All 40 standard Scott & Burgan fuel models are hardcoded `isDynamic = true`, including the 23 static ones

[jimmynotjim]

Disclosure: This came up during a wildfire fuel modeling exploration being run by an AI agent. I'm a human submitting it, but I did not discover this on my own and typically work in JS/TS rather than C++ so I'm not personally familiar with the workings of this library. I have confirmed the source myself by adding a unit test locally and am preparing a PR with the test and fix.

In src/behave/fuelModels.cpp, FuelModels::populateFuelModels() passes isDynamic = true to setFuelModelRecord(...) for every standard Scott & Burgan (2005) fuel model (model numbers 101–204), regardless of whether the model is actually dynamic. As a result, getIsDynamic() returns true for 23 models that are static per S&B 2005 — the ones with no transferable live herbaceous load.

This surfaced downstream in the pyrothermel Python bindings (j-tenny/pyrothermel#2), where FuelModel.from_existing(code).is_dynamic returns True for e.g. SH4, TU5, TL1, SB1.

The static vs. dynamic distinction

A fuel model is "dynamic" when its live herbaceous load transfers to dead based on curing (S&B 2005) — i.e. exactly the models with non-zero fuelLoadLiveHerbaceous. Per S&B 2005 (and the load data in this very file), the dynamic standard models are GR1–GR9, GS1–GS4, SH1, SH9, TU1, TU3 — and only those.

What's wrong

The entire 101–204 block passes isDynamic = true (the 2nd-to-last argument of setFuelModelRecord). For the static models that's incorrect. The 23 mislabeled records:

Code	Model #	Name (note the trailing (S))	live-herb load	isDynamic now	should be
SH2	142	Moderate load, dry climate shrub (S)	0	true	false
SH3	143	Moderate load, humid climate shrub (S)	0	true	false
SH4	144	Low load, humid climate timber-shrub (S)	0	true	false
SH5	145	High load, dry climate shrub (S)	0	true	false
SH6	146	Low load, humid climate shrub (S)	0	true	false
SH7	147	Very high load, dry climate shrub (S)	0	true	false
SH8	148	High load, humid climate shrub (S)	0	true	false
TU2	162	Moderate load, humid climate timber-shrub (S)	0	true	false
TU4	164	Dwarf conifer understory (S)	0	true	false
TU5	165	Very high load, dry climate timber-shrub (S)	0	true	false
TL1–TL9	181–189	all (S)	0	true	false
SB1–SB4	201–204	all (S)	0	true	false

(GR1–9, GS1–4, SH1, SH9, TU1, TU3 are already correctly true.)

Impact: metadata only (which is why it likely went unnoticed)

isDynamic only does work when there's live herbaceous load to transfer to dead. Every mislabeled record has fuelLoadLiveHerbaceous = 0, so the dynamic load-transfer step is a no-op for them — fire-behavior outputs (rate of spread, flame length, etc.) are unaffected. The defect is purely in the reported getIsDynamic() attribute. It only matters to consumers that read the flag as metadata — e.g. pyrothermel exposing is_dynamic.

Minimal repro (data-level)

For any of the 23 records above, getIsDynamic(n) returns true while getFuelLoadLiveHerbaceous(n, ...) returns 0. For example, TU5 (#165):

getIsDynamic(165)                    -> true   // expected false
getFuelLoadLiveHerbaceous(165, ...)  -> 0      // static: nothing to transfer

A sweep of the standard burnable set (101–204) shows isDynamic == true for all of them, including every (S) model.

References

• Scott, J.H. & Burgan, R.E. (2005). Standard Fire Behavior Fuel Models: A Comprehensive Set for Use with Rothermel's Surface Fire Spread Model. RMRS-GTR-153. (Table 1 gives the dynamic/static designation.)
• Downstream report: FuelModel.from_existing(...).is_dynamic returns True for all 40 standard burnable models, including static SH/TU/TL/SB models j-tenny/pyrothermel#2