Insights into CO2 simulations from the Irish Blackwater peatland using ECOSSE model

Non-degraded peatlands are known to be important carbon sink; however, if they are exposed to anthropogenic changes they can act as carbon source. This study forms a part of the larger AUGER project (http://www.ucd.ie/auger). It uses the ECOSSE process-based model to predict CO2 emissions [heterotrophic respiration (Rh)] associated with different peatland management (Smith et al., 2010). The work aims to provide preliminary insights into CO2 modelling procedures for drained and rewetted sites from Blackwater, the former Irish raised bog. After drainage in 1950’s (due to peat-extraction) and cessation of draining in 1999, the landscape developed drained ‘Bare Peat’ (BP), and rewetted ‘Reeds’ (R) and ‘Sedges’ (S) sites (Renou-Wilson et al., 2019). Modelling of CO2 from these sites was done using ECOSSE-v.6.2b model (‘site-specific’ mode) with water-table (WT) module (Smith et al., 2010), and default peatland vegetation parameters. The other model-input parameters (including soil respiration, WT and other soil parameters) were obtained from measurements reported in Renou-Wilson et al. (2019). Simulations on drained BP site were run starting from 1950 and on rewetted R and S sites starting from 1999 (which is the year of cessation of drainage). The climate data inputs (2010-2017) were obtained from ICHEC (EPA_Climate-WRF, 2019). The long-term average climate data for model spin-up were obtained from Met Éireann (2012) with potential evapotranspiration estimated by Thornthwaite (1948) method. Daily ecosystem respiration (Reco) data for May/June 2011 to Aug 2011 obtained from raw CO2 flux measurements (Renou-Wilson et al., 2019) were used. For vegetated sites Rh was estimated from Reco using method explained in Abdalla et al. (2014). Daily CO2 simulations were compared to Reco for BP site (r2 =0.20) and to Rh for R site (r2 = 0.35) and S site (r2 = 0.55). The preliminary results showed some underestimation of simulated CO2 indicating the need for further modelling refinements for satisfactory results. The results from BP site further indicated on the importance of including long-term drainage period (i.e. from 1950 on) because avoiding this step resulted in a large overestimation of predicted CO2.