Search for: All records

As wildfire activity increases worldwide, developing effective methods for estimating how fast it can spread is critical. This study aimed to develop and validate a computer vision algorithm for fire spread estimation. Using visual flame data from laboratory experiments on excelsior and pine needle fuel beds, we explored fire spread predictions for two types of experiments. In the first, the experiments were conducted in an environment where the flame was maintained visually undisturbed while in the second, real-world scenarios were simulated with visual obstructions. Algorithm performance evaluation was conducted by computing the index of agreement and normalized root mean square deviation (NRMSD) error. Results show that the algorithm estimates fire spread well in pristine visual environments with varying accuracy depending on the fuel type. For instance, the index of agreement between the rate of spread values estimated by the algorithm and the measured values is 0.56 for excelsior fuel beds and 0.51 for pine needle fuel beds. For visual obstructions, varying impacts on the rate of spread predictions were observed. Adding an orange background behind the flame had the least effect on algorithm performance (IAmedian = 0.45), followed by placing a Y-shape element resembling a branch (IAmedian = 0.31) and adding an LED light near the flame (IAmedian = 0.30). 

Terrestrial LiDAR scans (TLS) offer a rich data source for high-fidelity vegetation characterization, addressing the limitations of traditional fuel sampling methods by capturing spatially explicit distributions that have a significant impact on fire behavior. However, large volumes of complex, high resolution data are difficult to use directly in wildland fire models. In this study, we introduce a novel method that employs a voxelization technique to convert high-resolution TLS data into fine-grained reference voxels, which are subsequently aggregated into lower-fidelity fuel cells for integration into physics-based fire models. This methodology aims to transform the complexity of TLS data into a format amenable for integration into wildland fire models, while retaining essential information about the spatial distribution of vegetation. We evaluate our approach by comparing a range of aggregate geometries in simulated burns to laboratory measurements. The results show insensitivity to fuel cell geometry at fine resolutions (2–8 cm), but we observe deviations in model behavior at the coarsest resolutions considered (16 cm). Our findings highlight the importance of capturing the fine scale spatial continuity present in heterogeneous tree canopies in order to accurately simulate fire behavior in coupled fire-atmosphere models. To the best of our knowledge, this is the first study to examine the use of TLS data to inform fuel inputs to a physics based model at a laboratory scale.