Peatland fire occurrence across Sweden from 2001 to 2023 along drainage, land-cover, and hydroclimatic gradients

Abstract Background Peatlands combine large belowground organic fuel stores with high moisture conditions that often limit combustion, making peatland fire occurrence difficult to interpret from fuel availability or wetness alone. In Sweden, extensive drainage has altered peatland hydrology and vegetation, but it remains unclear whether mapped drainage intensity explains where peatland fires occur. We conducted an exploratory national geospatial analysis of fire occurrence from 2001 to 2023 by intersecting two satellite-derived fire datasets (mapped burned-area perimeters and thermally anomalous active-fire footprints) with peatland extent, mapped land cover, hydroclimatic indicators, ditch density, and road density within a Swedish analysis domain defined by forestland and open-wetland land-cover classes. Results Peatlands were not underrepresented among fire-affected areas. Nationally, peatlands accounted for 27.4% of mapped burned area and 35.5% of thermal-anomaly footprint area, compared with 24.8% of areas without detected fire. In southern Sweden, peatlands were overrepresented in both datasets. During 2001–2023, 71.3 km 2 of peatland was mapped as burned area and 439.0 km 2 as thermal-anomaly footprint area. Fire activity was strongly episodic, with most burned peatland area concentrated in 2014 and 2018, and these high-fire years coincided with dry or hot-dry hydroclimatic conditions. Fire-affected peatland areas had a lower forested-peatland share than peatland areas without detected fire. Mapped ditch density was generally not higher in fire-affected peatland areas; where differences occurred, they were weakly negative. Road and ditch densities covaried, suggesting that ditch density partly reflects broader land-use and access gradients. Conclusion Peatland fire occurrence in Sweden appears episodic and hydroclimatically conditioned, but it is not explained by mapped ditch density alone. These results do not show that drainage or rewetting is irrelevant to peatland fire processes. Rather, they indicate that mapped ditch density alone is not a reliable predictor of where fire occurs on Swedish peatlands. Future work should link event-level fire data with water-table observations, burn severity, peat soil properties, fuel moisture, and a more mechanistic drainage-intensity index.