Abstract
Cities measure the structure and function of their urban forest resource to optimize forest managementand the provision of ecosystem services. Measurements made using plot sampling methods yield useful results including citywide or land-use level estimates of species counts, leaf area, biomass, and air pollution reduction. However, these quantities are statistical estimates made over large areas and thus are not spatially explicit. Maps of forest structure and function at the individual tree crown scale can enhance management decision-making and improve understanding of the spatial distribution of ecosystem services relative to humans and infrastructure. In this research we used hyperspectral imagery and waveform lidar data to directly map urban forest species, leaf area index (LAI), and carbon storage in downtown Santa Barbara, California. We compared these results to estimates produced using field-plot sampling and the i-Tree Eco model. Remote sensing methods generally reduced uncertainty in species-level canopy cover estimates compared to field-plot methods. This was due to high classification accuracy for species with large canopies (e.g.,
Platanus racemosa with ∼90% average accuracy,
Pinus pinea at ∼93%,
Quercus agrifolia at ∼83%) and high standard error of the plot-based estimates due to the uneven distribution of canopy throughout the city. Average LAI in canopy, based on lidar measurements was 4.47 while field measurements and allometry resulted in an LAI of 5.57. Citywide carbon storage, based on lidar measurements and allometry was estimated at 50,991 metric tons (t) and 55,900 t from plot-sampling. As others have noted, carbon density varied substantially by development intensity based largely on differences in fractional cover but less so when only evaluating canopy biomass. Using separate biomass equations for each leaf type (broadleaf, needleleaf, palm) resulted in a more accurate carbon map but a less accurate citywide estimate.
Citation
Alonzo, Michael; McFadden, Joseph P.; Nowak, David J.; Roberts, Dar A. 2016. Mapping urban forest structure and function using hyperspectral imagery and lidar data. Urban Forestry & Urban Greening. 17: 135-147.
