Identifying Forest Canopy Gaps Using Satellite and Aerial Imagery

Submitted by atripp on Thu, 06/03/2021 - 12:44

In the early 2000s, Pictured Rocks National Lakeshore, Michigan, was infested by Beech Bark Disease (BBD). The disease kills the American beech tree (Fagus grandidentata), one of the dominant tree species in the upland forests of Pictured Rocks.  The loss of the beech trees creates gaps in the forest canopy and deprives wildlife species of a major nesting, shelter, and food source.  To understand how wildlife is reacting to the more fragmented canopy cover and altered food and habitat availability, park biologists need information about how forest canopy connectivity has changed over time. To address this data need,  the USGS Upper Midwest Environmental Sciences Center, as part of the Natural Resources Preservation Project (NRPP), is creating forest canopy gap layers from two imagery sources: aerial imagery collected in September 2005 and DigitalGlobe satellite imagery collected in 2016, 2018, and 2019. Forest canopy gaps will be distinguished using a multi-resolution segmentation algorithm in eCognition software. Canopy gaps can be identified using spectral values to distinguish between brighter cell values (tree canopies) and darker cell values (shadows or holes in the tree canopy). Segmentation algorithm parameters established by Nymageroh et al. will be used to determine settings in this project. Forest canopy gap shapefiles will be created to assess forest canopy change over time and to locate potential areas of impact from BBD infestation.

The yellow polygons represent forest canopy gaps identified by Trimble eCognition software from aerial imagery collected in 2005.

Author Name
Stephanie Sattler
Author Email