Tree Canopy

stat.stpete.org | Last Updated 21 May 2024

<DIV STYLE="text-align:Left;"><DIV><DIV>The Pinellas county LiDAR project was collected using Geiger Mode LiDAR at 20 ppsm and later decimated to approximately 10 ppsm for the post processing. The primary purpose of the LiDAR was to create DEM’s for hydraulic modeling. Additional deliverables included automated tree canopy polygons. Various samples were provided showing minimum polygon sizes of 100 square feet, 1000 square feet, and 43,560 square feet. The county chose a minimum area of 100 square feet.The LiDAR data was further decimated to increase processing speeds. Vegetation that was higher than 15 feet was used in the tree canopy polygon creation. LP360 was used to automatically create polygons using a Point Cloud Task customized from the Point Group Tracing template: The Grow Window parameter is a moving windowsize used to group points based on the <A href="mk:@MSITStore:C:\Program%20Files%20(x86)\QCoherent\LP360%20for%20ArcGIS\Help\lp360.chm::/Contents/buildingextractorpp_tracecls.htm">Boundary Trace Class</A>(Class 5 –High Vegetation). The moving window will be a square with length and width twice the value. This moving window is fit around an initial point and a surface growingprocess occurs grouping points with the same classification and within a distance specified by grow window parameters. The grow window value approximates the ground sample distance, or be slightly larger.The extracted outlines are saved to shapefiles with attributes that describe the traced objects. Attributescaptured for each object include the Area. The extraction of the polygons was computed for each of six sub-blocks of the entire county. For each one of the blocks, the polygons were run through the ‘Repair Geometry’in Arc Toolbox and then a conversion of Multipart to Single part was run. Because automated routines rarely produce perfect results, a manual review was performed. The polygons were superimposed over the LiDAR data and any Class 5 group of points that were not extracted were digitized using a manual approach. When the blocks were merged into one county-wide data set, the polygons along the block edges needed to be edited to merge adjacent polygons so that there were no discontinuities along the block edges. The areas of the polygons were re-computed and those that were less than 100 square feet were deleted. </DIV></DIV></DIV>