Develop a multistage stochastic program with recourse for scheduling prescribed burning based fuel treatments with consideration of future wildland fires and fire suppressions
Date
2015
Authors
Nguyen, Dung Tuan, author
Wei, Yu, advisor
Rideout, Douglas B., committee member
Reich, Robin M., committee member
Leisz, Stephen J., committee member
Journal Title
Journal ISSN
Volume Title
Abstract
In this study, I present a multistage stochastic linear program with recourse for scheduling prescribed burning based fuel treatments under the influences of random future windland fires and fire suppressions across multiple planning periods. Prescribed burning decreases future wildfire’s spread rate and intensity. Future wildfire uncertainties are characterized by sequences of independent and identical (i.i.d.) fire samples across the entire planning horizon. Each simulated sample fire ignites at a random location and spreads for a random duration under the influence of a randomly selected wind direction and speed. This stochastic program explicitly addresses the spatial and temporal relationships between fire behavior, prescribed burning, and suppression in multiple fire-planning periods. It uses sample average approximation and minimizes the sum of average discounted management cost plus average discounted fire loss across a planning horizon. Test cases are designed to examine fire-and-management situations on an artificial forested landscape, and are focused on selecting good quality first period prescribed burning locations. Results provide a wide range of optimal solutions for allocating the first period prescribed burning to handle risks from future wildfires.