NGD BISICLES

Dynamic Land-Ice Modeling

On average, Earth’s glaciers and ice sheets are shrinking. It is estimated that Antarctica holds about 70 percent of the world’s freshwater and 90 percent of the total global ice mass, while Greenland contains much of the remaining 10 percent. Understanding how these huge ice reserves respond to changing earth systems is critical for projections of sea level rise (SLR) in the 21st century and beyond. Because they interact with the ocean, marine ice sheets are particularly vulnerable to rapid change. In these ice sheets, ice flows in relatively fast-moving ice streams from the interior toward the ocean and into large floating ice shelves, which in turn push back against (“buttress”) the flow of the ice streams. The point at which the ice separates from the land and begins to float is known as the “grounding line” (GL). As ice shelves weaken and collapse due to mechanisms like warm-water incursions or surface melting that leads to crevassing and hydrofactures, the buttressing effect is lost and the feeder ice streams accelerate and thin. This can lead to dramatic grounding line retreat, a greater rate of ice loss, and amplified contributions to sea level rise.

A computer rendering illustrates the present-day Antarctic ice velocity field as computed by the BISICLES model. The inset image at the left details adaptive mesh refinement placement of fine spatial resolution near grounding lines (red) at the Pine Island Glacier, currently the fastest melting glacier in Antarctica, responsible for about 25 percent of Antarctica’s ice loss.

To meet this challenge, DOE’s earth and environmental system modeling has been developing ice sheet models, coupling them into the full earth system, and doing experiments to see how the ice sheets respond under changing environmental conditions. Much of this work has been done in collaboration with the Advanced Scientific Computing Research (ASCR) Office, under the Scientific Discovery through Advanced Computing (SciDAC) program, which supports partnerships between computational experts and earth system modelers. One outcome of this partnership is the Berkeley Ice Sheet Initiative for CLimate ExtremeS (BISICLES) ice sheet model. BISICLES uses DOE-developed adaptive mesh refinement (AMR) techniques to enable fully resolved modeling of the marine ice-sheet dynamics found in Antarctica and Greenland.

Vulnerability of the Antarctic ice shelf to regional ice shelf collapse – Orange circles indicate potential sea level rise contributions resulting from localized ice shelf collapse in each Antarctic sector, denoted by thick black lines. Blue colormap shows present-day ice velocity field, as computed by BISICLES. From Martin, Cornford, and Payne, GRL (2019)

There is substantial evidence that accurately modeling the relevant ice sheet dynamics requires very fine spatial and temporal resolution. Using AMR, the BISICLES ice sheet model dynamically applies high resolution only to where it is needed to resolve the dynamics of the ice sheet. This allows accurate and efficient projections of ice-sheet response to climate forcing and the resulting contributions to sea level rise at a tractable computational cost. For example, Antarctica’s contribution to sea level rise is expected to be dominated by the interaction of the marine ice sheets of the West Antarctic Ice Sheet and the ocean. If the most vulnerable ice in Antarctica is lost, global sea level is projected to rise by as much as four meters, impacting worldwide populations, economies, and the environment. For an example of BISICLES simulations of Antarctic ice sheet response, see this blog post which explains in more detail the following two videos.

Coupling dynamic land ice models fully into the climate system is an important target for the E3SM project, and efforts have been underway for some time to implement this coupling. As one of the dynamical cores developed by the PISCEES and PRoSPect ice sheet modeling SciDAC Application Partnerships, the BISICLES model is a natural fit with the E3SM project. Scientist are implementing full coupling of BISICLES into E3SM, along with the appropriate verification testing and initialization needed to make it a reliable and useful addition to the E3SM modeling system.

Movie frame from a coupled ice-ocean model simulation of the Antarctic ice sheet and the Southern Ocean: The white-light blue colors (as shown in the upper-left legend) indicate highest velocities of the grounded ice, while red-yellow (as shown in the upper-right legend) indicates the highest melt rates.

Built on the SciDAC-supported Chombo framework, BISICLES uses block-structured adaptive mesh refinement (AMR) to dynamically and adaptively deploy very fine spatial resolution where needed to accurately resolve the dynamics of features like grounding lines and ice streams. Recent work by Cornford, Martin, et al (2016) demonstrated that very fine (sub-km) resolution is essential to accurately resolve the dynamics of marine ice sheets, like the West Antarctic Ice Sheet. Without such fine resolution, models may compute solutions which are qualitatively and quantitatively incorrect. Uniformly resolving the entire Antarctic ice sheet at such fine resolution would be prohibitively expensive, with an enormous amount of unnecessary computational effort. Because the fine resolution must follow grounding lines as they retreat over hundreds of kilometers, BISICLES is an ideal tool for investigating the response of the Antarctic Ice sheet to marine forcing and the consequent grounding line retreat and contribution to sea level rise. AMR enables the required resolution while also maintaining a reasonable time to solution.

Adding BISICLES as an ice sheet model option will enhance E3SM’s predictive capability for understanding the role of ice sheets in the fully-coupled climate system and the resulting contributions to SLR in key ways, while simultaneously providing a useful companion to the in-development MALI model for verification and validation purposes. Access to two ice sheet models coupled to the same climate model (E3SM) will allow for more confidence in predictions of SLR. Two ice sheet models will also provide estimates for the structural uncertainty of SLR predictions from each of these capabilities which would make E3SM unique among large ESM efforts within the international community and would greatly improve E3SM’s ability to answer its cryosphere-related questions.

Developers are currently coupling BISICLES to E3SM using the existing E3SM coupling infrastructure. Scientists are drawing on their previous experience coupling BISICLES to the POP2X ocean model, while leveraging the significant progress already made toward coupling the MALI model to E3SM. This work is coordinated closely with the “Ecosystem” ProSPect BISICLES development effort at LBNL.

Get BISICLES

Instructions for downloading the BISICLES ice-sheet modeling code and software documentation are available at http://bisicles.lbl.gov/.

References

• S.L. Cornford, D.F. Martin, D.T. Graves, D.F. Ranken, A.M. Le Brocq, R.M. Gladstone, A.J. Payne, E.G. Ng, W.H. Lipscomb,  “Adaptive mesh, finite volume modeling of marine ice sheets”, Journal of Computational Physics, 232(1):529-549, 2013. https://doi.org/10.1016/j.jcp.2012.08.037.
• S. L. Cornford, D. F. Martin, A. J. Payne, E. G. Ng, A. M. Le Brocq, R. M. Gladstone, T. L. Edwards, S. R. Shannon, C. Agosta, M. R. van den Broeke, H. H. Hellmer, G. Krinner, S. R. M. Ligtenberg, R. Timmermann, D. G. Vaughan,  “Century-scale simulations of the response of the West Antarctic Ice Sheet to a warming climate”, The Cryosphere, August 18, 2015, https://doi.org/10.5194/tc-9-1579-2015.
• S.L. Cornford, D.F.Martin, V. Lee, A.J. Payne, E.G. Ng, “Adaptive mesh refinement versus subgrid friction interpolation in simulations of Antarctic ice dynamics”, Annals of Glaciology, September 2016, 57 (73), https://doi.org/10.1017/aog.2016.13.
• Daniel F. Martin, Stephen L. Cornford, Antony J. Payne, “Millennial‐scale Vulnerability of the Antarctic Ice Sheet to Regional Ice Shelf Collapse”, Geophysical Research Letters, January 9, 2019, https://doi.org/10.1029/2018GL081229.