Description: The January 15, 2014 Camp Swift Prescribed Research Burns, at the time they occurred, had the goal of collecting well-characterized datasets of environmental variables (e.g. local weather, topography, vegetation) and fire behavior (response of fire to environmental variables; overhead imagery of fire line progression; fire depth, and point measurements of radiant flux). These data are needed to support the testing and development of physics-based fire behavior models such as the Wildland-Urban Interface (WUI) Fire Dynamics Simulator (WFDS) and other physics-based fire models. As such, the objective of Camp Swift was to bring together researchers in the fire behavior, fire weather, fuels, models, experimental design, fire protection engineering, unmanned aerial vehicles, field data collection and prescribed burn implementation to collect these data and place these data on a central Google Drive. These were the only disciplines deemed necessary at the Camp Swift effort. It was assummed by those conducting Camp Swift that the above disciplines collecting data, performing some, but relatively little office data processing and dumping the data on a Google Drive where the steps necessary to achieve the above objective and provide a data set for model validation. Initial exploratory analysis rapidly identified the need for an integrated data set, produced following best practices in interdisciplinary research (Strang and McCleish) and geospatial science and technology (GSAT). Best practices in interdisciplinary research are required for these efforts because of the need for measurement specialists to work together and not perform individual research, as is typical of there training. Best practices in GSAT are required because of the need to manage data using established techniques, document data so that full methods employed can be understood by others, and integrate data in space and time. Ininitial examination indicated that all these needs appear to be required for producing a "well-characterized" dataset for physics-based fire model evaluation. A GSAT approach, for example, would immediately demand specifications and definitions of what a well-characterized dataset means, but was not present in the efforts leading up to the Camp Swift effort. Both interdisciplinary research and GSAT where not disciplines present at the Camp Swift Effort. Given that initial assessment indicated the data to be at best extremely questionable for the use of WFDS Validation, though subsequent evaluation by modelers might prove this statement wrong, a second objective, after the burns, became to use the Camp Swift effort to evaluate integration of scientific and engineering disciplines, teams, data integration, and data dissemination for assessment of the Fire Disturbance Continuum in small plot, controlled burn settings building towards providing specifications for a set of "well-characterized", integrated in space and time, data products representing pertinent information to support the validation of fire behavior models. Specifically, these efforts were being done to support the Wildland-Urban Interface (WUI) Fire Dynamics Simulator. On January 15, 2014 the National Institute of Standards and Technology (NIST), the United States Forest Service (USFS), Texas Forest Service (TFS), Joint Fire Science Program (JFSP), and other public and private organizations for three research burns on Camp Swift Military Base. The burns were conducted at Camp Swift Military Instillation which is a 4 718 hectares (11 659 acres) Texas Army National Guard Training Site located in Bastrop County in Central Texas. Burn plots were located on the southeast most side of the military facility in a fire unit referred to as the “Drop Zone”. The Drop Zone burn unit is 80.9 hectares (200 acres) and is managed for military operations such as personnel drops; therefore, the fire frequency and disturbance regime is higher in this unit than comparable grassland burn units at Camp Swift. The fire management regime in the Drop Zone is largely driven by requirements needed to maintain a safe and sustainable military training environment. Over a 2.5 hour time period three research burns, as portrayed in this dataset, were intentionally lit to support the validation of fire behavior models. Specifically, the effort was guided by validation needs of WFDS. The weather conditions were also chosen for this purpose, while staying within conditions that allowed for a safe prescribed burn.The fire practitioners timed the burns to coincide with similar weather parameters of relative humidity, temperature, wind speed and direction during the research burn periods. Minimum and maximum weather parameters for this ignition were as follows: wind directions: North, Northwest or SouthwestWind Speed: 9.7-24km/hr (6-15 mph)Relative Humidity: 20-25%Temperature: 1.6-29°C (35-85°F)The ignition procedure was designed to ensure, as much as possible from the ground, given the lack of homogeneity in the vegetation type and spatial distribution, that the initial fire established itself as a continuous fire line. The winds were chosen to have a higher enough speed to promote a continuous fireline. These are conditions important for WFDS validation.A validation dataset (as described above) in this case, might be defined as data that is sufficiently resolved (i.e., spatial and temporal scales) to allow a comparison of predicted and observed fire spread and fire depth. Ideally, sufficiently resolved heat and mass fluxes, temperature and wind (relevant to the fire) fields during flaming and smoldering. In the context of physics-based fire behavior models, an experiment whose objective is model validation should have a known, or expected fire behavior in the global observables (e.g. spread rate and fire depth). This is desirable because it reduces experimental uncertainty (this is not the same as measurement uncertainty, as documented in associated datasets) and allows for well founded detailed, and local, measurements targeted to validate models for the component physical processes. It’s the coupling of the physical process models that results in the global fire behavior. What this means, practically speaking, is that any local measurements of velocity, temperature, heat flux, wind fields should be understood in the context of the global fire behavior or they have marginal usefulness for WFDS or likely any physics-based fire model validation. For example, if it is not known if a particular set of instruments were in flanking versus heading versus merging fire lines, it is uncertain how to use the measured data for validation since the predicted global behavior might be wrong. This Camp Swift effort was not expected to meet all the above requirements but was conducted with the above understanding. This Environmental Systems Research Institute (ESRI) File Geodatabase polygon feature class portrays the spatial representation of the plot layouts for the Camp Swift research burns. Based on loacl expert knowledge and ground-site assessment, areas with the most continuous fuels were selected. Expert knowledge indicated that blue-stem would likely yield the highest fire intensities and most desirable fire behavior and areas were choosen from the ground accordingly. The high disturbance regime resulted in pre-fire publiclly available aerial imagery of the area not being representative of the conditions at the time of the burn. Logistical requirements did not allow for aerial flights before plot setup. Minimal disturbance to the interior of the plot was also an objective but the plots did have extensive military vehicular traffic at unknown times prior to the burn as can be seen in tire track observable in the pre-fire unmanned aerial image, which was not observable from the ground. This vegetation disturbance did not appear to affect head fire behavior greatly, though did possibly affect flanking fire situations. Nonetheless, this approach recognized that fires burning in a similar fuel type can be different given the actual burning conditions for each plot and time of day for ignition due to changes in weather conditions throughout the day. Different ignition patterns were employed so repeatability of burns was not addressed in these case studies. The design, however, allowed for the examination of multiple sampling methodologies and technologies to collect measurements of pre-fire, active fire, and post-fire conditions and demonstrate overcoming key logistical challenges present in conducting large field efforts. This design, coupled with the availability of the data, allowed for integration of the data as presented in this and associated data packages.
Copyright Text: Funding for this project provided by the Joint Fire Science Program (JFSP 11-1-3-29): https://www.firescience.gov. Funding also provided by the USDA Forest Service, Rocky Mountain Research Station and Pacific Northwest Research Station. Funding also provided by the National Institute of Standards and Technology.