Analysis, modelling and estimation of wildfire fuel load in north-central BC forests

Improving our ability to assess fuel loads across forested landscapes will aid our ability to manage and mitigate wildfire risk in the short- and long-term. Forest stands exhibit complex spatial variation in fuel load amount and characteristics across landscapes.  Conventional stand-level inventory methods often fail to adequately capture this variability, particularly for fuel load specific metrics such as canopy fuel load (CFL), canopy bulk density (CBD), canopy base height (CBH) and coarse woody debris (CWD). Airborne laser scanning (ALS) presents an opportunity to model and predict these metrics at fine resolution, over landscape-level areas. In this study, I assessed fuel loading in four basic forested stand types common in northcentral BC:  live conifer, dead conifer, mixedwood and deciduous stands.  I secondly evaluated the potential for using ALS to model CFL, CBD, CBH, CWD and other pertinent fuel metrics across stand types.  ALS derived models were then used to project the above-noted metrics across the entire study area, and to compute critical surface fire intensity (CSI) at a ten-meter resolution for the 1,800 square kilometer land base that comprises the Chinook Community Forest and surrounding lidar coverage. Results demonstrate that live conifer, mixedwood, deciduous leading, and MPB killed pine stands differ considerably with respect to CFL, CBD, crown length and CWD, and have similar CBH  Live conifer stands had the highest canopy load, CFL and CBD. Fuel loads in mixedwood stands were not statistically different from live conifer stands for most fuel load metrics. Dead pine stands did not constitute a significant canopy fuel hazard; however, their surface fuel loads were very high. My results demonstrate that many of the forest metrics important for fuel load characterization can be modeled with high accuracy and mapped at a fine grain (i.e. 100 m2) using ALS derived data. My work demonstrates that modelling fuel loads using ALS is broadly applicable across the stand types in the study area, enabling the projection of fuel load metrics without being constrained by different stand types. This thesis contributes to the science of wildfire fuel load management by establishing a replicable framework for quantifying fine-resolution fuel load metrics using ALS data across diverse stand types and providing valuable insights for effective wildfire risk assessment and mitigation strategies in forest management.