Integrating aquatic and terrestrial biogeochemical model to predict effects of reservoir creation on CO₂ emissions

Abstract. There is considerable debate on the role of hydroelectric reservoirs for the emission of CO₂ and other greenhouse gases. To quantify CO₂ emissions from a newly created reservoir that was formed by flooding the boreal landscape we developed a daily time-step reservoir model by integrating a terrestrial and an aquatic ecosystem model. We calibrated the model using the measurements of dissolved organic and inorganic carbon (C) in a ~ 600 km² boreal hydroelectric reservoir, Eastmain-1, in northern Quebec, Canada. A major constraint we dealt with is the dearth of basic environmental data for the Boreal region so we took a parsimonious approach for required inputs. We then evaluated the model performance against observed CO₂ fluxes data from an eddy covariance tower in the middle of the EM-1 reservoir for the period from 2006 to 2012 and compared internal variables such as water column respiration, chlorophyll-a concentration, and sedimentation rate to measurements from field campaigns during 2006–2008. The model predicted the seasonal and inter-annual variability of CO₂ emissions reasonably well compared to the observations. Discrepancies between simulation results and observations usually occurred near ice-off dates when there was large amount of dissolved CO₂ under ice-cover. We applied the model to assess the effects of reservoir creation on C dynamics over the estimated “engineering” reservoir lifetime (i.e., 100 years). We found that the reservoir acts as a net C source over its lifetime and simulated CO₂ fluxes were 204 g C m⁻² yr⁻¹ in the first year after flooding, steeply declined in the first three years, and then steadily decreased to ~110 g C m⁻² yr⁻¹ with increasing reservoir age. Sensitivity analyses revealed that the amount of terrestrial organic C flooded and oxygen effects can positively enhance benthic respiration and CO₂ fluxes across air–water interface, but the effects on CO₂ emissions were not significant. Higher temperatures dramatically stimulate CO₂ emissions by enhancing CO₂ production in both the water column and the sediment, and extending the duration of the open water period over which emissions can occur. Changing wind speeds had large uncertainties on annual CO₂ emissions, given that wind speeds not only affect the gas transfer rate but also the open water period by affecting the surface energy balance. The model is useful for the estimation of CO₂ emissions from reservoirs to the atmosphere and could be used to assist the hydro-power industry and others interested in emissions evaluate the role of boreal reservoirs as sources of greenhouse gas emissions.