 |
|||||||||||||||||||
|
|||||||||||||||||||
|
|||||||||||||||||||
 |
|||||||||||||||||||
|
CAOS Workshop: Speakers AbstractMatthew Piggott: "Unstructured Meshes and Adaptivity for 3D Multi-Scale Ocean Modelling" Applied Modelling and Computation Group, Department of Earth Science and Engineering, Imperial College, London, UK; http://amcg.ese.ic.ac.uk; Email: m.d.piggott@imperial.ac.uk Over the past decade there has been growing interest in the use of unstructured mesh based methods in ocean modelling. Moving from structured to unstructured meshes offers many potential benefits. In particular it allows for an excellent representation of complex coastlines and bathymetry, and the ability to use wildly different resolutions in different parts of the domain. For example enhanced resolution may be employed to better resolve important localised phenomena such as boundary layer separations and overflows, and also regions of particular socio-economic or scientific importance. Importantly, unstructured meshes allow for the efficient representation of any interaction between a range of coupled spatial scales. Due to their geometric flexibility this can be achieved with unstructured meshes without resorting to the unsatisfactory approach of nesting due to the fact that smooth variations in mesh resolution are easily achieved Unstructured meshes are also the natural framework within which to formulate robust adaptive mesh capabilities. Extending the above multi-scale ability of the mesh, adaptive methods can be used to optimally resolve and track the formation and evolution of localised features in prior unknown and/or evolving locations. This would be impossible with any fixed mesh, whether unstructured or not. When utlising adaptive algorithms a model is able to automatically allocate computational resources in an optimal and dynamic manner, as dictated by evolving solution fields and estimates of model and discretisation errors. The aim is that this will lead to more efficient calculations, i.e. overall less mesh nodes would be required to yield a particular solution to a given accuracy; also for a given computational resource problems can be solved which are more complex than is currently feasible. In this presentation we will describe some of our experiences with constructuring a three-dimensional non-hydrostatic finite element ocean model using fully unstructured adaptive mesh techniques. The model is being developed as an open source community project with the aim of maximising flexibility in terms of what it can simulate without compromising on computational power and novelty. Of particular importance and focus will be mesh generation; mesh optimisation operations and mesh anisotropy; error measures; techniques for accurately describing model states close to hydrostatic and geostrophic balance on arbitrary irregular meshes; sharp interface representation; arbitrary mesh movement; and load-balanced parallelisation. With any new modelling approach validation is crucial. Here a range of standard benchmarks will be presented where comparisons with laboratory and other well-validated numerical codes are possible. Applications and validation against overflow problems, internal waves, flow past bathymetry, western boundary currents, basin and global scale tides, tsunami, wetting and drying in estuaries, and baroclinic circulation in the North Atlantic will also be presented. We will conclude by highlighting future plans for the model including the incorporation of biology and sediments, and also the inclusion of a sea ice module. We also plan to develop the capability to simulate the three-dimensional ocean circulation under ice shelves. Gavin Schmidt: "Challenges in Incorporating Ice-Sheet Models in Coupled GCMs" NASA Goddard Institute for Space Studies, New York, NY; http://www.giss.nasa.gov; Email: gschmidt@giss.nasa.gov Most IPCC-class Coupled GCMs do not include any significant interactive ice sheet component, and yet the sensitivity of those ice sheets are one of the key unanswered questions in projections of potential future climate. I will outline some of the constraints and challenges that incorporating ISMs will present and the prospect of near-term progress. Adrian Jenkins: "On the Sensitivity of Ice Shelf Basal Melting to the Temperature of the Antarctic Continental Shelf Seas" British Survey, Cambridge, United Kingdom; http://www.antarctica.ac.uk; Email: ajen@bas.ac.uk (Click Here for PDF Version) David Holland: "Ice-Ocean Physics to Include in Climate Models: Observations from the Ilulissat Ice Fjord" New York University, Courant Institute of Mathematical Science, New York, NY; http://www.cims.nyu.edu; Email: holland@cims.nyu.edu The Ilussiant Ice Fjord (a.k.a. Jakobshavn) may play an important role in connecting the waters of the North Atlantic to the Greenland Ice Sheet. At the eastern end of the fjord, the Jakobshavn Isbrae terminates abruptly at a deep ice front. At the western end, a shallow submarine sill partially separates the waters of Disko Bay from those of the fjord. To date there has been little knowledge gained of the bathymetry or oceanographic conditions in the fjord, and thus little ability to draw inferences about the possible influence of ocean waters on the observed retreating behavior of the Jakobshavn Isbrae. During June of 2007, we measured the depth on these measurements and describe future plans for using them in a computer modeling study in an effort to understand a possible coupling between oceanography and glaciology in this environment. Greg Flato: "Representation of the Cryosphere in the CCCma Climate Model" Canadian Centre for Climate Modelling and Analysis, Victoria, British Columbia, CANADA; http://www.cccma.ec.gc.ca; Email: greg.flato@ec.gc.ca The CCCma global climate model includes a dynamic-thermodynamic sea-ice model, based on the cavitating fluid scheme of Flato and Hibler (1992). The large ice sheets of Greenland and Antarctica are represented as ice surfaces in the land surface scheme, with snow accumulation and melt computed explicitly. In order to close the hydrological budget, net snow accumulation is routed to the ocean to crudely represent ice sheet flow and iceberg calving. Future plans include conversion to the more comprehensive CICE code for sea ice and, perhaps further in the future, a more realistic representation of the surface mass balance on, and ice discharge from, large ice sheets. Robert Bindschadler: "Ocean-Ice Sheet Interaction: The Glaciological Perspective" NASA Goddard Space Flight Center, Greenbelt, MD; http://www.gsfc.nasa.gov; Email: robert.bindschadler@nasa.gov Modern observations leave little doubt that ice sheets are affected by surrounding waters. This is part of a larger paradigm shift in the multiple roles of water in ice-sheet dynamics. In both Greenland and Antarctica, the largest changes in ice discharge are concentrated in the deepest outlet glaciers. That these changes are caused by ice shelf thinning brought about by ocean-induced melting is a tempting but as yet, unproven inference. Supporting evidence is the spatial pattern of change, theoretical work on the buttressing effect of ice shelves; and observations on the Antarctic Peninsula supporting the concept that ice shelves buttress the feeding glaciers. Also relevant may be the analogue of tidewater glaciers, a class of glaciers that can exhibit a drastic retreat mode quickly traversing along an overdeepened channel. Miles McPhee: "Heat and Mass Exchange at the Ice-Seawater Interface: What is known and what is not" McPhee Research Company, Naches, WA; Email: mmcphee@starband.net Heat and mass transfer at an ice/seawater boundary have been studied extensively in the context of sea ice formed at the ocean surface. Salt exerts significant control over the thermodynamic balance at the interface, as evidenced by the fact that basal melt of Arctic pack ice may reach several decimeters during summer, even though water temperature of the ice melt. The talk will review what has been learned from measurements under sea ice about both exchanges at the ice/ocean interface, and the impact of buoyancy flux from freezing or melting on turbulent exchange in the ice/ocean boundary layer (IOBL). Applying these principles to the exchanges between seawater and glacial ice presents some obvious (as yet unresolved) problems, including:
Frank Pattyn: "Higher-Order Modelling of Ice Sheet/Ice Stream/Ice Shelf Transition" Laboratoire de Glaciologie (DSTE), Universite Libre de Bruxelles, Brussels, Belgium; http://www.ulb.ac.de; Email: fpattyn@ulb.ac.de From an ice-dynamical point of view, the interaction of the ice sheet with ocean involves a stress transfer through longitudinal stress gradients across the grounding line, as well as the migration of the grounding line across the bedrock. Factors that influence both stress transfer and grounding line migration are basal ice shelf and grounding line melting, buttressing, basal sliding and other topographic controls of the embayment. Here we present the different computational aspects of including longitudinal stress schemes in ice sheet models together with the results of a recent intercomparison exercise based on different approximations to the Stokes equations. Marine ice sheet sensitivity to grounding line migration is investigated as well with respect to the different factors influencing enhanced grounding line retreat and glacier acceleration. Robert H. Thomas: "Ice Sheet Contributions to Sea Level Rise: Answers, and Questions from Remote Sensing" NASA Wallops, Wallops, VA; http://www.nasa.gov/centers/wallops/home/index.html; Email: robert_thomas@hotmail.com Changes in the volumes of the Greenland and Antarctic ice sheets have by far the largest potential impact on global sea-level rise (SLR). Comparatively low rates of SLR over the last two or three thousand years indicate that such changes have been slow, whereas recent measurements, from three independent techniques, indicate that the ice sheets are now contributing almost 1 mm/yr to SLR, and that this contribution is increasing with time. Despite this, IPCC and other model simulations of future ice-sheet contributions are quite modest, primarily because they cannot include the possibility of major changes in ice discharge along outlet glaciers and ice streams, changes which observations clearly show are already happening. A different approach, assuming SLR to be proportional to increasing global temperatures, yields a far more rapid future SLR. Here, I review available estimates of ice-sheet mass balance, I assess the implications for future ice-sheet behaviour if these higher predictions are correct, and I explore the possibility of even higher rates of SLR. Anny Cazenave: "Present-Day Sea Level Rise: Current Understanding and Major Uncertainties" LEGOS-CNES, Toulouse, FRANCE; http://www.cnes.fr; Email: anny.cazenave@cnes.fr Measuring sea level change and understanding its causes has improved considerably in the recent years, essentially because new in situ and remote sensing data sets have become available. Here we report on current knowledge of present-day sea level change. We first discuss sea level measurements (from satellite altimetry since 1993 and tide gauge reconstruction for the past few decades). Then we discuss recent estimates of climate/anthropogenic contributions (ocean thermal expansion, land ice and terrestrial waters). We focus on three different time spans: past four decades, 1993 - 2003 and since 2003. Since early 1993, sea level variations are accurately measured by Topex/Poseidon satellite altimetry, complemented for the recent years by Jason-1 altimetry. This ~15 year long data set indicates that, in terms of global mean, sea level is presently rising at a rate of ~3.1 +/- 0.4 mm/yr, a value significantly higher than the mean rate recorded by tide gauges over the past 50 years (on the order of 1.8 +/- 0.3 mm/yr). This eventually suggests that sea level rise is currently accelerating due to enhanced land ice melting and/or increased ocean warming. Owing to its global coverage, altimetry also reveals high regional variability in the rates of sea level change, some regions exhibiting sea level rates 5 - 10 times larger than the global mean. Quasi-global ocean temperature data sets collected since 1950 allow quantitative estimate of thermal expansion. Results indicate that over the past 40 years, thermal expansion accounts for ~0.4 mm/yr sea level rise, i.e. 23% of the observed rise. For the recent years (1993 - 2003), enhanced thermal expansion accounts for 1.6 mm/yr (i.e. 50% of observed sea level rise). For both periods, there is ~1.2 -1.5 mm/yr global mean residual not explained by thermal expansion, and thus attributed to land ice melt and water mass exchange with land. Mountain glaciers mass balance studies report contributions to sea level rise of 0.5 mm/yr and 0.8 mm/yr over 1960-2003 and 1993 - 2003 respectively. Mass balance of Greenland and Antarctica ice sheets have been the object of intensive studies in the recent years. Results indicate ice sheet contributions of 0.2 mm/yr and 0.4 mm/yr over the two time spans. We thus note that over 1993-2003, the sea level budget is almost closed (sum of climate contributions amounting 2.8 mm/yr). The small difference could well be attributed to land water storage change. Indeed recent GRACE-based estimates of total land water storage change (surface and underground waters plus snow; climatic plus anthropogenic) over 2002 - 2007 report a small positive contribution to sea level of 0.2 mm/yr. Although the value refers to the recent years only, it is of the right order of magnitude to explain the small difference between observed sea level rise and climate contributions over 1993 - 2003. For 1960-2003, we are unable to close the sea level budget, some 0.7 mm/yr remaining unexplained. However, we cannot exclude that thermal expansion was underestimated because of limited ocean temperature coverage of remote southern hemisphere and Deep Ocean in the past. We cannot exclude either some non negligible land water contribution, in particular of anthropogenic origin (reservoir filling, irrigation, groundwater