Comparing groundbased and aerial data at field scale during dry down on golf course fairways
Academic Article
Overview
Research
Identity
Additional Document Info
Other
View All
Overview
abstract
AbstractSpatiotemporal data and analyses are gaining traction in the turfgrass industry as valuable tools to enable datadriven management practices, but to date there has been minimal research in practical settings. The objective of this work was to quantify relationships between soil volumetric water content (VWC), proximal normalized difference vegetation index (NDVI), and several aerial measurements (visible, NDVI, and thermal infrared) collected in a realworld application at field scale during a dry down. Data collection surveys were conducted in 2020 the morning of 25 Feb, afternoon of 25 Feb, and morning of 27 February on three golf course fairways in CA, USA. The first survey was initiated following an irrigation event, and then no additional irrigation or rainfall occurred prior to the second and third surveys. Groundbased data were collected using the Precision Sense 6000 (The Toro Company, Bloomington, MN) and aerial data were collected using an unmanned aerial vehicle (GreenSight Inc., Boston, MA). Data were appropriately georeferenced and analyzed to determine correlation between VWC and proximal NDVI, proximal and aerial NDVI, and VWC and aerial measurements. A significant, weak correlation (r=0.21, p<.05) was found between VWC and proximal NDVI measurements, but only for the first survey immediately following the irrigation event. Significant, moderate to strong correlations were found between proximal and aerial NDVI during all three surveys (r=0.63, p<.001; r=0.64, p<.001; r=0.85, p<.001), respectively]. Volumetric water content was significantly correlated with aerial NDVI measurements (r=0.26, p<.001) but the relationship was weak and only existed for the first survey following the irrigation event. This study demonstrates the complexity of scaling remote sensing technologies from small plots to realworld applications and identifies several barriers to providing quantitatively predictive and actionable data to turfgrass managers.
