Impact of future wildfire spread on forest carbon seqeustrartion: A case study of South Korea

As the frequency and intensity of megafires continue to increase, projecting and managing future wildfire occurrence is becoming increasingly important. Wildfires damage forests as major carbon sinks, thereby posing substantial uncertainties to carbon uptake. Furthermore, climate change is expected to intensify wildfire spread, leading to greater overall damage. Therefore, future wildfire management requires assessments that incorporate wildfire spread patterns under changing climate conditions. However, most existing studies have focused primarily on estimating wildfire ignition locations, with limited consideration of wildfire spread under climate change. In this study, we trained wildfire spread patterns within wildfire events using a combination of remote sensing data and field survey observations. Based on these patterns, we estimated future wildfire occurrence and subsequent spread using a dynamic Bayesian network framework. We further analyzed changes in forest carbon uptake resulting from wildfire occurrence and spread. Our results indicate that simulations accounting for both wildfire occurrence and spread result in greater total burned area by 2050 compared to simulations considering wildfire occurrence alone. In particular, repeated wildfire occurrences and their spatial propagation expanded the cumulative damaged areas. When wildfire spread was included, forest carbon uptake declined more sharply, with some regions projected to shift from net carbon sinks to net carbon sources. These findings demonstrate that excluding wildfire spread leads to an underestimation of wildfire damage and associated carbon sequestration.