Estimating wildfire fuel consumption with multitemporal airborne laser scanning data and demonstrating linkage with MODIS-derived fire radiative energy

• Published in: Remote sensing of environment 2020-12, Vol.251, p.112114, Article 112114
• Main Authors: McCarley, T. Ryan, Hudak, Andrew T., Sparks, Aaron M., Vaillant, Nicole M., Meddens, Arjan J.H., Trader, Laura, Mauro, Francisco, Kreitler, Jason, Boschetti, Luigi
• Format: Article
• Language: English
• Subjects: Airborne lasers; Airborne lidar; Bandelier national monument; Biomass; Biomass burning; Bulk density; Canopies; Carbon emissions; Eastern cascade range; Emissions; Estimation; Fire; Forest biomass; Fuel consumption; Laser applications; Lasers; MODIS; Occlusion; Pixels; Scanning; Spectroradiometers; Understory; Wildfires
• ISSN: 0034-4257; 1879-0704
• DOI: 10.1016/j.rse.2020.112114

Characterizing pre- and post-fire fuels remains a key challenge for estimating biomass consumption and carbon emissions from wildfires. Airborne laser scanning (ALS) data have demonstrated effectiveness for estimating canopy, and to a lesser degree, surface fuel components at fine-scale (i.e., 30 m) across landscapes. Using pre- and post-fire ALS data and corresponding field data, this study estimated consumption of canopy fuel (ΔCF), understory fuel (ΔUF), total fuel (ΔTF), and canopy bulk density (ΔCBD) for the 2012 Pole Creek fire in Oregon, USA (10,760 ha), and portions of the 2011 Las Conchas fire in New Mexico, USA (4,934 ha). Additionally, the feasibility of predicting fuel consumption was tested using separate pre- and post-fire models (PrePost), models combining all pre- and post-fire data (Pooled), and models using all data from both fires (Global). Estimates of ΔTF were then compared to fire radiative energy (FRE, units: MJ) derived from Fire Radiative Power (FRP, units: MW) observations from the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor onboard NASA Terra and Aqua satellites to mechanistically derive a biomass combustion coefficient (BCC, units: kg MJ−1). The PrePost and Pooled approaches yielded similar results at Las Conchas, but at Pole Creek insufficient pre-fire field data resulted in erroneous fuel consumption estimates outside the fire perimeter using the PrePost models. These results demonstrated that pre-fire field data were less important for these models than having field data which represent the full range of fuel conditions likely to exist across the landscape. Estimated total biomass consumed for the PrePost, Pooled, and Global models were 226 Gg, 224 Gg, and 224 Gg at Las Conchas, and 581 Gg, 713 Gg, and 552 Gg at Pole Creek. Comparisons between estimated ΔTF and FRE yielded an average BCC for both fires of 0.367 (s.d. ± 0.049) kg MJ−1 based on pixels with at least five MODIS observations. Both higher MODIS observations per pixel and accounting for canopy occlusion of FRE improved the relationship between ΔTF and MODIS-FRE. This study suggested a practical modelling approach for future efforts using only post-fire field observations and quantified a landscape-scale relationship between MODIS-derived FRE and fine-scale fuel consumption consistent with prior experiments. •Multitemporal ALS quantifies consumption of canopy and understory fuel.•Capturing full range of forest fuels is important for accurate modellin