Mapping forest structure and uncertainty in an urban area using leaf-off lidar data

Forests are important to nutrient cycling, hydrology, climate and aesthetics in urban ecosystems. Effective forest management in urban environments requires detailed data on the spatial distribution and structure of urban forests, but the lidar which are best for mapping the complexity of these forests are often unavailable or prohibitively expensive for municipalities. However, leaf-off small footprint lidar originally collected for topographic mapping are increasingly available, and will soon become accessible to forest managers in the U.S. through 3DEP (3D Elevation Program). In this paper, we demonstrated the opportunistic use of existing leaf-off lidar to map forest structure and associated uncertainties in Madison, Wisconsin and neighboring municipalities. Using empirical models, we were able to map five structural variables and aboveground biomass with accuracies comparable to or better than other studies using comparable data and with errors generally < 20 % of the data range. Highest uncertainties in our forest structure maps occurred in residential neighborhoods and along forest edges. From the results, we present maps of forest structure and, to our knowledge, first of a kind pixel-wise uncertainty maps for an urban area. These maps provide the basis for a spatially comprehensive assessment of forest resources and are effective for urban inventory and change assessment. For example, the maps enabled comprehensive comparison of carbon storage by urban trees among cities, with a range in our study of 1.2 kg/m(2) to 5.6 kg/m(2), and with major variations due to differences in city development patterns and ages.