Temporal variability of air-sea CO2 exchange in a low-emission estuary

There is the need for further study of whether global estimates of air-sea CO2 exchange in estuarine systems capture the relevant temporal variability and, as such, the temporal variability of bulk parameterized and directly measured CO2 fluxes was investigated in the Danish estuary, Roskilde Fjord. The air-sea CO2 fluxes showed large temporal variability across seasons and between days and that more than 30% of the net CO2 emission in 2013 was a result of two large fall and winter storms. The diurnal variability of Delta pCO(2) was up to 400 during summer changing the estuary from a source to a sink of CO2 within the day. Across seasons the system was suggested to change from a sink of atmospheric CO2 during spring to near neutral during summer and later to a source of atmospheric CO2 during fall. Results indicated that Roskilde Fjord was an annual low-emission estuary, with an estimated bulk parameterized release of 3.9 8.7 mol CO2 m(-2) y(-1) during 2012-2013. It was suggested that the production-respiration balance leading to the low annual emission in Roskilde Fjord, was caused by the shallow depth, long residence time and high water quality in the estuary. In the data analysis the eddy covariance CO2 flux samples were filtered according to the H2O-CO2 cross-sensitivity assessment suggested by Landwehr et al. (2014). This filtering reduced episodes of contradicting directions between measured and bulk parameterized air-sea CO2 exchanges and changed the net air-sea CO2 exchange from an uptake to a release. The CO2 gas transfer velocity was calculated from directly measured CO2 fluxes and Delta pCO(2) and agreed to previous observations and parameterizations. (C) 2016 Published by Elsevier Ltd.