An ArcGIS Pro workflow to extract vegetation indices from aerial imagery of small-plot turfgrass research

Abstract

Collection of multispectral imagery from an aerial sensor is a means to obtain plot-level vegetation index (VI) values; however, postcapture image processing and analysis remains a challenge for small-plot researchers. An ArcGIS Pro workflow of two task items was developed with established routines and commands to extract plot-level VI values (normalized difference VI [NDVI], ratio VI [RVI], and chlorophyll index–red edge [CI-RE]) from multispectral aerial imagery of small-plot turfgrass experiments. Users can access and download task items from the ArcGIS Online platform for use in ArcGIS Pro. The workflow standardizes processing of aerial imagery to ensure repeatability between sampling dates and across site locations. A guided workflow saves time with assigned commands, ultimately allowing users to obtain a table with plot descriptions and index values within a.csv file for statistical analysis. The workflow was used to analyze aerial imagery from a small-plot turfgrass research study evaluating herbicide effects on St. Augustinegrass [Stenotaphrum secundatum (Walt.) Kuntze] grow-in. To compare methods, index values were extracted from the same aerial imagery by TurfScout, LLC and were obtained by handheld sensor. Index values from the three methods were correlated with visual percentage cover to determine sensitivity (i.e., the ability to detect differences) of the different methodologies. Index values collected by handheld sensor were more sensitive to visual cover than those extracted from aerial imagery. Index values extracted by TurfScout, LLC were generally more sensitive to visual percentage cover than the workflow extraction method, but both detected similar trends of increasing index values as percentage cover increased.