NSFGEO-NERC: Adjoint tomography of mantle viscosity using deglacial sea level observations

The viscosity of Earth's mantle varies radially and laterally as a result of its complex thermal structure. While it is known from seismology, geodynamics, mineral physics, and sea level observations that these variations exist, they are difficult to constrain. Through glacial isostaticadjustment (GIA) - the viscoelastic response of the Earth to waxing and waning ice sheets - this incomplete knowledge propagates into questions of past ice sheets and sea level change. This proposal constitutes an effort to better understand Earth rheology and its implications for cryosphere evolution. The novel aspect proposed here is to invert sea level, geodetic, and gravitational observationsfor rheology and deglacial ice sheet changes using gradient-based optimization. Model gradients will be efficiently calculated using the adjoint method. This framework allows us to move beyond a limited set of forward simulations and enables us for the first time to efficiently assimilate sea level data into a 3D GIA model. We will use a newly compiled dataset with over 13,000 datapoints of deglacial sea level to produce the first tomographic image of Earth's internal viscosity structure. We will furthermore use this approach to produce ice sheet reconstructions that are consistent with 3D Earth models. In addition to exploring trade-offs between the Earth and ice structure, we will develop and implement second-order adjoint equations to assess uncertainty propagation. Our model output will allow us to address two targeted research questions, (1) the amount of melt or readvance of different ice sheets during the Holocene and (2) the present-day contribution of GIAand its uncertainty to sea level change in major coastal cities.