Optimizing fuel treatment plans to reduce burn probability: Importance of navigating context, priorities and trade-offs

View article.

Given the large size of landscapes, limited management budgets and diverse (sometimes competing) objectives, it can be extremely difficult to know where and how fuel treatments are best undertaken to reduce wildfire risks. While optimization algorithms can help to navigate such complex decisions, the computational cost of applying simulation-based models for predicting wildfire risk has prevented us from using optimization to guide decision-making. To implement optimization by leveraging ‘metamodelling’ approaches that can efficiently estimate the burn probability outputs of simulation models. We use a simulation-optimization approach that links a burn probability (BP) metamodel with the multi-objective optimization algorithm NSGA-II, to develop fuel treatment plans that optimization the trade-offs between different risk reduction objectives and the area treated (AT) by fuel treatment plans in a South Australian case study area. Optimization improves the reduction in BP per area managed by at least 81–284% when compared with existing approaches in our study area. Optimization develops highly effective fuel treatment plans that balance trade-offs between different BP-based objectives and/or levels of resources available for management. Optimization can improve strategic landscape management and offers the potential to help communities better achieve their risk reduction objectives.