Spatial probabilistic calibration of a high-resolution Amundsen Sea Embayment ice-sheet model with satellite altimeter data

Abstract. Probabilistic predictions of the sea level contribution from Antarctica often have large uncertainty intervals. Calibration with observations can reduce uncertainties and improve confidence in projections, particularly if this exploits as much of the available information as possible (such as spatial characteristics), but the necessary statistical treatment is often challenging and can be computationally prohibitive. Ice sheet models with sufficient spatial resolution to resolve grounding line evolution are also computationally expensive. Here we address these challenges by adopting a novel dimension-reduced approach to calibration combined with statistical emulation of the adaptive mesh model BISICLES. We find the most likely contribution to global mean sea level rise from the Amundsen Sea Embayment (ASE) over the next 50 years is 10.4 [0.6, 23.3] mm (mode and 5–95 % probability interval), a substantial reduction in uncertainty from the uncalibrated estimates of 9.6 [−5.9, 78.2] mm. We predict retreat of the grounding line along most parts of the ASE coast with high confidence, with a maximum inland extent of around 28 km at Smith Glacier. The model behaviour is much more consistent with observations if, instead of Bedmap2, a modified bedrock topography is used that most notably removes a topographic rise near the initial grounding line of Pine Island Glacier, though this does influence the future mass loss less than basal traction and viscosity scaling parameters. The ASE dominates the current Antarctic sea level contribution, but other regions have the potential to become more important on centennial scales. These larger spatial and temporal scales would benefit even more from methods of fast but exhaustive model calibration. Our approach therefore has the potential to improve projections for the Antarctic ice sheet on continental and centennial scales by efficiently improving our understanding of model behaviour, and substantiating and reducing projection uncertainties.

Download Full-text

Spatial probabilistic calibration of a high-resolution Amundsen Sea Embayment ice sheet model with satellite altimeter data

The Cryosphere ◽

10.5194/tc-14-1459-2020 ◽

2020 ◽

Vol 14 (5) ◽

pp. 1459-1474

Author(s):

Andreas Wernecke ◽

Tamsin L. Edwards ◽

Isabel J. Nias ◽

Philip B. Holden ◽

Neil R. Edwards

Keyword(s):

Sea Level ◽

Principal Components ◽

Statistical Treatment ◽

Ice Sheet ◽

Altimeter Data ◽

Model Parameters ◽

Amundsen Sea ◽

Thickness Change ◽

Wide Range ◽

Grounding Line

Abstract. Probabilistic predictions of the sea level contribution from Antarctica often have large uncertainty intervals. Calibration of model simulations with observations can reduce uncertainties and improve confidence in projections, particularly if this exploits as much of the available information as possible (such as spatial characteristics), but the necessary statistical treatment is often challenging and can be computationally prohibitive. Ice sheet models with sufficient spatial resolution to resolve grounding line evolution are also computationally expensive. Here we address these challenges by adopting and comparing dimension-reduced calibration approaches based on a principal component decomposition of the adaptive mesh model BISICLES. The effects model parameters have on these principal components are then gathered in statistical emulators to allow for smooth probability density estimates. With the help of a published perturbed parameter ice sheet model ensemble of the Amundsen Sea Embayment (ASE), we show how the use of principal components in combination with spatially resolved observations can improve probabilistic calibrations. In synthetic model experiments (calibrating the model with altered model results) we can identify the correct basal traction and ice viscosity scaling parameters as well as the bedrock map with spatial calibrations. In comparison a simpler calibration against an aggregated observation, the net sea level contribution, imposes only weaker constraints by allowing a wide range of basal traction and viscosity scaling factors. Uncertainties in sea level rise contribution of 50-year simulations from the current state of the ASE can be reduced with satellite observations of recent ice thickness change by nearly 90 %; median and 90 % confidence intervals are 18.9 [13.9, 24.8] mm SLE (sea level equivalent) for the proposed spatial calibration approach, 16.8 [7.7, 25.6] mm SLE for the net sea level calibration and 23.1 [−8.4, 94.5] mm SLE for the uncalibrated ensemble. The spatial model behaviour is much more consistent with observations if, instead of Bedmap2, a modified bedrock topography is used that most notably removes a topographic rise near the initial grounding line of Pine Island Glacier. The ASE dominates the current Antarctic sea level contribution, but other regions have the potential to become more important on centennial scales. These larger spatial and temporal scales would benefit even more from methods of fast but exhaustive model calibration. Applied to projections of the whole Antarctic ice sheet, our approach has therefore the potential to efficiently improve our understanding of model behaviour, as well as substantiating and reducing projection uncertainties.

