Assessing day and night UAV-based estimates of crop characterization using LiDAR

Advancements among remote sensing technologies and platforms have provided opportunities to collect spatially continuous biophysical parameters for the benefit of precision agriculture. Specifically, unmanned aerial vehicles with mounted Light Detection and Ranging (LiDAR) sensors have promise in characterizing crop biophysical parameters, such as canopy height (CH) and leaf area index (LAI). This study aimed to demonstrate the assessment of CH and LAI using UAV LiDAR-based methodologies under varying light conditions (e.g. day or night). The methods incorporated the use of the cloth simulation filter (CSF) to assign ground/vegetation status of the LiDAR point cloud for CH calculations, and then further benefited from the inclusion of laser penetration and intensity-based LiDAR indices to estimate LAI. These methods were replicated with identical day and night flights for a multi-planting date (seven dates) corn experiments conducted in Temple, Texas. Findings suggest that both CH ( R 2 ≥0.91; RMSE ≤15.43) and LAI ( R 2 ≥0.77, RMSE =0.52) measurements are highly correlated when comparing in-situ and LiDAR-based measurements. When comparing between LiDAR data collected during day and night flights, CH showed very little difference (Δ R 2 ≤0.01) and were found to be highly replicable. LAI measures presented more differences (Δ R 2 ≤0.05) between the day and night collections that we attribute to leaf turgor due to the relatively minor differences among models. The findings within provide the foundation for future works aiming to incorporate day and/or night LiDAR-based precision agriculture.