The primary purpose was to test a system for measuring fire behavior developed in grasslands (FM 3 & 1), in a woodland/forest situation (FM 2ish)
Secondary purposes included demonstrating fire science and providing data for the Blackfoot TREX.
Three FeatherFlame thermocouple datalogger systems were deployed ahead of three prescribed fires at the 2025 Blackfoot TREX.
Low fuel loads in these forests largely impeded fire spread under prescription conditions. Visual observation indicated that in many cases, fire spread was limited to small patches of needle cast fuels directly ignited by drip torches. In the absence of such fuels, fire rarely spread from ignition lines. Many instances of burning stumps and downed woody debris, and torching trees, were observed.
The FeatherFlame sensor is effective in measuring rate of spread in these forest fuels when fire spreads through the array. Substantial variability due to the factors described above means a greater, or targeted, sampling effort is required to adequately sample behavior of freely-moving flame fronts away from their point of ignition.
Care must be taken in interpreting FeatherFlame data from burns in which fire spread was poor. Sensors can reflect elevated temperature due to heat in the ambient environment rather than direct flame contact provided by true fire spread.
Soil heating in these forest fuelbeds can be significant and organic matter in the duff layer might be more responsible for carrying fire than conventional surface fine fuels.
Raw time-temperature curves for three above-surface temperature thermocouples (TC) from each of three datalogger units (F4, F5, F6) deployed in three prescribed fires at the Blackfoot TREX.
Raw time-temperature curves for two temperature thermocouples (TC) located at the soil surface from each of three datalogger units (F4, F5, F6) deployed in three prescribed fires at the Blackfoot TREX.
Temperature results were challenging to interpret. As seen above, many sensors recorded elevated air temperatures without the distinct peak of flame coming into contact with the thermocouple bead.
Temperatures in the flame zone in arrays through which fire had actually spread only exceeded 100 deg C six times total, and only averaged above 100 for all three sensors at a point twice, both on Baldy C3.
Interestingly, in four cases, including all three loggers at Baldy C3, soil surface temperatures substantially exceeded flame zone temperatures, suggesting these fires spread by smoldering through forest duff and litter?
Mean maximum temperatures reached in the flame zone and soil surface on three prescribed burns at the Blackfoot TREX.
For comparison, maximum temperatures recorded in tallgrass prairie fuels at the NerdTREX in northern Missouri were as follows:
| Year | Level | Min | Mean | Max |
|---|---|---|---|---|
| 2024 | Flame | 100 | 352 | 597 |
| Surface | 10 | 236 | 492 | |
| 2025 | Flame | 422 | 530 | 674 |
| Surface | 65 | 388 | 658 |
Below are the data from individual sensors summarized above:
| event | logger | level | TC1 | TC2 | TC3 | TC4 | TC5 | Mean maximum temp |
|---|---|---|---|---|---|---|---|---|
| Baldy C4 | F4 | flame | 58 | 62 | 70 | 63.5 | ||
| Baldy C4 | F4 | surface | 37 | 87 | 62.25 | |||
| Baldy C4 | F5 | flame | 53 | 118 | 66 | 79.25 | ||
| Baldy C4 | F6 | flame | 62 | 62 | ||||
| Baldy C4 | F6 | surface | 258 | 257.8 | ||||
| Baldy C3 | F4 | flame | 33 | 32 | 32 | 32.42 | ||
| Baldy C3 | F4 | surface | 51 | 41 | 46 | |||
| Baldy C3 | F5 | flame | 95 | 80 | 137 | 103.9 | ||
| Baldy C3 | F5 | surface | 374 | 125 | 249.5 | |||
| Baldy C3 | F6 | flame | 184 | 258 | 256 | 233.1 | ||
| Baldy C3 | F6 | surface | 359 | 426 | 392.8 | |||
| Wedge 2.0 | F4 | flame | 34 | 36 | 34 | 34.33 | ||
| Wedge 2.0 | F4 | surface | 38 | 36 | 36.88 | |||
| Wedge 2.0 | F5 | flame | 46 | 134 | 84 | 87.67 | ||
| Wedge 2.0 | F5 | surface | 53 | 66 | 59.38 | |||
| Wedge 2.0 | F6 | flame | 62 | 44 | 40 | 48.67 | ||
| Wedge 2.0 | F6 | surface | 42 | 37 | 39.62 |
In addition to temperatures, the FeatherFlame system is designed to calculate rate of spread based on the arrival times of the flame front at each of the three flame-zone sensors.
Fuels at the Blackfoot TREX were very challenging for ROS measurement and interpretation, because the data do not necessarily reflect flame fronts as much as heating from the general fire environment. In fact, in several cases, the area under or even around the sensors remained green after the fires, and thermocouples detected general increases in surrounding air temperature. Data from these arrays were excluded from the ROS presentation.
| date | logger | event | ROS (m/min) |
|---|---|---|---|
| 01 May | F4 | Baldy C4 | 0.33 |
| 01 May | F5 | Baldy C4 | 1.3 |
| 02 May | F4 | Baldy C3 | 0.034 |
| 02 May | F5 | Baldy C3 | 0.45 |
| 02 May | F6 | Baldy C3 | 1.7 |
For comparison, rates of spread through two tallgrass prairie at the 2025 NerdTREX in Northern Missouri, measured with these same instruments, ranged from 5 to 20 m/min, and in three fires in 2024, ranged from 2 to 12 m/min.