Download Full-text

Ocean-forced evolution of the Amundsen Sea catchment, West Antarctica, by 2100

The Cryosphere ◽

10.5194/tc-14-1245-2020 ◽

2020 ◽

Vol 14 (4) ◽

pp. 1245-1258

Author(s):

Alanna V. Alevropoulos-Borrill ◽

Isabel J. Nias ◽

Antony J. Payne ◽

Nicholas R. Golledge ◽

Rory J. Bingham

Keyword(s):

Sea Level ◽

21St Century ◽

Adaptive Mesh Refinement ◽

Ice Sheet ◽

Adaptive Mesh ◽

Ocean Temperature ◽

Ice Shelf ◽

Ice Streams ◽

Amundsen Sea ◽

Grounding Line

Abstract. The response of ice streams in the Amundsen Sea Embayment (ASE) to future climate forcing is highly uncertain. Here we present projections of 21st century response of ASE ice streams to modelled local ocean temperature change using a subset of Coupled Model Intercomparison Project (CMIP5) simulations. We use the BISICLES adaptive mesh refinement (AMR) ice sheet model, with high-resolution grounding line resolving capabilities, to explore grounding line migration in response to projected sub-ice-shelf basal melting. We find a contribution to sea level rise of between 2.0 and 4.5 cm by 2100 under RCP8.5 conditions from the CMIP5 subset, where the mass loss response is linearly related to the mean ocean temperature anomaly. To account for uncertainty associated with model initialization, we perform three further sets of CMIP5-forced experiments using different parameterizations that explore perturbations to the prescription of initial basal melt, the basal traction coefficient and the ice stiffening factor. We find that the response of the ASE to ocean temperature forcing is highly dependent on the parameter fields obtained in the initialization procedure, where the sensitivity of the ASE ice streams to the sub-ice-shelf melt forcing is dependent on the choice of parameter set. Accounting for ice sheet model parameter uncertainty results in a projected range in sea level equivalent contribution from the ASE of between −0.02 and 12.1 cm by the end of the 21st century.

Download Full-text

Ocean forced evolution of the Amundsen Sea catchment, West Antarctica, by 2100

10.5194/tc-2019-202 ◽

2019 ◽

Author(s):

Alanna V. Alevropoulos-Borrill ◽

Isabel J. Nias ◽

Antony J. Payne ◽

Nicholas R. Golledge ◽

Rory J. Bingham

Keyword(s):

Sea Level ◽

21St Century ◽

Adaptive Mesh Refinement ◽

Ice Sheet ◽

Adaptive Mesh ◽

Ocean Temperature ◽

Ice Shelf ◽

Ice Streams ◽

Amundsen Sea ◽

Grounding Line

Abstract. The response of ice streams in the Amundsen Sea Embayment (ASE) to future climate forcing is highly uncertain. Here we present projections of 21st century response of ASE ice streams to modelled local ocean temperature change using a subset of Coupled Model Intercomparison Project (CMIP5) simulations. We use the BISICLES adaptive mesh refinement (AMR) ice sheet model, with high resolution grounding line resolving capabilities, to explore grounding line migration in response to projected sub-ice shelf basal melting. We find a contribution to sea level rise of between 2.0 cm and 4.5 cm by 2100 under RCP8.5 conditions from the CMIP5 subset, where the mass loss response is linearly related to the mean ocean temperature anomaly. To account for uncertainty associated with model initialisation, we perform three further sets of CMIP5 forced experiments using different parameterisations that explore perturbations to the prescription of initial basal melt, the basal traction coefficient, and the ice stiffening factor. We find that the response of the ASE to ocean temperature forcing is highly dependent on the parameter fields obtained in the initialisation procedure, where the sensitivity of the ASE ice streams to the sub-ice shelf melt forcing is dependent on the choice of parameter set. Accounting for ice sheet model parameter uncertainty results in a projected range in sea level equivalent contribution from the ASE of between −0.02 cm and 12.1 cm by the end of the 21st century.

