Network separation modeling and quantum computing for developing wildfire fuelbreak strategy.

Area of Science:

• Wildfire Management
• Computational Optimization
• Quantum Computing Applications

Background:

• Traditional fuelbreak placement relies on experienced fire managers, often lacking scalability for large areas.
• Existing formal placement strategies face challenges in efficiently managing extensive landscapes.

Purpose of the Study:

• To develop and present an efficient fuelbreak placement strategy using equal graph partitioning and quantum computing.
• To compare the performance of this novel strategy against traditional methods and other solvers.

Main Methods:

• Formulated fuelbreak placement as a quadratic constrained binary optimization problem.
• Utilized D-Wave's hybrid quantum optimization tool for efficient placement determination.
• Compared quantum solver performance with traditional SCIP and CPLEX solvers on a subsection.

Main Results:

• The quantum computing approach determined placements in seconds, outperforming traditional methods in efficiency.
• Two alternative placements showed improved land separation equality (2.9% and 12.4%) with varying acreage cleared.
• D-Wave's hybrid solver demonstrated competitive speed, with potential for greater speed-up on larger-scale problems.

Conclusions:

• Equal graph partitioning is an effective strategy for optimizing fuelbreak placement.
• D-Wave's hybrid quantum solvers show significant potential for enhancing wildfire management efficiency and effectiveness.