What determines area burned in large landscapes? Insights from a decade of comparative landscape-fire modelling

Understanding what determines area burned in large landscapes is critical for informing wildland fire management in fire-prone environments and for representing fire activity in Dynamic Global Vegetation Models. For the past ten years, a group of landscape-fire modellers have been exploring the relative influence of key determinants of area burned in temperate and boreal forested landscapes using a suite of landscape-scale fire models from around the world. Three major simulation comparisons have been completed thus far, each incorporating five models and examining several different factors including fuel management, weather variability, climate change and landscape attributes. Comparison one found that variation in weather and climate was more important in determining simulated area burned than variation in fuel-load pattern and terrain in the majority of models. Comparison two demonstrated that year-to-year variation in weather and variation in efforts to reduce ignitions had a greater effect on area burned than fuel management across the range of modelled ecosystems. Finally, the third comparison explored the relative importance of vegetation dynamics, along with variation in climate and weather, in determining area burned. In this case, variation in climate was found to be a critical factor in most models - although vegetation succession was also important in some cases - emphasising the critical role that future climate change will have in controlling wildland fire. Similar to the concept of the "wisdom of crowds", the multiple-model perspective gained through these efforts provides considerable insights into examining commonalities across a range of ecosystems and simulation models