Download Full-text

Contrasting the modelled sensitivity of the Amundsen Sea Embayment ice streams

Journal of Glaciology ◽

10.1017/jog.2016.40 ◽

2016 ◽

Vol 62 (233) ◽

pp. 552-562 ◽

Cited By ~ 31

Author(s):

ISABEL J. NIAS ◽

STEPHEN L. CORNFORD ◽

ANTONY J. PAYNE

Keyword(s):

Sea Level Rise ◽

Sea Level ◽

Adaptive Mesh Refinement ◽

Adaptive Mesh ◽

Mass Conservation ◽

Physical Parameters ◽

Ice Streams ◽

Amundsen Sea ◽

West Antarctic ◽

Grounding Line

AbstractPresent-day mass loss from the West Antarctic ice sheet is centred on the Amundsen Sea Embayment (ASE), primarily through ice streams, including Pine Island, Thwaites and Smith glaciers. To understand the differences in response of these ice streams, we ran a perturbed parameter ensemble, using a vertically-integrated ice flow model with adaptive mesh refinement. We generated 71 sets of three physical parameters (basal traction coefficient, ice viscosity stiffening factor and sub-shelf melt rate), which we used to simulate the ASE for 50 years. We also explored the effects of different bed geometries and basal sliding laws. The mean rate of sea-level rise across the ensemble of simulations is comparable with current observed rates for the ASE. We found evidence that grounding line dynamics are sensitive to features in the bed geometry: simulations using BedMap2 geometry resulted in a higher rate of sea-level rise than simulations using a rougher geometry, created using mass conservation. Modelled grounding-line retreat of all the three ice streams was sensitive to viscosity and basal traction, while the melt rate was more important in Pine Island and Smith glaciers, which flow through more confined ice shelves than Thwaites, which has a relatively unconfined shelf.

Download Full-text

Comparing ice discharge through West Antarctic Gateways: Weddell vs. Amundsen Sea warming

The Cryosphere Discussions ◽

10.5194/tcd-9-1705-2015 ◽

2015 ◽

Vol 9 (2) ◽

pp. 1705-1733 ◽

Cited By ~ 9

Author(s):

M. A. Martin ◽

A. Levermann ◽

R. Winkelmann

Keyword(s):

Sea Level ◽

Ice Sheet ◽

Back Stress ◽

Step Function ◽

Ocean Warming ◽

Weddell Sea ◽

Amundsen Sea ◽

Function Type ◽

West Antarctic ◽

Grounding Line

Abstract. Future changes in Antarctic ice discharge will be largely controlled by the fate of the floating ice shelves, which exert a back-stress onto Antarctica's marine outlet glaciers. Ice loss in response to warming of the Amundsen Sea has been observed and investigated as a potential trigger for the marine ice-sheet instability. Recent observations and simulations suggest that the Amundsen Sea Sector might already be unstable which would have strong implications for global sea-level rise. At the same time, regional ocean projections show much stronger warm-water intrusion into ice-shelf cavities in the Weddell Sea compared to the observed Amundsen warming. Here we present results of numerical ice sheet modelling with the Parallel Ice Sheet Model (PISM) which show that idealized, step-function type ocean warming in the Weddell Sea leads to more immediate ice discharge with a higher sensitivity to small warming levels than the same warming in the Amundsen Sea. This is consistent with the specific combination of bedrock and ice topography in the Weddell Sea Sector which results in an ice sheet close to floatation. In response to even slight ocean warming, ice loss increases rapidly, peaks and declines within one century. While the cumulative ice loss in the Amundsen Sea Sector is of similar magnitude after five centuries of continued warming, ice loss increases at a slower pace and only for significantly higher warming levels. Although there is more marine ice stored above sea level in close vicinity of the grounding line compared to the Weddell Sea Sector, the ice sheet is farther from floatation and the grounding line initially retreats more slowly.

Download Full-text

Thwaites Glacier grounding-line retreat: influence of width and buttressing parameterizations

Journal of Glaciology ◽

10.3189/2014jog13j117 ◽

2014 ◽

Vol 60 (220) ◽

pp. 305-313 ◽

Cited By ~ 11

Author(s):

David Docquier ◽

David Pollard ◽

Frank Pattyn

Keyword(s):

