Passive and Active Microwave Remote Sensing and Modeling of Layered Snow

• Main Author: Fuller, Mark Christopher
• Other Authors: Yackel, John, Derksen, Chris
• Format: Doctoral or Postdoctoral Thesis
• Language: English
• Published: Graduate Studies 2015
• Subjects: Geophysics; Physical Geography; Remote Sensing; Snow; Snow Physical Modeling; Snow Active Microwave Remote Sensing; Snow Passive Microwave Remote Sensing; Snow Layering; Layered Snow over Land; Layered Snow over Sea Ice; Snow Assimilation Technique; Snow Water Equivalent; Snow Microstructure; Snow Stratigraphy; Snow Microwave Backscatter Modeling; Snow Microwave Emission Modeling; Depth Hoar; Ice Layers; Wind Slab; Sea ice
• Online Access: http://hdl.handle.net/11023/2394; https://doi.org/10.11575/PRISM/27265

This thesis investigates the effects of complexly-layered snow on passive and active microwave remote sensing observations and models, employing detailed in-situ geophysical measurements over various landcover types. First, I present observed and simulated C-band backscatter signatures for complexly-layered snow on smooth, landfast first-year sea ice. Detailed in-situ measurements describe snow structure. A multilayer backscatter model is used to assess the impacts of layered components. The backscatter from a complexly-layered snow cover on smooth first-year sea ice is higher than from a simple snow cover. Sensitivity analysis suggests that rough ice layers within the snow cover and superimposed at the snow-ice interface influence brine volume, and are mechanisms that increase surface and volume scattering. This has implications for sea ice mapping, geophysical inversion, and snow thickness retrievals. Second, I present a snow layer excavation experiment to compare observed and modeled brightness temperatures at 19 and 37 GHz, with regard to snow water equivalent (SWE), snow type, grain size, and layered structure. In-situ snow measurements forced a multi-layer snow emission model. Emission scattering from depth hoar was disproportionate to its SWE contribution, and masked observed scattering contributions from upper snow layers. The simulations diverged from observations above 130 mm SWE, as simulations did not capture snowpack emission. This may impact the effective grain size optimization process of the GlobSnow assimilation technique. Third I present the application of meteorological reanalysis data to the SNTHERM snow model for comparison with in-situ snow measurements, and observed and simulated C-band backscatter of snow on first-year sea ice. Application of in-situ salinity profiles to one SNTHERM snow profile resulted in simulated backscatter close to in-situ measurements. In other cases simulations remained 4 to 6 dB below observations. Although, there is the possibility of achieving comparable ...