Measurements of the carbon, energy and water balance in an Italian alpine peatland and simulating the vegetation and carbon dynamics in three different peatlands

The stored carbon in the soil of peatlands is a result of a long-term imbalance between the decomposition and the biomass production. Peatlands occur all over the world but are mainly located in the northern hemisphere. Up to now, peatlands in the Alps have not received much attention, because they are not the dominant ecosystem. In this study, we investigated the carbon, water and energy balance of a small-scale fen in the Italian Alps over four years (2012-2015). During the four-year period, the carbon fluxes were measured by an eddy covariance system. Over this period, the peatland was a carbon source based on CO2 emissions (net ecosystem exchange (NEE): 103.5, 262.9, 175.7, 1807.7 g C-CO2 m-2 yr-1). During a 10-month period, December 2013- September 2014 a methane analyser was installed at the peatland, which measured small methane (CH4) fluxes (3.2 g C-CH4 m-2). Dissolved Organic Carbon (DOC) plays an important role in the carbon balance of ecosystems. To assess the DOC fluxes of the peatland the hydrology of the peatland was modelled with the model GEOtop. GEOtop is a process-based distributed model of the water and energy budget and was applied for four years. The modelled energy fluxes are comparable to the fluxes measured by the on-site eddy covariance tower. The model was able to simulate the volumetric water content temperature accurately over the four years. During snow cover, the model had difficulties simulating the soil temperature due to insulation by the snow. In 2014-2015, the DOC concentrations were measured. The modelled water cycle was used to quantify the loss of dissolved organic carbon (DOC) and to calculate the carbon balance of the peatland. Based on the DOC measurements and the modelled discharge, an extrapolation of the DOC export was made, which resulted in an average loss of DOC 7.7, 12.3, 13.8, 8.0 g C m-2 yr-1 over the four years. The combination of the DOC, CO2 and CH4 fluxes indicate that the peatland is acting as a carbon source for all four subsequent years with a carbon balance of 112.3, 273.8, 190.8 and 95.3 g C m-2 yr-1 for 2012, 2013, 2014 and 2015 respectively. To study the functioning of three different peatlands over the world the NUCOM-Bog (NUtrient cycling and COmpetition Model) was applied. Bayesian techniques have been used to calibrate the model for each site specific and for a multi-site calibration. The combination of site-specific and multi-site calibration has the potential to gain insights in the functioning of different peatlands and to identify how the vegetation of peatlands will change over time in respect to climate change. The marginal parameter uncertainty of the multi-site calibration compared to the site-specific calibration, indicate that with the multi-site calibration the parameter uncertainty can be reduced by using information from multiple sites