DYNAMIC Fire risk indicator implementation

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

R.D. Strand and L.M. Kristensen: An implementation, evaluation and validation of a dynamic fire and conflagration risk indicator for wooden homes. In volume 238 of Procedia Computer Science, pp. 49-56, 2024. Proceedings of the 15th International Conference on Ambient Systems, Networks and Technologies Networks (ANT). Online: https://www.sciencedirect.com/science/article/pii/S187705092401233X

Compared to the original repository and the associated PyPI package, 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 datasetas input 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.

Make sure that you have uv installed.

You can test the library by running

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

where ./bergen_2026_01_09.csv is an example CSV file demonstrating the input format which comes bundled with this repository.

This should provide the following output:

Computing FireRisk for given data in 'dynamic-frcm-simple/bergen_2026_01_09.csv' (132 datapoints)

timestamp,ttf
2026-01-07T00:00:00+00:00,6.072481167177002
2026-01-07T01:00:00+00:00,5.7243022443357905
2026-01-07T02:00:00+00:00,5.511503568040109
2026-01-07T03:00:00+00:00,5.3486591429746895
2026-01-07T04:00:00+00:00,5.220546121174422
2026-01-07T05:00:00+00:00,5.122507965859784
2026-01-07T06:00:00+00:00,5.043379927801035
2026-01-07T07:00:00+00:00,4.97841119689852
2026-01-07T08:00:00+00:00,4.9228585219502525
2026-01-07T09:00:00+00:00,4.873359414165299
2026-01-07T10:00:00+00:00,4.836470479216353
...

You can also build the project as a 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.

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.