| Summary: | Thesis (M.S.) University of Alaska Fairbanks, 2024 Seasonal snowpack plays a critical role in hydrologic and ecologic processes. In boreal forest regions snow depth is known to be markedly different across land cover types. Identifying the vegetation metrics responsible for possible snow depth and snow water equivalent (SWE) spatial variability continues to be a challenge. Airborne lidar has advanced our understanding of links between forest snow distribution and vegetation impacts. This study analyzes high resolution (0.5 meter) lidar data sets acquired during NASA's SnowEx field campaign in Alaska and compares them statistically across the vegetation metrics of land cover class and lidar-derived canopy height. Airborne lidar data was collected for a boreal forest site, the Caribou Poker Creeks Research Watershed (CPCRW), during snow-off and peak snow-on accumulation in March of 2022 and May of 2023. Lidar snow depth (98 ± 15 cm) and canopy height maps, both at 0.5 m resolution, were created from lidar data sets. Lidar snow depth and canopy height maps were resampled to 1.5 m resolution to account for spatial autocorrelation. A total of 85.9 million lidar snow depth and canopy height values were available for this study. Three subsets totaling 6.1 million snow depths and canopy heights were processed to run the analysis. A USGS National Land Cover Database (NLCD) 2016 map of Alaska was used to determine land cover classes. Extensive in situ field snow depth measurements were collected concurrently with the peak snow-on lidar survey and were used to validate lidar accuracy. Analysis results from the three subsets showed statistically significant differences in median snow depths for all land cover classes and canopy height (p < 2.2e-16). Statistical comparison within land cover classes showed the largest significant difference in snow depths between shrub and deciduous forest (6-15 cm) and shrub and wetlands (7-14 cm). For canopy height classes, forest and treeless (12-14 cm) and forest and shrub/short ...
|
|---|