Optimizing Fire Observation Towers

Fire monitoring towers are an effective way to mitigate the spread of forest fires. This will become a more prevalent issue as the effects of climate change worsen and human development increases. This project was able to find an optimal solution of fire monitoring towers for the Chebucto Peninsula using the Location Set Coverage Problem (LSCP) location allocation model. Human development is growing, and with it the risk of human-caused fires, so human settlements were taken into account by the use of a registry of Canadian civic addresses.

Suitable sites for towers were identified using suitability analysis. Suitability was determined by distance to roads, slope, land cover, and topography (peaks). These factors were created from shape files and digital elevation models (DEM) using spatial and 3D analysis tools in ArcGIS Pro. The LSCP will find the least amount of fire towers to cover all the demand zones. The demand zones were created using two different methods. Each community in the peninsula had a mean center based on civic addresses, and polygon centers based on the community borders. The model was run separately using both centers. The LSCP used a coverage matrix to solve the allocation problem. Coverage was determined through testing which demand centers were covered by an eight kilometer buffer placed around each potential fire tower site. The optimal solution used the least amount of towers to cover all demand centers.

The optimal solution for both models was six towers. The mean center model covered much more of the civic addresses with only a slight decrease in area covered. The optimal solution proved to be above 97% accuracy in covering civic addresses and above 90% in covering area. For small study areas such as in this project, an LSCP can find the optimal tower sites to cover all of the study area. For further application, larger areas should consider using a maximum covering location problem (MCLP) in order to maximize coverage whilst conforming to a budget.