Fire module in WRF 3.3

This page tracks the evolution of the fire module from the submission to the 3.3 release.

Release branch

To track the files, branch WRF_release_3.3 was created starting from the files as submitted, applied one submitted bug fix, and ending with the files from the WRF 3.3 release at http://www.mmm.ucar.edu/wrf/src/WRFV3.3.TAR.gz

Links to versions

• Version Nov 20, 2010 as originally submitted to WRF committee for the 3.3 release
• Bug fix Jan 18, 2011. This is the final version submitted to the WRF committee.
• Files after changes at NCAR in the submitted fire module files.
• Final files in the release after changes to the rest of the files

Changes at NCAR prior to the release

• Janice's changes in fire module files. See also her post describing the additional changes. The code changes were made in this commit on Mar 4 2011 (Jon is not the author, we just mirror the commit.)
  • Removed the wind factors feature [1] [2]. This limits the code to ideal runs with a single fuel, see details. This is the only change of substance in the code.
  • Edited comments [3] [4] [5] [6] [7], plus a number of smaller changes to the comments.
  • Small (max about 1%) changes to default values of fuel data variables [8]
  • Renamed variable ros_back to ros_base [9], added local variable ros_back [10] used only in chapparal. Two lines change results: [11] in the MacArthur formula and [12] in chaparral have the effect of increasing the spread rate by adding the zero-wind rate twice.
  • Deleted the identification of the git commit the files come from [13] [14] (appropriate when the files are no longer the same)
  • Replaced all examples by two new ones.
• Routine software engineering changes: Added ONLY to USE statements [15] [16] [17] [18] [19] [20]

Details on the wind factors

The change has the same effect as setting all windrf=1 in the data with the submitted code. The default value of fire_wind_height was not changed in the registry and remains at 6.096m; this results in 2-3 times increase of wind-driven spread rates when running the code at defaults. The intention was to set fire_wind_height to a smaller value in the namelist, but then the value is the same for all fuels.

The winds factors feature is important for the computations reported in our GMDD paper. It was surpassed later by Wind interpolation to different heights for different fuels, which is not in the release.