Erosion is accelerating along many stretches of the coastal Arctic, putting critical infrastructure at risk and threatening local communities. These permafrost-laden coastlines are increasing vulnerable to erosion due to declining sea ice and increasing duration of open water, more frequent storms during ice-free periods, and warming permafrost soils. However, predicting shoreline erosion rate remains extremely challenging because of the highly non-linear behavior of the coupled and changing environmental system. Although the Arctic comprises one-third of the global coastline and has some of the fastest eroding coasts, current tools for quantifying permafrost erosion are unable to explain the episodic, storm-driven erosion events. In this talk I will present the details of the development and calibration efforts for the Arctic Coastal Erosion (ACE) Model at Sandia National Laboratories. The ACE Model is a multi-physics numerical tool that couples oceanographic and atmospheric conditions with a terrestrial permafrost domain to capture the thermo-chemo-mechanical dynamics of erosion along permafrost coastlines. It is based on the finite-element method and solves the governing equations for conservation of energy (heat conduction with phase change), and conservation of linear and angular momenta using a plasticity material model. Oceanographic and atmospheric boundary conditions force evolution of a permafrost environment, consisting of porous media made of sediment grains and pore fluid. An oceanographic modeling suite (external software packages) produces time-dependent water level, temperature, and salinity boundary conditions for the terrestrial domain. Atmospheric temperature is obtained from the ECMWF Reanalysis v5 (ERA5) dataset. Driven by these boundary conditions, 3-D solutions of temperature, stress, and displacement develop in the terrestrial domain in response to the plasticity model that is controlled by the frozen water content. Material is removed when the stress or strain within an element exceeds the yield strength or strain limit of the material and is followed by grid adaptation that captures the new geometry. This modeling approach enables failure from any allowable deformation (e.g., block failure, slumping, thermal denudation) and can treat erosion behavior over single events, seasonally, or over several years.

BIO: Jennifer Frederick earned her Ph.D. in 2013 from the Earth & Planetary Science Department at U.C. Berkeley, a M.S. in Mechanical Engineering at U.C. Berkeley, and a B.S. in Bioengineering from the University of Illinois at Chicago. She is a computational geoscientist who develops open-source, massively parallel, subsurface simulators for multiphase, multicomponent, and multiscale reactive flow and transport processes in the subsurface to help solve problems in national security at Sandia National Laboratories. She is a project leader under several laboratory-wide strategic initiatives, including the Arctic Science and Security Initiative, and more recently the Climate Change Security Strategic Initiative, with expertise on climate change and national security concerns such as greenhouse gas emissions from permafrost-associated gas hydrate, terrestrial and submarine permafrost, and infrastructure issues associated with Arctic coastal erosion. In the field, she uses distributed temperature sensing (DTS) and distributed acoustic sensing (DAS) to better understand the character and evolution of submarine permafrost and seabed processes.