Computational Fire Acoustics: Numerical Modeling of Fire-Generated Infrasound Using 3D Finite-Difference Time-Domain Methods

Wildfires pose an increasing risk to forested and urban areas. Combustion related to fire has been shown to generate low-frequency acoustic waves and infrasound (sound below 20 Hz) (Marcillo et al., 2025). Here, acoustic waves from a semi-realistic linear fire source are modeled in a 3x3km simulated forest medium using infraFDTD, a 3-D numerical solver for the acoustic<br />wave equation using the Finite-Difference Time-Domain (FDTD) method (Kim and Lees, GRL 2011). The model is validated by comparing it to acoustic observations from a prescribed burn at Eglin Air Force Base in March 2023 (Marcillo et al., 2025). Model parameters are altered to constrain the source dimensions, geometry, and source-time function necessary to create an observable and distinct acoustic signal from a plume during a fire. Results show that the dimensions, intensity, and source function of firelines have a significant effect on signal strength<br />and frequency. All frequencies are constrained from 0-25 Hz. Longer firelines produce prolonged high-amplitude, low-frequency signals, with dominant signals in the 0-4 Hz range. More intense fires produce fewer low-amplitude, high-frequency signals (6-14 Hz) and stronger, longer-lasting signals in the 0-4 Hz range. Dominant frequencies of a semi-realistic fire signal range from 0-5 Hz, with weaker signals in the 5-15 Hz range, and higher frequencies (15-25 Hz) produced by noise. Comparative analysis with literature also confirms that fires with extensive flame fronts<br />produce stronger infrasound signals at 0.1 Hz. By utilizing modeling to give constraints on source properties, fires can be better identified from acoustic observations in the field. Additionally, back-propagation modeling using the Reverse Time Migration (RTM) technique (Kim and Lees, GJI 2015) demonstrates the potential for imaging the acoustic source of fires from observation stations below 15 Hz.&nbsp;