Skip to main content
ARS Home » Pacific West Area » Tucson, Arizona » SWRC » Research » Publications at this Location » Publication #334411
Title: UAV hyperspectral and lidar data and their fusion for arid and semi-arid land vegetation monitoring

Author
item SANKEY, T.T. - Northern Arizona University
item MCVAY, J. - Northern Arizona University
item SWETNAM, T. - University Of Arizona
item MCCLARAN, MITCHEL - University Of Arizona
item Heilman, Philip - Phil
item Nichols, Mary

Submitted to: Remote Sensing in Ecology and Conservation
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/27/2017
Publication Date: 4/1/2017
Citation: Sankey, T., Mcvay, J., Swetnam, T., Mcclaran, M., Heilman, P., Nichols, M.H. 2017. UAV hyperspectral and lidar data and their fusion for arid and semi-arid land vegetation monitoring. Remote Sensing in Ecology and Conservation. 4(1):20-33. https://doi.org/10.1002/rse2.44.
DOI: https://doi.org/10.1002/rse2.44

Interpretive Summary: Unmanned aerial vehicles (UAVs), sometimes referred to as drones, provide a new platform for collecting high spatial and temporal resolution measurements of vegetation and the earth surface. However, testing is needed to determine the capability of various sensors to characterize plant communities under field conditions. In this paper we demonstrate a combination of unmanned aerial vehicle (UAV) lidar and hyperspectral imagery (272 spectral bands) for individual plant species identification and 3D characterization of the earth surface at sub-meter scales on the Walnut Gulch Experimental Watershed in southeastern Arizona, USA. UAVs are capable of low altitude flights, which result in much high resolution data than is possible from sensors on satellites or airplanes. We thought both data sources together would perform better than hyperspectral alone in arid ecosystems with sparse vegetation. Hyperspectral classification by itself provided 82% accuracy. The combined approach provided 89% overall accuracy in target species classification at the canopy scale, leveraging a wide range of target spectral responses in the hyperspectral data and a high point density (50 points/m2) in the lidar data. Acquired at a relatively low cost, the UAV lidar-derived digital elevation model (DEM) is also strongly correlated with manned airborne lidar-derived DEM (R2=0.98). These applications can characterize earth systems at a level of detail which was impossible using traditional remote sensing systems.

Technical Abstract: We demonstrate a unique fusion of unmanned aerial vehicle (UAV) lidar and hyperspectral imagery for individual plant species identification and 3D characterization of the earth surface at sub-meter scales in southeastern Arizona, USA. We hypothesized that the fusion of the two different data sources would perform better than either data type alone in the arid ecosystem with sparse vegetation. The fusion approach provides 89% overall accuracy in target species classification at the canopy scale, leveraging a wide range of target spectral responses in the hyperspectral data and a high point density (50 points/m2) in the lidar data. Acquired at a reduced cost, the UAV lidar-derived digital elevation model (DEM) is also strongly correlated with manned airborne lidar-derived DEM (R2=0.98). High spatial and temporal resolution measurements of earth surface patterns critically needed in ecological, geologic, hydrologic, and geomorphic process models are possible with UAV sensors. These applications will illuminate new details in earth system phenomena, which were previously unobtainable or unaffordable using traditional remote sensing systems.