Use of ultra-high spatial resolution aerial imagery in the estimation of chaparral wildfire fuel loads.

Development of methods that more accurately estimate spatial distributions of fuel loads in shrublands allows for improved understanding of ecological processes such as wildfire behavior and postburn recovery. The goal of this study is to develop and test remote sensing methods to upscale field estimates of shrubland fuel to broader-scale biomass estimates using ultra-high spatial resolution imagery captured by a light-sport aircraft. The study is conducted on chaparral shrublands located in eastern San Diego County, CA, USA. We measured the fuel load in the field using a regression relationship between basal area and aboveground biomass of shrubs and estimated ground areal coverage of individual shrub species by using ultra-high spatial resolution imagery and image processing routines. Study results show a strong relationship between image-derived shrub coverage and field-measured fuel loads in three even-age stands that have regrown approximately 7, 28, and 68 years since last wildfire. We conducted ordinary least square analysis using ground coverage as the independent variable regressed against biomass. The analysis yielded R values ranging from 0.80 to 0.96 in the older stands for the live shrub species, while R values for species in the younger stands ranged from 0.32 to 0.89. Pooling species-based data into larger sample sizes consisting of a functional group and all-shrub classes while obtaining suitable linear regression models supports the potential for these methods to be used for upscaling fuel estimates to broader areal extents, without having to classify and map shrubland vegetation at the species level.