An investigation of hyperspectral technology for use in tree identification

Abstract

Hyperspectral imaging is a niche remote sensing modality that captures Earth’s surface reflectance across hundreds of spectral bands. Its high spectral resolution enables the detection of subtle spectral variations that correspond to changes in chemistry (e.g., pigments, moisture, minerals) and biological processes. Despite this potential, adoption remains limited, partly due to the lack of established spectral relationships for novel applications. More importantly, the fundamental principles connecting spectral differences to physiochemical changes are often overlooked in disciplines that have been slower to adopt hyperspectral methods. Accordingly, the first research chapter of this thesis outlines these principles and reviews existing applications to guide remote sensing practitioners towards informed innovation. The following chapter applies these principles to an early-stage framework for investigating foliar canopy reflectance in an understudied spectral range. Finally, this thesis closes with a commentary on how the most common uses today may be falling short of hyperspectral imaging’s true potential.