Modeling inorganic carbon dynamics in the Seine River continuum in France

Abstract. Inland waters have been recognized as an active component of the carbon cycle where transformations and transports are associated with carbon dioxide (CO2) outgassing. We propose a modeling approach by formalizing an inorganic carbon module integrated into the biogeochemical model, pyNuts-Riverstrahler, to estimate the carbon fate in the aquatic continuum. Our approach was implemented on the human-impacted Seine River (France) taking into account point sources (including the largest wastewater treatment plant in Europe, reaching a treatment capacity of 6 106 inhab eq), and diffuse constraints to the model. Both sources were characterized by field measurements in groundwater and in wastewater treatment plants, and by existing databases. In average, we calculated DIC concentrations from 25 to 92 mgC L−1 depending of the aquifers while in WWTP effluents our measurements of DIC averaged 70 mgC L−1. On the period studied (2010–2013), yearly averaged simulated CO2 emissions from the hydrosystem were estimated at 364 ± 99 Gg C yr−1. Simulations of dissolved inorganic carbon, total alkalinity, pH and CO2 concentrations showed good agreement with observations, and seasonal variability could be reproduced. Metabolism in the Seine hydrographic network highlighted the importance of benthic activities in small head streams while planktonic activities were mainly observed downstream in larger rivers. In contrast to the 1990s, the net ecosystem productivity remained negative throughout all the years and at every place within the river network, highlighting the heterotrophy of the basin. In parallel, CO2 supersaturation with respect to atmospheric concentrations of the basin was shown. Outgassing was the most important in lower order streams while peaks were simulated downstream of the major wastewater treatment effluent.