Inferring forest structure and disturbance dynamics by combining a canopy model with LiDAR remote sensing data
The major uncertainty identified through simulations of atmospheric carbon dioxide levels over the next century can be ascribed principally to the carbon flux associated with forests. This uncertainty must be addressed through the application of effective forest dynamic models; all the major ecological processes influencing net carbon release should be accounted for, from those acting at the level of the stand to those at the landscape level. This forms the basis of the ‘balanced complexity’ approach to modelling the global carbon cycle, from which the ultimate aim is to construct a data constrained model that incorporates all the important processes across the entire range of system levels.
My research aims to create a link model enabling recently developed canopy dynamics models, given various canopy metrics derived from LiDAR data, to be employed to infer forest stand structure. The resultant model will therefore provide a tool for predicting stand structure across a large forested area, from which current carbon stocks can be measured and future stocks can be predicted. This model will then be employed to map the stand structure of a large area of temperate and hardwood forest in central Ontario, Canada. This will then be used in combination with the locations of known disturbances across the region in an effort to characterise the effects of different disturbances. These characterised stands will consequently be used to test the hypothesis that the practiced harvesting regimes emulate the effects of the dominant natural disturbances in the region.