Eddy-covariance fluxes of CO2, CH4 and N2O in a drained peatland forest after clear-cutting

• Main Authors: Tikkasalo, Olli-Pekka, Peltola, Olli, Alekseychik, Pavel, Heikkinen, Juha, Launiainen, Samuli, Lehtonen, Aleksi, Li, Qian, Martínez-García, Eduardo, Peltoniemi, Mikko, Salovaara, Petri, Tuominen, Ville, Mäkipää, Raisa
• Other Authors: Ilmatieteen laitos, Finnish Meteorological Institute, orcid:0000-0002-0612-128X
• Format: Article in Journal/Newspaper
• Language: English
• Published: Copernicus Publications 2025
• Subjects: emissions; carbon dioxide; greenhouse gases; forests; peatlands; peatland forests; blogs; climate changes; atmosphere (earth); machine learning; päästöt; hiilidioksidi; kasvihuonekaasut; metsät; turvemaat; suometsät; suot; ilmastonmuutokset; ilmakehä; koneoppiminen; Fennoscandia
• Online Access: http://hdl.handle.net/10138/593712

Rotation forestry based on clear-cut harvesting, site preparation, planting and intermediate thinnings is currently the dominant management approach in Fennoscandia. However, understanding of the greenhouse gas (GHG) emissions following clear-cutting remains limited, particularly in drained peatland forests. In this study, we report eddy-covariance-based (EC-based) net emissions of carbon dioxide (CO2), methane (CH4) and nitrous oxide (N2O) from a fertile drained boreal peatland forest 1 year after wood harvest. Our results show that, at an annual scale, the site was a net CO2 source. The CO2 emissions dominate the total annual GHG balance (23.3 t CO2 eq. ha−1 yr−1, 22.4–24.1 t CO2 eq. ha−1 yr−1, depending on the EC gap-filling method; 82.0 % of the total), while the role of N2O emissions (5.0 t CO2 eq. ha−1 yr−1, 4.9–5.1 t CO2 eq. ha−1 yr−1; 17.6 %) was also significant. The site was a weak CH4 source (0.1 t CO2 eq. ha−1 yr−1, 0.1–0.1 t CO2 eq. ha−1 yr−1; 0.4 %). A statistical model was developed to estimate surface-type-specific CH4 and N2O emissions. The model was based on the air temperature, soil moisture and contribution of specific surface types within the EC flux footprint. The surface types were classified using unoccupied aerial vehicle (UAV) spectral imaging and machine learning. Based on the statistical models, the highest surface-type-specific CH4 emissions occurred from plant-covered ditches and exposed peat, while the surfaces dominated by living trees, dead wood, litter and exposed peat were the main contributors to N2O emissions. Our study provides new insights into how CH4 and N2O fluxes are affected by surface-type variation across clear-cutting areas in forested boreal peatlands. Our findings highlight the need to integrate surface-type-specific flux modelling, EC-based data and chamber-based flux measurements to comprehend the GHG emissions following clear-cutting and regeneration. The results also strengthen the accumulated evidence that recently clear-cut peatland forests are significant ...