Identification of plant communities on barrier islands by using spatial properties derived from close-range and low-altitude UAV photos

Abstract:

Barrier islands reduce impacts of storms and hurricanes on coastal areas, and provide habitat for migrating waterfowl and other species such as turtles. Scientific investigations are prerequisites to effective habitat management. Different habitats usually associate with different prevalent plant associations. For examples, emergent areas closest to estuarine waters consist of regularly flooded salt-tolerant grasses. In brackish areas, species composition changes to brackish-water assemblage. Conventional mapping methods include visual interpretation on aerial photographs and field investigation, which is conducted to compare various plant communities in the field with corresponding photographic signatures on aerial photos for mapping purposes. The low-cost, low-altitude flying UAV provides an opportunity to obtain the vital plant community information. The UAV remote sensing system usually adopts off-the-shelf cameras. Although the cameras have low spectral and radiometric resolution, they typically have high spatial resolution. The prevalent plant associations may have similar spectral reflectance, however they usually have different appearances. The hyperspatial images acquired by close-range cameras and low-altitude UAV cameras are used to generate geometric characterization of plant communities. The experiment has two study areas, transects on Mustang Island and beach on South Padre Island. The images used in the experiment consists of close-range photos along transects on Mustang Island and the UAV images over South Padre Island beach. After interior orientation and geo-orthorectification, predominant spatial properties such as texture and geometric properties are calculated on these images. The effectiveness of these parameters for identifying plant communities will be evaluated. The experiment tries to build the spatial signatures of typical coastal plant communities. Positive results will further the use of UAV technology into coastal environmental management.