mining, etc.). Since 2003, ocean thermal expansion (mainly based on the ARGO system) shows a plateau while satellite altimetry indicates that sea level is still rising. However, altimetry and tide gauges observations suggest that the observed rate of rise is slightly less that during the previous decade (on the order of 2 mm/yr between 2002 and 2007). Accounting for the recent acceleration of ice mass loss from Greenland and West Antarctica reported by very recent studies, and assuming a glacier contribution similar to the previous decade and accounting for the GRACE-based land water component, we note that the past 4 -5 years (reduced) rate of sea level rise can be explained solely by land ice and land waters. We also note that the latter contributions tend to become more and more important through time (0.7 mm/yr, 1.2 mm/yr and 1.9 mm/yr over the past 4 decades, 1993 - 2003 and 2003 - 2007 respectively). Bill Lipscomb: "Coupling Ice Sheet Models to Global Climate Models" Los Alamos National Laboratory, Los Alamos, NM: http://www.lanl.gov; Email: lipscomb@lanl.gov There is increased urgency to include realistic ice sheet models in coupled climate models for sea-level prediction. The LANL climate modeling group, which has developed the POP ocean model and CICE sea ice model, is now working to improve ice sheet models and include them in coupled simulations. I have recently added the GLLIMMER ice sheet model to the Community Climate System Model (CCSM). GLIMMER runs as a fifth physical component of CCSM (along with atmosphere, ocean, sea ice, and land) and exchanges fields with the land model via the coupler. The ice surface mass balance is computed in the land model on a coarse (~100 Km) grid in ~10 elevation classes and downscaled to the finer (~10 km) ice sheet grid. This innovative approach improves energy consistency, avoids code duplication, and reduces computational cost. Coupled climate simulations with a dynamic Greenland ice sheet will begin later this year. The current version of GLIMMER is based on the shallow-ice approximation and does not include ice shelves. We have begun a long-term effort with many collaborators to add ice streams and ice shelves (with higher-order dynamics), move toward finer grids, and improve the treatment of basal sliding, grounding-line migration, iceberg calving, and other key processes. The ice sheet model has not yet been coupled to the ocean. In response to theory and observations suggesting that the West Antarctic ice sheet could retreat abruptly, we have submitted a proposal to model ice sheet/shelf-ocean interactions at fine (~5 km) resolution. Todd Ringler will describe this project in a follow-up talk. Jesse V. Johnson: "Ice Sheet Models: Physical Basis, Numerical Treatments, and Software Development" University of Montana, Missoula, MT: http://www.umt.edu; Email: johnson@cs.umt.edu Ice sheet models are formulated from the conservation mass, momentum, and energy. The resulting equations are familiar to those accustomed to working with fluids, but distinctive due to the non-Newtonian rheology of ice and the formation of fast flowing ice streams. From the physics, one sees that are at least two areas that are marked by significant uncertainty. They are 1) prescription of the boundary conditions, and 2) simplification of the full set of equations as is appropriate for ice sheet, stream and shelf flow. Numerous numerical approaches have been used on the equations, but for cases where models have been directly compared, all are capable of correctly representing the behavior of the system. Presently, the chief numerical concerns include combining multiple regions of different flow into a single model while preserving the conservation of mass, and correctly tracking the migration of the grounding line. Ultimately, many of the concerns of ice sheet modeling relate to how science is conducted. It is increasingly clear than a comprehensive ice sheet model is beyond a single researcher's ability to construct and maintain. Additionally, as many of active areas of research are at the interface between ice sheet models and other Earth systems components, there is a need for a recognized, well documented ice model. There are several efforts gaining attention in this space, most notably GLIMMER and PISM. Both models are publicly accessible via the Internet, and open source. If this approach to science is to gain momentum, additional efforts must be made to assure that ice sheet models are appealing to the widest possible audience, and represent ice sheet modelings present state of the art. Stan Jacobs: "Ocean Observations near the Antarctic Ice Sheet" Lamont-Doherty Earth Observatory, The Earth Institute at Columbia University, New