| Summary: | Thesis (M.S.) University of Alaska Fairbanks, 2022 Understanding and modeling the permafrost system, hydrologic cycle, energy balance, and biologic systems in the Arctic are dependent, in part, on snow depth and snow distribution. Point-source snow measurements provide ground-truth observations of snow depth and snow water equivalent, although these measurements may be limited in their spatial and temporal distributions. Satellite-derived remote sensing products and gridded model output provide spatial coverage of snow properties, but their applicability is affected by their balance of resolution, computational speed, and accuracy confidence. The goal of this research is to assess the performance of three snow data products derived from remote sensing techniques as well as model output across the North Slope of Alaska with the International Land Model Benchmarking (ILAMB) Project software. Historic ground-based snow data, collected by agencies, academia, and industry, and dating from 1902 to 2021, was curated to create an ILAMB-compatible benchmark dataset for end-of-winter (EOW) snow depth and snow water equivalent (SWE) for the evaluation of the three snow data products: Canadian Sea Ice and Snow Evolution (CanSISE) network SWE; Arctic Boreal Vulnerability Experiment (ABoVE) snow depth; and Energy Exascale Earth System Model (E3SM) Earth Land Model (ELM) snow depth. The ILAMB evaluation results showed that the ABoVE data product is effective in providing the average EOW snow depth for regions of the North Slope but lacks representation of interannual and spatial variability of snow depth. Comparatively, the CanSISE data product and ELM results are inaccurate in magnitude for applicability across the North Slope of Alaska in addition to lacking representation of snow condition spatial variability. In interpreting ILAMB results, factors to consider were representation bias from inconsistent benchmark site distribution throughout the evaluated time period, the range of dates considered to represent the spring ...
|
|---|