In the NEMOSID project, I first looked at the physics community's state-of-the-art models for granular mechanics. These mechanical models were promising, as they reconcile observations with physical behavior deemed realistic for general sediment mechanics. However, the models were untested and uncalibrated for glacial sediments. The next task consisted of collecting glacial sediment in the field and performing rigorous laboratory studies where specialized geotechnical equipment deformed the sediment under conditions similar to those in glacial environments. I used the laboratory results to tweak the granular mechanics model. In the process, I discovered that changes in water pressure are, by far, the most critical factor controlling the magnitude of sediment transport. This realization implies that daily glacier melt and tidal oscillations may be the primary governers of how glacial landforms emerge. Next, I put the calibrated granular mechanics model into a popular ice-sheet model called PISM (Parallel Ice Sheet Model). Climate modelers worldwide use PISM to understand past glaciations and predict the future of the Antarctic and Greenland ice sheets in a warming climate. This task required diving deep into the source code of PISM to understand its internals and then figure out how to bolt my model into it. I found two solutions and published them in open repositories so that other ice-sheet modelers can benefit from the advance. The advance in PISM means that we can now simulate how glaciers reshape their beds and test the influence of this process on sea-level rise. I modeled various glacial settings and found that sediment transport results in glacial landforms, as seen in the geological record from past glaciations. Additionally, the formation of sedimentary landforms can conditionally stabilize ice sheets against sea level rise.