York, NY; http://www.Ideo.columbia.edu; Email: sjacobs@Ideo.columbia.edu Wherever the Antarctic Ice Sheet extends into the sea, it becomes vulnerable to ocean temperature, circulation and climate change. Across most of the deep circumpolar continental shelf break, a sharp landward transition occurs to ocean properties that are dominated by freezing and melting. This is less so in the Southeast Pacific, where 'warm' seawater intrusions and deep glacier grounding lines can drive basal melt rates more than 100 times higher than beneath the larger ice shelves. Over recent decades an increased melting rate of continental ice and a decreased formation rate of sea ice in the Amundsen and Bellingshausen Seas may account for observed ocean freshening downstream in the Ross Sea. Such freshening can in turn influence circulation, melting and sea level change. I will review Southern Ocean temperature and salinity measurements, mainly near the western WAIS, noting variability, forcing and impacts of potential interest to modeling work. Cornelis Vanderveen: "Subglacial Geology and Basal Boundary Conditions for Ice-Sheet Models" Center for Remote Sensing of Ice Sheet, University of Kansas, Lawrence, KS; http://www.ku.edu; Email: cjvdv@ku.edu Several recent studies have pointed to the influence that subglacial geology may have on the onset of fast flow. Observations suggest localized melting in northern Greenland that can only be explained if large spatial variations in geothermal heat flux are invoked. These variations can be caused by variations in the thickness of the crust under the ice, or by pronounced subglacial topographic relief, and can have an important effect on the thermal regime of the basal ice. At present, geothermal heat flux variations have been estimated for a few places in Green land and Antarctica, but for modeling the evolution of these ice sheets, better constraints on heat supplied to the base of the ice are needed. Airborne and satellite gravity and magnetic data offer opportunities to extrapolate geologic features from the margins into the ice-covered interior to obtain more detailed spatial models of inferred geothermal heat flux. Further characterization of the basal temperature regime can be inferred from airborne radar sounding. By recovering basal reflection characteristics, discrimination of subglacial water is feasible, allowing the extent of water under the ice sheets to be mapped. Such a map would provide an important validation for thermo-mechanical ice-sheet models. A second important control on fast flow and possibly grounding-line stability is the availability of subglacial sediments and sediment transport. Seismic profiling across the margin of a tributary ice stream in West Antarctica has provided evidence that the margin directly overlies the boundary of a sedimentary basin. Presumably, sedimentary till offers much reduced frictional resistance to ice flow, thereby allowing the glacier to reach greater speeds. As material is eroded and transported by the ice, net deposition may take place at the grounding line where the ice looses contact with its bed. Recent modeling studies suggest that this deposition may act to strabilize grounding lines of ice streams. Related to the question of subglacial sediment transport is the nature of the subglacial water drainage system. Several studies based on satellite observations posit that water moves beneath the ice streams through a series of discharge events through a "cascading" network of subglacial lakes, which suggests the need for new models for subglacial hydrology and water flow. Edward L. Bueler: "Dynamics, Numerics, and Plain Old Ignorance: Progress and Problems in Ice Sheet/Stream/Shelf Modelling" University of Alaska, Fairbanks, AK; http://www.uaf.edu; Email: ffelb@uaf.edu Ice sheet modeling is apparently still young, but I will attempt an outsider's overview which compares to other geophysical flows like ocean circulation and porous media. Ice sheets are hard to observe. The temperature within the body of the ice, the friction at the base, and whether ice is freezing or melting at a warm base (grounded and floating), are each hard to see but critical for prediction. Ice sheet modelers must do (at least appreciate!) inverse modeling, but still there will be few constraints on certain initial conditions. In fact, numerical ice sheet models are built to provide the missing "laboratory experiments" for continuum modelers. Continuum fluid dynamics issues include: How much shallowness?, When must membrane stresses be included?, How to model deformation of the basal ice and till below?, How to couple flow regimes together? Numerical issues include: How to approximate mass continuity (for