The identification of tree species can provide a useful and efficient tool for forest managers for planning and monitoring purposes. Hyperspectral data provide sufficient spectral information to classify individual tree species. Two non-parametric classifiers, support vector machines (SVM) and random forest (RF), have resulted in high accuracies in previous classification studies. This research takes a comparative classification approach to examine the SVM and RF classifiers in the complex and heterogeneous forests of Muir Woods National Monument and Kent Creek Canyon in Marin County, California. The influence of object- or pixel-based training samples and segmentation size on the object-oriented classification is also explored. To reduce the data dimensionality, a minimum noise fraction transform was applied to the mosaicked hyperspectral image, resulting in the selection of 27 bands for the final classification. Each classifier was also assessed individually to identify any advantage related to an increase in training sample size or an increase in object segmentation size. All classifications resulted in overall accuracies above 90%. No difference was found between classifiers when using object-based training samples. SVM outperformed RF when additional training samples were used. An increase in training samples was also found to improve the individual performance of the SVM classifier.

An accelerating trend of global urbanization accompanying population growth makes frequently updated land use and land cover (LULC) maps critical. LULC maps have been widely created through the classification of remotely sensed imagery. Maps of urban areas have been both dichotomous (urban or non-urban) and entailing of discrete urban types. This study incorporated multispectral built-up indices, designed to enhance satellite imagery, for introducing new urban classification schemes. The indices examined are the new built-up index (NBI), the built-up area extraction index (BAEI), and the normalized difference concrete condition index (NDCCI). Landsat Level-2 data covering the city of Miami, FL, USA was leveraged with geographic data from the Florida Geospatial Data Library and Florida Department of Environmental Protection to develop and validate new methods of supervised and unsupervised classification of urban area. NBI was used to extract discrete urban features through object-oriented image analysis. BAEI was found to possess properties for visualizing and tracking urban development as a low-high gradient. NDCCI was composited with NBI and BAEI as the basis for a robust urban intensity classification scheme superior to that of the United States Geological Survey National Land Cover Database 2016. BAEI, implemented as a shadow index, was incorporated in a novel infill geosimulation of high-rise construction. The findings suggest that the proposed classification schemes are advantageous to the process of creating more detailed cartography in response to the increasing global demand.

The clear need for accurate landslide susceptibility mapping has led to multiple approaches. Physical models are easily interpreted and have high predictive capabilities but rely on spatially explicit and accurate parameterization, which is commonly not possible. Statistical methods can include other factors influencing slope stability such as distance to roads, but rely on good landslide inventories. The maximum entropy (MaxEnt) model has been widely and successfully used in species distribution mapping, because data on absence are often uncertain. Similarly, knowledge about the absence of landslides is often limited due to mapping scale or methodology. In this paper a hybrid approach is described that combines the physically-based landslide susceptibility model “Stability INdex MAPping” (SINMAP) with MaxEnt. This method is tested in a coastal watershed in Pacifica, CA, USA, with a well-documented landslide history including 3 inventories of 154 scars on 1941 imagery, 142 in 1975, and 253 in 1983. Results indicate that SINMAP alone overestimated susceptibility due to insufficient data on root cohesion. Models were compared using SINMAP stability index (SI) or slope alone, and SI or slope in combination with other environmental factors: curvature, a 50-m trail buffer, vegetation, and geology. For 1941 and 1975, using slope alone was similar to using SI alone; however in 1983 SI alone creates an Areas Under the receiver operator Curve (AUC) of 0.785, compared with 0.749 for slope alone. In maximum-entropy models created using all environmental factors, the stability index (SI) from SINMAP represented the greatest contributions in all three years (1941: 48.1%; 1975: 35.3; and 1983: 48%), with AUC of 0.795, 0822, and 0.859, respectively; however; using slope instead of SI created similar overall AUC values, likely due to the combined effect with plan curvature indicating focused hydrologic inputs and vegetation identifying the effect of root cohesion. The combined approach––using either stability index or slope––highlights the importance of additional environmental variables in modeling landslide initiation.

LiDAR-derived slope models may be used to detect abandoned logging roads in steep forested terrain. An object-based classification approach of abandoned logging road detection was employed in this study. First, a slope model of the study site in Marin County, California was created from a LiDAR derived DEM. Multiresolution segmentation was applied to the slope model and road seed objects were iteratively grown into candidate objects. A road classification accuracy of 86% was achieved using this fully automated procedure and post processing increased this accuracy to 90%. In order to assess the sensitivity of the road classification to LiDAR ground point spacing, the LiDAR ground point cloud was repeatedly thinned by a fraction of 0.5 and the classification procedure was reapplied. The producer’s accuracy of the road classification declined from 79% with a ground point spacing of 0.91 to below 50% with a ground point spacing of 2, indicating the importance of high point density for accurate classification of abandoned logging roads.

The benefits of meadow restoration can be assessed by understanding the connections among geomorphology, hydrology, and vegetation; and multispectral imagery captured from unpiloted aerial systems (UASs) can provide the best method in terms of cost, resolution, and support for vegetation indices. Our field studies were conducted on northern Sierra montane meadows (with ≤70 km² watershed area). The meadows exist in various stages of ecological restoration. Field survey methods included GPS + laser-leveling channel survey, cross-sections, LiDAR, vegetation sampling, soil measurements, and UAS imaging. A sensor captured calibrated blue (465–485 nm), green (550–570 nm), red (663–673 nm), near infrared (NIR) (820–860 nm), and red-edge (712–722 nm) bands at 5.5 cm resolution (as well as thermal at 81 cm resolution) and provided multispectral images and derivative vegetation indices such as the normalized difference vegetation index (NDVI) and red-edge chlorophyll index (Cl_re). This fine-scale imagery extended our morphometric assessment of post-restoration channel bedform patterns and sinuosity related to Carex-influenced soil properties and Salix influence, and also documented groundwater-related effects via Carex patterns evident from spring snowmelt images, as well as NDVI and Cl_re (derived from spring and summer images) in growing to senescent phenological stages. Carex was significantly associated with low bulk density and high soil moisture, NDVI, and Cl_re in low-lying areas, and channel sinuosity was significantly associated with willow influence. Our methods can be applied by restoration managers to assess where projects are threatened by renewed incision and to document levels of carbon sequestration significant to addressing climate change.

Quantifying forest structure to assess changing wildfire risk factors is critical as vulnerable areas require mitigation, management, and resource allocation strategies. Remote sensing offers the opportunity to accurately measure forest attributes without time-intensive field inventory campaigns. Here, we quantified forest canopy cover and individual tree metrics across 44 plots (20 m × 20 m) in oak woodlands and mixed-conifer forests in Northern California using structure-from-motion (SfM) 3D point clouds derived from unoccupied aerial systems (UAS) multispectral imagery. In addition, we compared UAS–SfM estimates with those derived using similar methods applied to Airborne Laser Scanning (ALS) 3D point clouds as well as traditional ground-based measurements. Canopy cover estimates were similar across remote sensing (ALS, UAS-SfM) and ground-based approaches (r² = 0.79, RMSE = 16.49%). Compared to ground-based approaches, UAS-SfM point clouds allowed for correct detection of 68% of trees and estimated tree heights were significantly correlated (r² = 0.69, RMSE = 5.1 m). UAS-SfM was not able to estimate canopy base height due to its inability to penetrate dense canopies in these forests. Since canopy cover and individual tree heights were accurately estimated at the plot-scale in this unique bioregion with diverse topography and complex species composition, we recommend UAS-SfM as a viable approach and affordable solution to estimate these critical forest parameters for predictive wildfire modeling.