Journal of Emerging Scholarship

The important ecological role of submerged aquatic vegetation (SAV) makes its year-to-year distribution of significant interest to environmental monitoring organizations. The use of drones to perform the task of SAV monitoring through multispectral analyses is a promising tool to achieve a methodology that is automatable, repeatable, time efficient, and accessible. A preliminary trial was conducted at Eagle Cove near Gibson Island on the Magothy River where a DJI Phantom 4 Drone with a Sentera special purpose camera captured multispectral digital images with five spectral bands. These were used to apply and compare four vegetation indices: Normalized Difference Vegetation Index (NDVI), Green Normalized Difference Vegetation Index (GNDVI), Modified Normalized Difference Vegetation Index (mNDVI), and Normalized Difference Aquatic Vegetation Index (NDAVI). Analyses was done using the geographic information systems program known as ArcGIS Pro. The images generated by each index show some measure of successful identification of SAV, though there are many false-positives due to a variety of factors. The effectiveness of each index in our images was estimated by comparing the amount of pixels identified as SAV in the area of observed SAV growth and outside of this area. The most effective index was indicated to be mNDVI. This methodology will continue to be developed at AACC, and future work will aim to improve upon this process and to make calculations of SAV acreage and density that can be compared to ground-truthed observations.

Download the full paper