| This study demonstrates a new method for analyzing discrete lidar data using pseudo-waveforms and a set of metrics designed to exploit the physical characteristics of these waves for the estimation of forest and canopy structure. Advantages of this method include: 1) model fits that were better than those created using quantile-based metrics; 2) potential for new metrics that capture unique patterns within the waves; 2) ability to explain metric selection based on structural forest characteristics; and 3) lower correlation among independent variables.;Lidar-acquired Canopy structure information was used in the development of a snowpack evolution model (SEM), which was designed to use a reduced set of site and climate data and be sensitive to a wide range of forest types without requiring specific information about forest age or composition. The model used remotely sensed and online data to focus on the major factors responsible for snowpack formation and ablation. Modeled results followed the general trend of observed snowpack evolution, with the best agreement between the two occurring during the melt period. Moreover, model simulations indicated that both canopy closure and PAI are important factors in determining the timing of snowmelt, its rate of delivery, and the total amount generated.;The SEM was enhanced to simulate snowpack formation and melt in a watershed with variable topography and heterogeneous forest cover. This model was then linked to a simple routing model to simulate the effects that forest structure can have on snowpack-derived streamflow in a forested watershed in northern Wisconsin, U.S.A. The reduced parameter set and model's ability to operate at fine spatial and temporal scales made it better suited to this type of simulation than existing hydrological models, which rarely consider land cover types beyond broad taxonomic classes, and typically operate at coarse spatial and temporal scales. The modeled flows followed the general trend measured during the winter of 2007-2008, with conspicuous divergences that correlated with rain-on-snow events that were not simulated by this version of the model. Despite this shortcoming, this model provides a greater understanding of how forest canopy structure can affect snowmelt-derived streamflows. |
