Alistair Smith | Search USGS

USGS2017

Landscape-scale quantification of fire-induced change in canopy cover following mountain pine beetle outbreak and timber harvest

Across the western United States, the three primary drivers of tree mortality and carbon balance are bark beetles, timber harvest, and wildfire. While these agents of forest change frequently overlap, uncertainty remains regarding their interactions and influence on specific subsequent fire effects such as change in canopy cover. Acquisition of pre- and post-fire Light Detection and Ranging (LiDAR) data on the 2012 Pole Creek Fire in central Oregon provided an opportunity to isolate and quantify fire effects coincident with specific agents of change. This study characterizes the influence of pre-fire mountain pine beetle (MPB; Dendroctonus ponderosae ) and timber harvest disturbances on LiDAR-estimated change in canopy cover. Observed canopy loss from fire was greater (higher severity) in areas experiencing pre-fire MPB (Δ 18.8%CC) than fire-only (Δ 11.1%CC). Additionally, increasing MPB intensity was directly related to greater canopy loss. Canopy loss was lower for all areas of pre-fire timber harvest (Δ 3.9%CC) than for fire-only, but among harvested areas, the greatest change was observed in the oldest treatments and the most intensive treatments [i.e., stand clearcut (Δ 5.0%CC) and combination of shelterwood establishment cuts and shelterwood removal cuts (Δ 7.7%CC)]. These results highlight the importance of accounting for and understanding the impact of pre-fire agents of change such as MPB and timber harvest on subsequent fire effects in land management planning. This work also demonstrates the utility of multi-temporal LiDAR as a tool for quantifying these landscape-scale interactions.

Oregon↗

USGS2017

Multi-temporal LiDAR and Landsat quantification of fire-induced changes to forest structure

Measuring post-fire effects at landscape scales is critical to an ecological understanding of wildfire effects. Predominantly this is accomplished with either multi-spectral remote sensing data or through ground-based field sampling plots. While these methods are important, field data is usually limited to opportunistic post-fire observations, and spectral data often lacks validation with specific variables of change. Additional uncertainty remains regarding how best to account for environmental variables influencing fire effects (e.g., weather) for which observational data cannot easily be acquired, and whether pre-fire agents of change such as bark beetle and timber harvest impact model accuracy. This study quantifies wildfire effects by correlating changes in forest structure derived from multi-temporal Light Detection and Ranging (LiDAR) acquisitions to multi-temporal spectral changes captured by the Landsat Thematic Mapper and Operational Land Imager for the 2012 Pole Creek Fire in central Oregon. Spatial regression modeling was assessed as a methodology to account for spatial autocorrelation, and model consistency was quantified across areas impacted by pre-fire mountain pine beetle and timber harvest. The strongest relationship (pseudo-r 2 = 0.86, p < 0.0001) was observed between the ratio of shortwave infrared and near infrared reflectance (d74) and LiDAR-derived estimate of canopy cover change. Relationships between percentage of LiDAR returns in forest strata and spectral indices generally increased in strength with strata height. Structural measurements made closer to the ground were not well correlated. The spatial regression approach improved all relationships, demonstrating its utility, but model performance declined across pre-fire agents of change, suggesting that such studies should stratify by pre-fire forest condition. This study establishes that spectral indices such as d74 and dNBR are most sensitive to wildfire-caused structural changes such as reduction in canopy cover and perform best when that structure has not been reduced pre-fire.

Oregon↗