized as a~significant source of carbon dioxide (CO2) to the atmosphere at the global scale. Fluxes of CO2 between aquatic systems and the atmosphere are calculated from the gas transfer velocity and the water-air gradient of the partial pressure of CO2 (pCO2). Nowadays, direct measurements of water pCO2 remain scarce in freshwaters and most published pCO2 data are calculated from temperature, pH and total alkalinity (TA). Here, we compare calculated (pH and TA) and measured (Equilibrator and headspace) water pCO2 in a large array of temperate and tropical freshwaters. The 761 data points cover a wide range of values for TA (0 to 14.2 mmol L−1), pH (3.94 to 9.17), measured pCO2 (36 to 23 000 ppmv), and dissolved organic carbon (DOC) (29 to 3970 μmol L−1). Calculated pCO2 were > 10% higher than measured pCO2 in 60% of the samples (with a median overestimation of calculated pCO2 compared to measured pCO2 of 2560 ppmv) and were > 100% higher in the 25% most organic-rich and acidic samples (with a median overestimation of 9080 ppmv). We suggest these large overestimations of calculated pCO2 with respect to measured pCO2 are due to the combination of two cumulative effects: (1) a more significant contribution of organic acids anions to TA in waters with low carbonate alkalinity and high DOC concentrations; (2) a lower buffering capacity of the carbonate system at low pH, that increases the sensitivity of calculated pCO2 to TA in acidic and organic-rich waters. We recommend that regional studies on pCO2 should not be based on pH and TA data only, and that direct measurements of pCO2 should become the primary method in inland waters in general, and in particular in acidic, poorly buffered, freshwaters.

No abstract

Rivers transport sediment and carbon (C) from the continents to the ocean, whereby the magnitude and timing of these fluxes depend on the hydrological regime. We studied the sediment and carbon dynamics of a tropical river system at two sites along the lower Tana River (Kenya), separated by a 385 km stretch characterized by extensive floodplains, to understand how the river regime affects within-river C processing as well as the C exchange between floodplain and river. Sampling took place during three different wet seasons (2012-2014), with extensive flooding during one of the campaigns. We measured the suspended sediment concentration, the concentration and stable isotope signature of three different carbon species (particulate and dissolved organic carbon, POC and DOC, and dissolved inorganic carbon, DIC) and other auxiliary parameters. During non-flooded conditions, the total C flux was dominated by POC (57-72%) and there was a downstream decrease of the total C flux. DIC was dominating during the flooded season (56-67%) and the flux of DIC and DOC coming from the inundated floodplains resulted in a downstream increase of the total carbon flux. Our data allowed us to construct a conceptual framework for the C dynamics in river systems, whereby nine major fluxes were identified. The application of this framework highlighted the dominance of POC during non-flooded conditions and the significant CO 2 emissions during the flooded season. Furthermore, it identified the exchange of POC with the floodplain as an important factor to close the C budget of the river.