Collaborative Research: Fuel Treatment Planning Optimization for Wildfire Management Grant uri icon

abstract

  • This award contributes to the welfare of the nation by addressing important challenges in reducing economic losses due to wildfires. Increased development and urbanization in fire-prone areas, coupled with evolving climate changes, have significantly increased the vulnerability of human communities and ecosystems to wildfires. Even with large expenditures for fire suppression efforts, annual economic losses, as well as loss of human life, due to wildfires remain high. This award supports research efforts to reduce wildfire activity by developing cost-effective methods to reduce risk through fuel treatment. Fuel treatment involves removing vegetation (i.e., fuel) from a landscape to reduce the potential and severity of large-scale fires. Fuel treatment, which forms a first line of wildfire defense, may include any combination of controlled burning, grazing, and various types of mechanical thinning. This award will contribute to better understanding of what types of fuel treatment options and associated decision-making strategies are more appropriate for particular fire-prone regions. This project will involve both graduate and undergraduate students as well as development of courses that expose students at all levels to quantitative methods to address large-scale societal problems. This award will support research into new sequential mixed-integer optimization methods to determine the appropriate location, timing and type of fuel treatments over multiple seasons in order to minimize the expected losses from wildfires in a region. The optimization framework will involve formulating and solving non-linear mathematical programming models for fuel accumulation and reduction under resource constraints. To take into account the inherent uncertainties with respect to fire ignition, the approach employs robust optimization techniques. The project will investigate analytical results that describe important structural properties of the models and will develop specialized numerical algorithms to solve realistically-sized instances of the problem. The algorithms will leverage and extend modern techniques at the intersection of robust and combinatorial optimization. The models will be calibrated and validated using historical data from the Texas A&M Forest Service, a state agency charged with overseeing forest management in the state of Texas. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

date/time interval

  • 2020 - 2023