Abstract. Peatlands, which contain large carbon stocks that must be accounted for in the global carbon budget, are poorly represented in many earth system models. We integrated peatlands into the coupled Canadian Land Surface Scheme (CLASS) and the Canadian Terrestrial Ecosystem Model (CTEM), which together simulate the fluxes of water, energy, and CO 2 at the land surface-atmosphere boundary in the family of Canadian Earth system models (CanESMs). New components and algorithms were added to represent the unique features of peatlands, such as their characteristic ground floor vegetation (mosses), the slow decomposition of carbon in the water-logged soils and the interaction between the water, energy, and carbon cycles. This paper presents the modifications introduced into the CLASS-CTEM modelling framework together with site-level evaluations of the model performance for simulated water, energy and carbon fluxes at eight different peatland sites. The simulated daily gross primary production (GPP) and ecosystem respiration are well correlated with observations, with values of the Pearson correlation coefficient higher than 0.8 and 0.75 respectively. The simulated mean annual net ecosystem production at the eight test sites is 87 g C m −2 yr −1 , which is 22 g C m −2 yr −1 higher than the observed annual mean. The general peatland model compares well with other site-level and regional-level models for peatlands, and is able to represent bogs and fens under a range of climatic and geographical conditions.

Elevated nitrogen deposition and climate change alter the vegetation communities and carbon (C) and nitrogen (N) cycling in peatlands. To address this issue we developed a new process-oriented biogeochemical model (PEATBOG) for analyzing coupled carbon and nitrogen dynamics in northern peatlands. The model consists of four submodels, which simulate: (1) daily water table depth and depth profiles of soil moisture, temperature and oxygen levels; (2) competition among three plants functional types (PFTs), production and litter production of plants; (3) decomposition of peat; and (4) production, consumption, diffusion and export of dissolved C and N species in soil water. The model is novel in the integration of the C and N cycles, the explicit spatial resolution belowground, the consistent conceptualization of movement of water and solutes, the incorporation of stoichiometric controls on elemental fluxes and a consistent conceptualization of C and N reactivity in vegetation and soil organic matter. The model was evaluated for the Mer Bleue Bog, near Ottawa, Ontario, with regards to simulation of soil moisture and temperature and the most important processes in the C and N cycles. Model sensitivity was tested for nitrogen input, precipitation, and temperature, and the choices of the most uncertain parameters were justified. A simulation of nitrogen deposition over 40 yr demonstrates the advantages of the PEATBOG model in tracking biogeochemical effects and vegetation change in the ecosystem

Changes in vegetation composition due to nitrogen deposition may influence the role of peatlands in the carbon cycle. The impact of nitrogen (N) on vegetation composition of the Mer Bleue Bog (Ontario) and the underlying mechanisms were studied using a coupled carbon (C) and nitrogen wetland model (PEATBOG). The model was applied to data from a long-term nutrient fertilization experiment and the results were compared with the observed C and N pools in plants, peat, and soil water after 8 years of fertilization with 1.6, 3.2, and 6.4 gN m -2 y -1 and additional phosphorus and potassium. The evaluated model was employed to simulate the vegetation dynamics in the peatland fertilized with different loads and concentrations of N. The model suggested a shift in plant functional types from moss-shrub dominated to graminoid dominated with increasing N load. Suppression of mosses by N fertilization was mitigated by daily deposition of N at a low concentration. Mosses became extinct in the simulations only when the concentration and load of N deposition were high, corroborating empirical results from the fertilization experiment. Tracking the deposited N in the system, the model indicated that it was primarily sequestered in the peat. Dissolved N concentrations in peat and export from the bog increased with N load and decreased with N concentration, leading to higher N uptake by vascular plants. The results further suggested that the observed detrimental effect of N on mosses was most likely due to toxicity caused by N uptake exceeding N assimilation. Exposure of mosses to high N concentrations in precipitation and evaporating water on moss tissues is thus a factor that needs attention when considering critical loads of N deposition on peatlands.

Abstract. Peatlands, which contain large carbon stocks that must be accounted for in the global carbon budget, are poorly represented in many earth system models. We integrated peatlands into the coupled Canadian Land Surface Scheme (CLASS) and the Canadian Terrestrial Ecosystem Model (CTEM), which together simulate the fluxes of water, energy and CO2 at the land surface–atmosphere boundary in the family of Canadian Earth System Models (CanESMs). New components and algorithms were added to represent the unique features of peatlands, such as their characteristic ground floor vegetation (mosses), the slow decomposition of carbon in the water-logged soils and the interaction between the water, energy and carbon cycles. This paper presents the modifications introduced into the CLASS-CTEM modelling framework together with site-level evaluations of the model performance for simulated water, energy and carbon fluxes at eight different peatland sites. The simulated daily gross primary production and ecosystem respiration are well correlated with observations, with values of the Pearson correlation coefficient higher than 0.8 and 0.75 respectively. The simulated mean annual net ecosystem production at the eight test sites is 87 g C m−2 yr−1, which is 22 g C m−2 yr−1 higher than the observed annual mean. The general peatland model compares well with other site-level and regional-level models for peatlands, and is able to represent bogs and fens under a range of climatic and geographical conditions.