Applications of Forest-DNDC to simulate daily carbon fluxes on the natural and inundated terrestrial ecosystems of Canada

Forest-DNDC is a process-based biogeochemistry model used to quantify carbon (C) and nitrogen (N) fluxes from both upland and wetland forests. Despite of its usefulness due to low input and parameter requirements, and intermediate modelling scale (day time and landscape), no study has attempted to predict C fluxes on Canadian forests and peatlands. In this study, we evaluate and modify Forest-DNDC for simulating ecosystem C fluxes (Gross Primary Production, Ecosystem Respiration, and Net Ecosystem Production) for mature black spruce forests and ombrotrophic bogs. We also test the performance of a “flooded” version of Forest-DNDC in an attempt to estimate the change in ecosystem C exchanges due to the inundation of forests and peatlands for the creation of hydro-electric reservoirs. The evaluations were conducted for a forest site (NOBS in Manitoba) and a bog site (Mer Bleue bog in Ontario) using 12 and 8 years flux records, respectively. The “flooded” Forest-DNDC behaviour was examined by comparing the model outputs with field observations from a recently impounded boreal forests and peatlands (Eastmain-1 reservoir in Québec). The simulations for the natural ecosystems generated reasonable estimations about daily C fluxes. Over the study periods, mean of daily GPP, ER, and NEP were resulted in 19.3, 17.5, and 1.8 kg C ha-1 d-1 at the black spruce site and 17.0, 14.0, and 3.0 kg C ha-1 d-1 at the bog site, and agreement index (d) between the modelled and measured fluxes in GPP, ER, and NEP were 0.96, 0.96, and 0.75, and 0.94, 0.95, and 0.76 for the forest and peatland, respectively. Daily NEP from “flooded” Forest-DNDC ranged between 0 and -28.6 kg C ha-1 d-1 for the flooded forest and 0 and -9.0 kg C ha-1 d-1 for the flooded peatland, but the comparisons with measured fluxes in both the areas had less agreement (e.g. d \u3c 0.70). However, it is not clear how to directly interpret the observations of the reservoir (i.e. atmosphere exchanges), causing uncertainties in comparison between the flux measurements and model outputs