As part of a larger Great Lakes Restoration Initiative - Remote Sensing project led by the FWS, Jim Klassen with SharedGeo automatically derived 2-m vegetation surface canopy digital elevation models from stereo submeter optical satellite imagery using the National Science Foundation's Blue Waters supercomputer. This is one early example where imagery was compared to previous models derived from digital stereo aerial imagery and lidar imagery. The advantage of the approach is to map seasonal vegetation canopy changes over time to better discriminate specific wetland species such as black ash forests or invasive phragmites. With the additional support of the University of Minnesota, Polar Geospatial Center, SharedGeo processed the entire stereo WorldView imagery archive covering the Great Lakes Basin in about a month. If this was done on a desktop workstation, it would have taken years.
Surface Model derived from stereo aerial digital imagery taken in 2008; rough surface water from windy conditions caused the speckled appearance in the center of the image at blue arrow. Top right: 50-cm Digital Surface Model from lidar taken in 2011.
Bottom left: 2-m Digital Surface Model derived from WorldView-3 stereo images taken on April 30, 2016 (leaf-off). Bottom center: 2-m Digital Surface Model derived from WorldView-3 stereo images taken on October 3, 2016. Bottom right: The same October 3 image with a red circle highlighting a new forest cut through the middle of Tallas Island. The two red circles in the lower right highlight the canopy disappearance of two black ash forest wetland clumps since they lose their leaves in mid-September.
Images courtesy of DigitalGlobe under the NextView license administered by the National Geospatial-Intelligence Agency.