Terrain influence landscape patterns of burn severity in subtropical forests of southern China

Context Understanding the landscape patterns of burn severity is vital for managing fire-prone ecosystems. Relatively limited research has been done about fire and burn severity patterns in subtropical forests. Methods: Using Landsat 8 OLI remote sensing imagery, this paper spatially mapped the burn severity of 27 forest fires in the subtropical broadleaved evergreen forest in Southern China from 2017–2021. The landscape pattern of patches with different burn severity was quantified using landscape indices. In addition, factors influencing the patterns of burn severity across the landscape were determined using the Geodetector model. Results: Burn severity of patches varied significantly. High burn severity was common in forest patches with low fragmentation, low patch density, and regular shape. In contrast, moderate and low burn severity was prevalent in patches with smaller patch size, high patch density, and complex shapes. Extensively burned forest patches were located at higher elevations, while more fragmented patches were located in gently sloping areas. Topographic factors were the most significant factors influencing variances in burn severity across the forest patches, followed by climatic conditions. A detailed understanding of burn severity patterns and driving factors in a landscape can help develop sustainable forest management and restoration strategies after severe forest fire disturbances.