Rothermel_SFIRE
Workflow information
Documentation page:
Version: 1.1
Date of record creation: 2024-12-03
Date of upload to firebench: 2024-12-03
Version/tag/commit firebench: 0.3.2a0
Configuration
Rate of spread model: Rothermel using
firebench.ros_models.Rothermel_SFIREimplementation.Number of point Sobol: 2^10
Specific inputs
The environmental variables chosen for this test are:
WIND_SPEEDfrom -15 m s-1 to 15 m s-1,SLOPE_ANGLEfrom -45 deg to 45 deg,FUEL_MOISTURE_CONTENTfrom 1% to 50%.
Hardware/software description
Apple M2, macOS 14.7.1
Python 3.10.8
Results
Fig.1 shows the execution time aggregated for all fuel classes (total) and for each fuel class. As fuel category 4 uses a different rate of spread calculation within the SFIRE implementation of Rothermel, a difference in performance is expected. Overall, the performance is very similar for each fuel category and a mean execution time of 6.698 \(\mu\)s over 106,496 samples.

Fig. 1 : Execution time boxplot for Rothermel rate of spread model using Anderson13 fuel model. Fliers points not shown on the figure.
Data
path to data:
01_generate_data.py:
ab95fdc101ae108b3d1592ec0291febf07621e3001c989b0973f2c9d69e48af102_plot_data.py:
6e8398298215938a08900555bb04390e4a9281cbb250f6fbc59a66eedb28f85403_create_record.py:
dfb138fe4ccaf999746eefdd9e23cc07453cd5b28ad8247f9eac15d7fa3d84edfirebench.log:
eb6ad262459db70de254592248239fe256d4e63f3af69bbbe35ac7d7feaae764output_data.h5:
421036734d13f57d84f9bc0e197255bd70523cf5b91e9e02d320ca805f189c04efficiency_box.png:
f1942c22f82c906f84be33c22ff596765e4336727ddfac80c860a470fe37dd7b