Predicting Ice Sheet and Climate Evolution at Extreme Scales (PISCEES)

Mass loss from the Greenland and Antarctic ice sheets is accelerating. Although ice sheet models have improved in recent years, much work is needed to make these models reliable and efficient on continental scales and to quantify their uncertainties. We therefore propose to develop and apply robust, accurate, and scalable dynamical cores ('dycores') for ice sheet modeling on structured and unstructured meshes with adaptive refinements; to evaluate ice sheet models using new tools and data sets for verification and validation (V&V) and uncertainty quantification (UQ); and to integrate these models and tools in the Community Ice Sheet Model (CISM) and Community Earth System Model (CESM). Using improved estimates of icesheet initial conditions, we will simulate decade-to-century-scale evolution of the Greenland and Antarctic ice sheets, using CISM both in standalone mode and coupled to CESM. We aim to provide useful, credible predictions, including uncertainty ranges, of future ice-sheet mass loss and resulting changes in climate and sea level. Building on recent successes of SciDAC and the Ice Sheet Initiative for Climate Extremes (ISICLES), we will develop two dycores: (1) a finite-volume dycore on a structured mesh, using the Chombo adaptive mesh refinement software framework, and (2) a finite-element dycore on an unstructured mesh, using the Model for Prediction Across Scales (MPAS) framework and Trilinos software packages. Both dycores will include a hierarchy of Stokes and higher-order solvers that can be applied at variable resolution in different regions. We will develop stable, efficient numerical schemes for ice-thickness evolution and will implement realistic, physics-based basal boundary conditions. These models will be engineered to optimize performance on new high-performance computers with heterogeneous architectures. We will also develop tools and frameworks for V&V and UQ. We will create verification test suites consisting of analytical and manufactured solutions, and we will compile the best available data sets for model validation. These tools will be assembled in a post-processing package that can be used to formally evaluate CISM in standalone runs and as a CESM component. Observational data sets will provide a target for initialization of the coupled model; we will spin up ice sheets to a state that closely resembles present conditions, while remaining in near balance with CESM surface forcing. Using the DAKOTA framework, we will develop non-intrusive and intrusive (adjoint-based) methods to quantify uncertainties associated with physics parameters, initial conditions, boundary forcing, and model structure. We will propagate these uncertainties in forward models to estimate confidence ranges for projections of ice-sheet evolution and sea-level rise. PISCEES will provide a coherent structure for ongoing collaboration among glaciologists, climate modelers, and computational scientists. We will work closely with the SciDAC institutes (FASTMath, QUEST, and SUPER) and the community of CISM and CESM developers. All source code and simulation results will be freely and publicly available.