Use of UAS-based LiDAR and multispectral data for estimating vegetation parameters

Abstract

This master's thesis focuses on estimating vegetation parameters using multispectral and Light Detection and Ranging (LiDAR) sensors mounted on unmanned aerial systems (UAS). The aim was to investigate different methodologies for green area index (GAI) and plant area index (PAI) estimation, specifically in the context of a winter wheat crop over an entire growing season. We employed three distinct methods for GAI estimation using multispectral data and two methods for PAI estimation using LiDAR data. The GAI estimation involved the use of the Fraction of Vegetation Cover (FVC) approach with the Normalized Difference Vegetation Index (NDVI), Normalized Difference Red and Red Edge (NDRRE), and Normalized Difference Red Edge (NDRE) indices. The PAI estimation utilized a modified Beer-Lambert light extinction model with ground and canopy returns and intensity ratios. Results indicate the potential of LiDAR intensity for estimating PAI. The GAI estimation using NDRRE also shows a promising future by being less affected by saturation and shadow effects compared to the two other indices. Although the obtained results may not have met the initial expectations, this thesis can serve as a foundation for further exploration and raises new questions that motivate future research in the field of remote sensing-based vegetation parameter estimation.