An optimization model for aggregation of prescribed burn units

Abstract

Fire is a natural component of many terrestrial ecosystems; however, uncontrolled intense wildfires can cause loss of human life and destruction of natural resources. Prescribed burning is a management activity undertaken for the purposes of both wildfire hazard reduction and the preservation of fire-adapted ecosystems. Achievement of prescribed burn targets is made difficult by operational constraints such as limited personnel, equipment and suitable weather for undertaking burning. Prescribed burn program implementation may benefit from efficiencies gained in undertaking larger burns. Here, we present a mixed-integer programming formulation for aggregation of prescribed burn units. The model minimizes the total prescribed burn perimeter requiring management, by aggregating existing ‘fundamental’ burn units into larger compact and contiguous units or ‘clusters’. This problem is a special case of the ‘supervised regionalization’ or the ‘ p -regions problem’. The model’s functionality is demonstrated on a test landscape and a number of extensions and implementation issues are discussed.