the free surface of ice sheets)? Is verification really possible? What parallelizes? How do we see the continuum truth through the numerical confusion? Progress, not just problems, will be emphasized. Todd Ringler: "Modeling Ice Shelf-Ocean Interaction" Los Alamos National Laboratory, Alamos, NM; http://www.lanl.gov; Email: ringler@lanl.gov Simulation of coupled ice shelf-ocean dynamics presents a host of computational challenges. The system is characterized by various spatial scales, many of which are significantly smaller than can be resolved in global climate models. The complex geometries of embayments along with important subtleties in the ocean shelf structure might require horizontal resolutions of 5 km or less. In addition, significant resolution of the vertical structure of the ocean circulation in and around these ice shelves may be required to accurately capture the heat transport into and out of these embayments. To meet these challenges we are working to develop a flexible modeling framework that can be utilized as a regional coupled model, yet can be extented to global climate system modeling. The first aspec t of this framework is to allow the vertical structure of mass and momentum to be modeled on seperate vertical grids. Mass, along with density and associated tracer fields of temperature and salinity, will reside on an Arbitrary Eulerian Lagrangian (ALE) vertical grid. Similar to isopycnal methods, this method allows density drven currents to propagate without excessive numerical diffusion. Momentum will reside on a fixed, z-level Eulerian grid in order to limit discretization error along steep slopes. The second aspect of the domains. The variable resolution allows to focus computational resources toward those aspects of the system that are the most difficult to simulate, such as ice streams, grounding lines and topographically-constraint ocean circulations. The method should allow us to capture the relevant scales of ice shelf-ocean interaction, yet retain a computational model that is tractable for global climate system modeling.The presentation will motivate the modeling framework, provide examples where this framework will be most effective and summarize our progress to date. Waleed Abdalati: "Remote Sensing Contributions to Understanding Ice Sheet Behavior" NASA Goddard Space Flight Center, Greenbelt, MD; http://www.gsfc.nasa.gov; Email: waleed@icesat2.gsfc.nasa.gov Since the beginning of this decade, our understanding of the behavior of the Greenland and Antarctic ice sheets has been revolutionized by the contributions of remote sensing observations from satellite and aircraft. In addition to enabling observationally-based estimates of ice sheet contribution to sea level, these tools have shed important new insights into the mechanisms of ice sheet change. Moreover, they have revealed that present-day climate conditions have major impacts on ice sheet behavior in ways that were not fully appreciated in the past. As a result of these observations, we now know that the ice sheets to be subject to far-reaching non-linear effects, such as the enhanced discharge as floating ice around the ice sheet margins disappears and seasonal acceleration in response to the rapid movement of surface melt water to the ice/bedrock interfaces. Remote sensing has enabled us to study large-scale dynamics, surface melt, basal melting, accumulation, calving front movement, groundling line movement and other aspects of these vast frozen reservoirs. In so doing, these observations continue to change profoundly our understanding of how these ice sheets behave in the changing climate, and have opened new questions as to what the future conditions of these ice sheets may mean for sea level. Chris Little: "GFDL's Climate and Cryosphere Modeling Efforts" Princeton University, Princeton, NJ; http://www.princeton.edu; Email: cmlittle@princeton.edu An important part of NOAA Geophysical Fluid Dynamics Laboratory's (GFDL) mission is to be a world leader in the development of earth system models. GFDL is thus working towards the inclusion of a dynamic ice sheet model capable of simulating paleoclimate and future sea-level changes. Currently, GFDL's coupled models include geopotential and isopycnal coordinate ocean models coupled to a finite-volume atmospheric model via a dynamic-thermodynamic sea-ice model. Ice sheets are static and are part of the land model; mass fluxes to the ice are passed to the closest coastal grid point. GFDL is searching for a candidate to lead the land-ice model development effort. When implemented, a land-ice model will become part of the flexible modeling framework (FMS). The FMS and other GFDL infrastructure will likely ease integration difficulties