Applying freely available remote sensing data products to improve natural resource management : case studies of street tree benefits analysis and wetlands detection

Abstract

Natural resource management increasingly relies on geographic information systems (GIS) to facilitate decision making, but resource managers must manage the costs of geospatial data products. This thesis is divided into two topics focused on the use of freely available geospatial data to improve natural resource management. For the first topic, freely available street level imagery and benefits modeling software are utilized to conduct a virtual street tree survey. This new methodology has the potential to provide time effective and inexpensive access to benefits modeling while still providing a baseline understanding of associated street tree benefits. This study produced results that were statistically similar on a tree by tree basis between field and virtual surveys, but had lower modeled benefits due to underestimation of tree diameters in the virtual survey. In the second topic, wetlands detection was explored using freely available high resolution data products. LiDAR (light detection and ranging) and NAIP (National Agriculture Imagery Program) imagery were utilized to create two different products which may aid in wetlands detection. The first product used NAIP aerial photographs to separate pixels which had high probability of being vegetation based on their spectral reflectance in red and infrared, and combined these with LiDAR to show the heights of the vegetative surfaces. The second product utilized LiDAR data to generate contour lines and locate concentric depression centers that may be indicative of wetlands. While these products may not specifically identify wetland areas, they may be helpful for focusing conservation efforts.