Cross-sites analysis of snowpack depth from LiDAR in Southern Sierra Nevada

Abstract:

To investigate on the differences and similarities of snow depth spatial variability over different watershed areas, five sites in Southern Sierra Nevada Critical Zone Observatory were selected for creating the snow depth maps using the snow-on and snow-off LiDAR datasets. The snow-on data were collected during the snow-peak time in 2010, while the snow-off data were collected during the summer in the same year. By subtracting the digital elevation models (DEM) of the snow-off data from the snow-on point clouds, snow-depth maps for these sites were created. Furthermore, canopy height, slope, and aspect are also appended with the snow-depth for digging out the impact on the snow distribution from these topography features. From the results, the snow depth in the open area increases at 14-15 cm/100 m elevation increasing is consistent across areas in the elevation range from 1850m to 2700m, while The results under the canopy presented an increasing rate about 2 cm/100 m higher but with around 20 cm lower of snow depth compared to that in the open area. Other than elevation, aspect also has a tremendous effect on snow distribution with the result showing that the ground facing to the northeast direction always having more snow accumulated than other areas regardless of vegetation existence.

Even though the results reveal strong consistency of the vegetation impact on the snow depth across sites, only about 35% of total area is under canopy in forested areas and less than 30% of LiDAR beams could be returned from the ground under the canopy. The LiDAR might overestimate the snowpack volume but is still an important index for blending with ground data and data from remote-sensing satellites. Also, implied from the tight connection between snow depth and aspect, it is suggested that solar radiation, wind speed and direction, temperature, as well as other environmental factors are interacting with topography features and playing important roles in snowpack redistribution and ablation. Therefore, data blending research work will be necessary for people to have better understanding of snow dynamics.