and facilitate coupled cryosphere-climate simulations. In collaboration with GFDL scientists (in particular, Robert Hallberg and Anand Gnanadesikan), I have developed model code representing a static ice shelf in GFDL's isopycnal ocean model (HIM). The physical basis for the ice-ocean interface is heavily indebted to earlier work in MICOM (e.g. Holland and Jenkins 2001, Jenkins and Holland 1999). We are now conducting idealized, regional analyses of oceanic constraints on circulation and melting, the melt response to external forcing, and our ability to effectively model basal melting (i.e. what constitutes "sufficient" vertical and horizontal resolution). These efforts provide insight into ice shelf basal melting and a tested, scalable ice interface for future coupling to continental scale ice sheet models.Eric Rignot: "Radar Remote Sensing Observations of Ice Shelf Bottom Melting in Antarctica" Jet Propulsion Laboratory, Pasadena, CA; http://www.jpl.nasa.gov; Email: eric.rignot@jpl.nasa.gov In this talk, we will present an update on prior work comparing remote sensing estimates of bottom melting with thermal forcing from the ocean, and an estimate of the fraction of the Antarctic mass balance that is controlled by ice-ocean interactions, namely the fraction of ice removed by iceberg calving vs that removed from bottom melting of ice shelves. Jeff Ridley: "The Reversibility of a Decline in the Greenland Ice Sheet" Met Office Hadley Centre, Exeter, UK; http://www.metoffice.gov.uk; Email: jeff.ridley@metoffice.gov.uk Climate model experiments which have a coupled thermodynamic and dynamic ice sheet model have been used by a number of modelling centres to predict the future development of the Greenland ice sheet. Such simulations predict the demise of the Greenland ice sheet, under high climate forcing scenarios, within a few thousand years. However, it is unlikely that such high forcing can be sustained for such long periods and consequently it is possible that ice sheet melting may be reversed. To investigate the reversibility of decline of the Greenland ice sheet a series of GCM experiments is undertaken in which the ice sheet, initialised from various states obtained during its modelling decline, is forced with a pre-industrial climate. Ice sheet growth is slow compared with rates of change during a force melt, and consequently an asynchronously coupled experiment design was adopted. Preliminary results, which indicate that there are multiple 'stable' states for the ice sheet, are presented here. Tad W. Pfeffer: "Unresolved Physics in Basal and Englacial Dynamics" University of Colorado, Boulder, CO; http://www.colorado.edu; Email: pfeffer@tintin.colorado.edu Several critical processes intimately connected to glacier dynamic response remain poorly constrained by observation or theory and cannot be reliably implemented in glacier or ice sheet numerical models. Among these processes are 1) surface meltwater routing to glacier beds (where it may influence dynamics) or directly out of the glacier system (raising sea level); 2) the role of water and subglacial hydrology in modulating basal sliding; and 3) mechanics of iceberg calving. The details of surface meltwater routing are very poorly understood, and no analytical capacity exists to predict the englacial path or and seasonal) empirical rules exist to characterize subglacial sliding and iceberg calving, but no physically-based and observationally confirmed laws exist to implement these processes on length and time scales needed for numerical modeling. I review relevant observational data and discuss research priorities needed to bridge this gap. Tony Payne: "Progress in Developing a Higher-Order Model of the West Antarctic Ice Sheet" (Along with Steve Price, Los Alamos National Laboratory, USA; Andreas Vieli, University of Durham, UK; Poul Christoffersen, Scott Polar Research Centre, UK; and Marion Bougamont, Scott Polar Research Centre, UK) University of Bristol, Bristol, UK; http://www.bristol.ac.uk; Email: a.j.payne@bristol.ac.uk This talk will report progress made in developing a new model of the Antarctic ice sheet and discuss future plans. The following aspects of the project will be covered. 1. Development of a solver for the three-dimensional, first-order stress balance equations. The solver is now largely complete and initial work on the Greenland ice sheet using a 10km spatial resolution will be reported (~ 500,000 grid cells). Work on the vertical resolution required indicates that ~10 cells should be sufficient, although there are issues concerning the often ~100 fold difference in ice rheology though a thickness of ice. The model currently uses the SLAP Sparse Linear Algebra Package (Lawrence Livermore National Laboratory, 1998), however, a modern (parallel) library is likely to improve performance significantly (e.g., PETSc). 