Quantifying structural change in UK woodland canopies with a dual-wavelength full-waveform terrestrial laser scanner

Vegetation structure provides a direct link between forest ecosystem productivity and earth-atmosphere fluxes, and is both a result and driver of these interactions. Therefore, the ability to collect objective, quantitative and three-dimensional measurements of vegetation structure is essential, particularly in light of climate change. However, a significant challenge still remains as to how to best measure changes in forests and prepare for future climatic scenarios. Terrestrial Laser Scanning (TLS) has shown its potential to provide such measurements, offering a new approach to monitoring how forest systems change through time and space. The overall aim of this thesis was to improve the characterisation of the seasonal dynamics of UK woodland vegetation structure using the Salford Advanced Laser Canopy Analyser (SALCA), a research TLS with dual-wavelength full-waveform capabilities. There were three key objectives to this research: (1) the development of a radiometric calibration for the SALCA instrument to produce an apparent reflectance product, (2) the separation of SALCA point clouds into leaf and wood on a tree and plot scale using dual-wavelength lidar, and (3) the examination of the seasonal dynamics of vegetation structure in a range of UK forest types. To address these objectives, two field campaigns were conducted. SALCA measurements of artificial reflectance targets were acquired from both field campaigns to generate a calibration dataset to address Objective 1. The two field campaigns comprised a tree-scale validation experiment at Alice Holt Forest (to address Objective 2), and a multi-temporal monitoring experiment using SALCA and hemispherical photography at Delamere Forest in five different plots (to address Objective 3). Key findings relating to Objective 1 have highlighted the complexities of SALCA intensity response, such as the effect of internal temperature. As a result, a novel approach to radiometric calibration was developed using artificial neural networks which produced an apparent reflectance product with measured accuracy comparable with other approaches. A key conclusion of this research relating to Objective 2, is that dual-wavelength TLS has the potential to aid separation of leaf and wood material. However, there still remain significant ecological, instrumental, and processing challenges to be overcome. Temporally and vertically resolved plot measurements have provided a quantitative analysis of foliage dynamics to address Objective 3 and results have shown how this differences between species. The research presented in this thesis has explored the use of dual-wavelength full-waveform TLS for improved characterisation of forest vegetation. Future research priorities should focus on the radiometric calibration and investigation of other methods for leaf-wood separation to extend and complement this research