Research

Pyroaerobiology modeling

Wildland fires are shown to be a primary global driver in the exchange of bio-aerosols between the biosphere and the atmosphere [1]. To simulate the transport of these particles through wildfire plumes at landscapes scales, we have developed a highly computationally-scalable and adaptive Langevin Dynamic model in the Lagrangian framework. The model can be coupled with various CFD solvers and is currently coupled with WRF-SFIRE. This model creates a unique capability to better understand the complex interplay between wildfires, natural ecosystems, and atmospheric processes.

The image on the right showcases a snapshot of our simulations for the Manning Creek prescribed fire. Specifically, the spatial distribution of the aerosolized microbial particle count, aggregated over a $20 \times 20 \: m^{2}$ area, is shown along with the heat flux from the fire. The extent of the computational domain is $17.5 \times 17.5 \: km^{2}$ and our preliminary results indicate that many bio-aerosols exit the computational domain, implying that the wildfire plume may significantly extend the impacts of the wildfire far beyond the burn perimeter.

[1] L. N. Kobziar et al., “Wildland fire smoke alters the composition, diversity, and potential atmospheric function of microbial life in the aerobiome,” ISME COMMUN., vol. 2, no. 1, Art. no. 1, Jan. 2022, doi: 10.1038/s43705-022-00089-5.

Heat transfer from Firebrands

Wildfire propagation is driven by two main mechanisms that are (1) the spread of local fire front through convection and radiation heat transfer during exposure of fuels to flames and (2) firebrand showers, also known as ember attacks; see the top figure on the left. Firebrand showers are the ignition of spot fires as a result of generation, transport, and deposition of firebrands away from the fire line. The accumulation of firebrands can lead to spot fires, which are known to be responsible for losses in Wildland Urban Interface (WUI) communities and may alter the fire behavior itself. In this NSF-supported project, we adopt a fully coupled numerical modeling approach to simulate the heat transfer from firebrands to the recipient surface through the boundary layer.

[2] A. Tohidi and N. B. Kaye, “Firebrands,” in Wildland Fire Dynamics, Fire Effects and Behavior from a Fluid Dynamics Perspective, K. Speer and S. Goodrick, Eds., Cambridge: Cambridge University Press, 2022, pp. 129–155. doi: 10.1017/9781108683241.005.

Inverse solver for wildfire forecast

Wildfire spread involves multi-scale and multi-physics interaction of phenomena ranging from micro-scale combustion reactions to large atmospheric interactions between the fire and ambient weather on a macro-scale. Recent advancements in fire science and high-performance scientific computing (HPC) have made significant progress in the physics-based modeling of some sub-processes at small to mid-flame scales. However, due to high computational costs, the findings are difficult to generalize to wildfire flame scales. This work adopts a data-centric approach to develop an inverse solver for wildfire propagation using the combination of physics-based models and observational data from prescribed burn experiments to improve the accuracy of wildfire forecasts. The figure on the right shows a sample of the corrected simulation by our inverse solver. SJSU Tower Foundation has supported this project through the IBM Public Impact Projects initiative.

[3] M. A. Finney, “FARSITE: Fire Area Simulator-model development and evaluation,” U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Ft. Collins, CO, RMRS-RP-4, 1998. doi: 10.2737/RMRS-RP-4.