2. Ice-thickness evolution. One consequence of the move away from the simplest Shallow Ice Approximation is that ice-thickness evolution must be solved separately from the stress balance. Previously the highly reduced stress-balance equations could be substituted into the thickness evolution equation, resulting in a straight forward (albeit non-linear) parabolic equation. The different flow regimes within an ice sheet (inland ice dominated by diffusion; ice shelves dominated by advection; and ice streams lying between the two) make the application of a single transport solver problematic, however options include the semi-lagrangian incremental remapping of Dukowicz and Baumgardner (2000) as well as more traditional methods. 3. Nested grids. The computational expense of the first-order stress balance solver implies that a nested-grid strategy may be appropriate. Several package exist within the public domain however, care is needed in assessing their applicability to ice-sheet simulations. In particular, the need for high resolution around the entire grounding line is an issue for block-based methods. an initial scheme will be presented with estimates of the likely computing requirements. 4. Basal processes. The work of coupling the Tulaczyk-Christoffersen-Bougamont model of basal sediment process has started. The key characteristics of this model are a relation between sediment yield strength and void ration (based on laboratory results); and a model of sediment hydrology that links sediment void ratio to water pressure. Two versions of this model will be used: one which explicitly resolves the vertical movement of pore water within the subglacial sediment and a box model which does not. The model will be driven by melt/freeze rates generated by the ice sheet model and provide basal traction field to the stress balance solver. Careful attention will need to be paid to the manner in which basal tractions (and indeed the whole concept of plastic till) are applied.5. Basal hydrology. The dependence of basal sediment properties on the quantity of available basal water and/or its pressure, suggests that a model is required for the lateral flow of subglacial water and, in particular, the link between large-scale drainage (which may or may not be channelized) and local sediment pore water (the quantity actually controlling sediment rheology). Results from a simple routing algorithm will be shown, as well as their use in calibrating a parameterization of basal traction for West Antarctica. 6. Coupling to the ocean. A two-dimensional plume-based model has recently been applied to Pine Island's Ice Shelf. Initially, the intention will be to use this model as a method of determining rates of melt from and freeze on to the underside of the ice shelves (the primary coupling to the ocean). This decision stems principally from the ease of use of the two-dimensional model, however issues exist concerning the properties of entrained sub-shelf water and whether they can be adequately prescribed at the ice-shelf front alone, or whether three-dimensional flow effects within the cavity are important. Ayako Abe-Ouchi: "Ice Sheet Modelling with MIROC GCM for the Past Ice Sheet Change and the Future Prediction" University of Tokyo, Tokyo, Japan; http://www.u-tokyo.ac.jp; Email: abeouchi@ccsr.u-tokyo.ac.jp One of the challenges of earth system modeling is to simulate the change of ice sheet and sea level in the past and confirm the theories of ice age cycle by simulating it and to understand the uniqueness or necessity of the present state of climate and cryosphere. Here we simulate the glacial cycles and investigate the origin of saw-tooth shape 100ka cycle using a three dimensional ice sheet model with the input obtained by a global climate model (IcIES with MIROC GCM). The ice sheet model includes the thermo-mechanical coupling process of ice sheet with the process of delayed isostatic rebound with a typical time constant. In order to estimate the climate sensitivity to Milankovitch forcing and atmospheric CO2 indicated by ice core data we used an atmospheric GCM (part of MIROC GCM, ice age cycles with a saw-tooth shape 100ka cycle and sharp terminations, the major NH ice sheets volume and the geographical distribution at the glacial maximum are simulated. It is shown by sensitivity studies that the ice sheet change is highly sensitive to the treatment of albedo feedback, elevation/lapse rate feedback and desertification feedback. We discuss on several implications and prospect for modeling the future change of ice sheet using MIROC GCM and IcIES. |