Planning Optimal UAS Flight Areas for Rare Plant Monitoring in the Blue Ridge Mountains of North Carolina

Traditional methods for mapping rare plant species are often resource intensive and subject to an individual's interpretation. However, rapid advances in UAS platforms and sensors provide powerful tools for rare plant monitoring. These tools facilitate the creation of high resolution photography and 3d models of inaccessible environments. In this study we present the results used to plan efficient flight areas based on a moving window resampling process from a maximum entropy species distribution model (SDM).

Know locations of Geum Radiatum, Solidago Spithamea and Houstonia Montana were each used as presence samples to train a maximum entropy SDM against 14 environmental variables in the Blue Ridge Mountains region of North Carolina. The resulting 20ft resolution rasters contained pixel values representing the probability of suitable conditions for each of the three species. A moving window with various geometries was used to resample and sum values into 10 ac² areas. The resampled rasters for each species were exported to polygon shapefiles and merged. The polygons with the highest values that intersected the presence samples were selected and exported as the boundaries for the UAS flight planning software.

Preliminary results suggest both the SDM and polygon flight areas accurately portrayed the highest probability habitat areas. Intersecting the highest value polygons with the presence samples ensured the flight plan geometry captured known sites as well as other highly suitable locations identified in the SDM.

Planning UAS flights from probabilistic modeling outputs will yield more relevant and precise datasets. Creating UAS flight plans from resampled SDM is a potentially useful technique for accurately monitoring rare plants; especially if the species is located in a challenging location where traditional methods could be hazardous and resource intensive.