DYNAMIC Fire risk indicator implementation

This repository contains a simplified version the implementation of the dynamic fire risk indicator described in the submitted paper:

R.D: Strand and L.M. Kristensen. An implementation, evaluation and validation of a dynamic fire and conflagration risk indicator for wooden homes.

Compared to the original version, this repository only contains the fire risk calculation itself (without the hard-wired integration with the https://met.no integration and the more complex API). The calculation takes a CSV dataset containing time, temperature, relative humidity, and wind speed data points and provides the resulting fire risk as time to flashover (ttf).

Installation

The project is based on using uv as the package manager. If you want to build this project on you own, make sure that uv is installed correctly.

afterwards you can build the package with:

uv build

Which will create the package wheel (file ending .whl) in the dist/ directory. This package can ordinarily be installed with pip install and integrated into existing Python applications or it can be run standalone using python -m.

Alternatively you can test FRCM directly by running:

uv run python src/frcm/__main__.py ./bergen_2026_01_09.csv

where ./bergen_2026_01_09.csv is an example CSV demostrating the input format which comes bundled with this repo.

Overview

The implementation is organised into the following main folders:

• datamodel - contains an implementation of the data model used for weather data and fire risk indications.
• fireriskmodel contains an implementation of the underlying fire risk model.

The central method of the application is the method compute() in fireriskmodel.compute.