The Canadian model for peatlands (CaMP): A peatland carbon model for national greenhouse gas reporting

•A model for GHG emission and removal reporting for Canadian peatlands was built.•The CaMP v2.0 was tested against observed CO2 and CH4 flux data.•Results validate the CaMP structure for large-scale emission estimates.•Data gaps in peatland research needed for further model development are outlined. A model framework for national greenhouse gas emission and removal estimation for Canadian peatlands (CaMP v2.0) was developed and tested. It provides a module that can work alongside the upland forest Generic Carbon Budget Model (GCBM) developed to eventually replace the Carbon Budget Model of the Canadian Forest Sector (CBM-CFS3) as the core model in Canada's National Forest Carbon Monitoring, Accounting and Reporting System. The CaMP (v2.0) provides a simple model foundation that can be applied nationally for 11 different peatland categories. It tracks the growth, turnover and decay in annual time steps of different vegetation components (foliage, branches, stems, and roots of trees, shrubs, sedges and mosses). It uses a Q10 relationship to model peat C pool decomposition as a function of mean annual temperature, and models methane flux response to deviations in annual water table depth. The CaMP takes a simple approach to modeling hydrology for large spatial scales by using the nationally-available Canadian Fire Weather Index Drought Code to predict long-term and annual water table depth. The CaMP (v2.0) provides the framework needed to model disturbances but only includes wildfire in this version. Model behavior and sensitivity were assessed, and evaluated against observed flux data. Results suggest that the CaMP (v2.0) provides an appropriate structure for large spatial- and temporal-scale estimation of emissions, owing to the model behaving as expected relative to shifts in environmental variables, and to reasonably small mean observed to modeled residuals. Methane was overestimated by the model on average by 6 g C ha−1y − 1 (n = 53 years of data across 11 peatland sites), and by 8 g C ha−1y − 1 when weighted by site location (n = 12 sites, ≥ 3 years of data per site). The model overestimated net ecosystem exchange (NEE) by 20 g C ha−1y − 1 (n = 36 years of data across 12 peatland sites), and by 2 g C ha−1y − 1 when weighted by site location (n = 11 sites, ≥ 3 years of data per site), and results demonstrate that inter-site variation is greater than temporal variation across NEE measures. Several aspects were identified as requiring further work to incre