Asking the pikas politely…
Estimating snow mechanical properties – such as elastic modulus, stiffness, and strength – is important for understanding how effectively a vehicle can travel over snow-covered terrain. Vehicle instrumentation data and observations of the snowpack are valuable for improving the estimates of winter vehicle performance. Combining in-situ and remotely-sensed snow observations, driver input, and vehicle performance sensors requires several techniques of data integration. I explored correlations between measurements spanning from millimeter to meter scales, beginning with the SnowMicroPenetrometer (SMP) and instruments applied to snow that were designed for measuring the load bearing capacity and the compressive and shear strengths of roads and soils. The spatial distribution of snow’s mechanical properties is still largely unknown. From this initial work, I determined that snow density remains a useful proxy for snowpack strength. To measure snow density, I applied multi-sensor electromagnetic methods. Using spatially distributed snowpack, terrain, and vegetation information developed in the subsequent chapters, I developed an over-snow vehicle performance model. To measure the vehicle performance, I joined driver and vehicle data in the coined Normalized Difference Mobility Index (NDMI). Then, I applied regression methods to distribute NDMI from spatial snow, terrain, and vegetation properties. Mobility prediction is useful for the strategic advancement of warfighting in cold regions. The security of water resources is climatologically inequitable and water stress causes international conflict. Water resources derived from snow are essential for modern societies in climates where snow is the predominant source of precipitation, such as the western United States. Snow water equivalent (SWE) is a critical parameter for yearly water supply forecasting and can be calculated by multiplying the snow depth by the snow density. In this work, I combined high-spatial resolution light detection and ranging (LiDAR) measured snow depths with ground-penetrating radar (GPR) measurements of two-way travel-time (TWT) to solve for snow density. Then using LiDAR derived terrain and vegetation features as predictors in a multiple linear regression, the density observations are distributed across the SnowEx 2020 study area at Grand Mesa, Colorado. The modeled density resolved detailed patterns that agree with the known interactions of snow with wind, terrain, and vegetation. The integration of radar and LiDAR sensors shows promise as a technique for estimating SWE across entire river basins and evaluating observational- or physics-based snow-density models. Accurate estimation of SWE is a means of water security. In our changing climate, snow and ice mass are being permanently lost from the cryosphere. Mass balance is an indicator of the (in)stability of glaciers and ice sheets. Surface mass balance (SMB) may be estimated by multiplying the thickness of any annual snowpack layer by its density. Though, unlike applications in seasonal snowpack, the ages of annual firn layers are unknown. To estimate SMB, I modeled the firn depth, density, and age using empirical and numerical approaches. The annual SMB history shows cyclical patterns representing the combination of atmospheric, oceanic, and anthropogenic climate forcing, which may serve as evaluation or assimilation data in climate model retrievals of SMB. The advancements made using the SMP, multi-channel GPR arrays, and airborne LiDAR and radar within this dissertation have made it possible to spatially estimate the snow depth, density, and water equivalent in seasonal snow, glaciers, and ice sheets. Open access, process automation, repeatability, and accuracy were key design parameters of the analyses and algorithms developed within this work. The many different campaigns, objectives, and outcomes composing this research documented the successes and limitations of multi-sensor estimation techniques for a broad range of cryosphere applications.
Items connected by shared entities, co-authorship, citations, or semantic similarity.