Modeling fire potential polygon networks for fire suppression 🔥

Authors: Minho Kim, Marc Castellnou, Marta C. Gonzalez

Abstract

The Catalan Fire Service pioneered an innovative approach to guide proactive fire management, manually drawing polygons of fire potential and connections using expected fire behavior on the landscape. However, this manual drawing process is time-consuming, subjective, and relies heavily on expert judgment. To address this limitation, we introduce a method to automatically generate fire potential polygons and connect them into a weighted network based on fire behavior. To this end, we use a cellular automata-based 2D fire growth model under dynamic weather conditions to simulate fire behavior. We calculate the elapsed time computed from fire spread simulations and propose a method inspired by basin delineation tools from hydrology to segment polygons on the landscape. These polygons are subsequently connected into a network, using a weighted rate of spread metric to characterize the connections. Our method produces automated polygons, networks of the polygons connected by various fire behavior metrics, major fire pathways, and network visualizations of simulated scenarios.

We validate our approach on two wildfire case studies in Catalonia (Spain) during the 2024 fire season. Our approach is applied during the initial attack of two fires that had the potential to grow large with a high risk of invoking catastrophic damage. In the wind-driven Ciutadilla fire (News Article), the polygon networks were able to identify high-risk polygons, connections, and critical fire pathways that aligned with real operations on the ground. In the Vilanova fire (News Article), we present how our modeling approach can integrate various suppression tactics and a prescribed burn to update the networks and assess the amount of time gained through the suppression.

Diagram of proposed methodology using fire spread simulations and hydrology-inspired modeling of fire potential polygons to build fire potential networks for decision support.

Highlights⚡

• Fire potential polygons are an innovative approach for wildfire risk management🔥.
• Generated polygons using fire spread simulations🔥 and hydrology-based basin delineation💧 tools.
• Constructed networks of polygons that prioritize suppression efforts🚒 and enhance decision-making💡.
• Evaluated our method in real-time initial attack operations🧑‍🚒 for two key wildfires in Spain.

Contents

1. File directories
2. Notebooks
3. Code Implementation
4. Code Requirements

File Directories

• data
• figures
• notebooks
• results
• src

Notebooks

• toy_example.ipynb: Example of 100 x 100 cell grid of homogeneous fuels used to generate fire potential polygons and networks for decision-making.

Code Implementation

1. Clone the Repository: Clone the repository containing the environment YAML file to your local system.

   git clone https://github.com/humnetlab/FirePolygons.git

2. Change your current directory to the repository directory.

   cd FirePolygons
   

3. Create the environment from the YAML file

   conda env create -f environment.yml
   
   

Code Requirements

• python
• matplotlib
• pandas
• geopandas
• rasterio
• shapely
• networkx
• pysheds
• pyflwdir

Acknowledgements

The authors acknowledge the support of C3.ai through the grant Multiscale analysis for Improved Risk Assessment of Wildfires facilitated by Data and Computation (Link). The authors acknowledge the support of the Catalan Fire Service for accommodation and access to the wildfires during suppression operations. The authors also acknowledge the support of the Pau Costa Foundation.

Citation

Please cite the journal paper if this code is useful and helpful for your research.

@article{kim2025,
  title={Modeling Fire Potential Networks for Suppression Strategies Using Fire Spread Simulations and Hydrological Tools},
  author={Kim, Minho and Castellnou, Marc and Gonzalez, C. Marta},
  }