Sea Level Rise ◽

Sea Level ◽

Ice Sheet ◽

Ice Shelf ◽

Amundsen Sea ◽

West Antarctic Ice Sheet ◽

West Antarctic ◽

Ice Stream ◽

Grounding Line ◽

Flow Convergence

AbstractMajor ice loss has recently been observed along coastal outlet glaciers of the West Antarctic ice sheet, mainly due to increased melting below the ice shelves. However, the behavior of this marine ice sheet is poorly understood, leading to significant shortcomings in ice-sheet models attempting to predict future sea-level rise. The stability of a marine ice sheet is controlled by the dynamics at the grounding line, the boundary between the grounded ice stream and the floating ice shelf. One of the largest contributors to current sea-level rise is the fast-flowing Thwaites Glacier, which flows into the Amundsen Sea. Here we use an ice-stream/ice-shelf model and perform a number of experiments along a central flowline to analyze the sensitivity of its grounding line on centennial timescales. In the absence of width and buttressing effects, we find that the grounding line retreats by ˜300 km in 200 years from the present day (rate of 1.5 km a–1). With variable glacier width implemented in the model, flow convergence slows the retreat of Thwaites grounding line at 0.3–1.2 km a–1. The parameterization of ice-shelf buttressing according to different observed scenarios further reduces the glacier retreat and can even lead to a slight advance in the most buttressed case.

Download Full-text

Twenty-first century response of Petermann Glacier, northwest Greenland to ice shelf loss

Journal of Glaciology ◽

10.1017/jog.2020.97 ◽

2020 ◽

pp. 1-11

Author(s):

Emily A. Hill ◽

G. Hilmar Gudmundsson ◽

J. Rachel Carr ◽

Chris R. Stokes ◽

Helen M. King

Keyword(s):

Sea Level ◽

First Century ◽

Ice Shelf ◽

Ice Shelves ◽

Sea Levels ◽

Grounding Line ◽

Global Mean Sea Level ◽

Near Future ◽

Global Mean ◽

Large Sections

Abstract Ice shelves restrain flow from the Greenland and Antarctic ice sheets. Climate-ocean warming could force thinning or collapse of floating ice shelves and subsequently accelerate flow, increase ice discharge and raise global mean sea levels. Petermann Glacier (PG), northwest Greenland, recently lost large sections of its ice shelf, but its response to total ice shelf loss in the future remains uncertain. Here, we use the ice flow model Úa to assess the sensitivity of PG to changes in ice shelf extent, and to estimate the resultant loss of grounded ice and contribution to sea level rise. Our results have shown that under several scenarios of ice shelf thinning and retreat, removal of the shelf will not contribute substantially to global mean sea level (<1 mm). We hypothesize that grounded ice loss was limited by the stabilization of the grounding line at a topographic high ~12 km inland of its current grounding line position. Further inland, the likelihood of a narrow fjord that slopes seawards suggests that PG is likely to remain insensitive to terminus changes in the near future.

Download Full-text

Results of the Marine Ice Sheet Model Intercomparison Project, MISMIP

The Cryosphere ◽

10.5194/tc-6-573-2012 ◽

2012 ◽

Vol 6 (3) ◽

pp. 573-588 ◽

Cited By ~ 146

Author(s):

F. Pattyn ◽

C. Schoof ◽

L. Perichon ◽

R. C. A. Hindmarsh ◽

E. Bueler ◽

...

Keyword(s):

Steady State ◽

Adaptive Mesh Refinement ◽

Ice Sheet ◽

Adaptive Mesh ◽

Stable Steady State ◽

Fixed Grid ◽

Approximate Analytical Solutions ◽

Grounding Line ◽

Intercomparison Project ◽

Line Positions

Abstract. Predictions of marine ice-sheet behaviour require models that are able to robustly simulate grounding line migration. We present results of an intercomparison exercise for marine ice-sheet models. Verification is effected by comparison with approximate analytical solutions for flux across the grounding line using simplified geometrical configurations (no lateral variations, no effects of lateral buttressing). Unique steady state grounding line positions exist for ice sheets on a downward sloping bed, while hysteresis occurs across an overdeepened bed, and stable steady state grounding line positions only occur on the downward-sloping sections. Models based on the shallow ice approximation, which does not resolve extensional stresses, do not reproduce the approximate analytical results unless appropriate parameterizations for ice flux are imposed at the grounding line. For extensional-stress resolving "shelfy stream" models, differences between model results were mainly due to the choice of spatial discretization. Moving grid methods were found to be the most accurate at capturing grounding line evolution, since they track the grounding line explicitly. Adaptive mesh refinement can further improve accuracy, including fixed grid models that generally perform poorly at coarse resolution. Fixed grid models, with nested grid representations of the grounding line, are able to generate accurate steady state positions, but can be inaccurate over transients. Only one full-Stokes model was included in the intercomparison, and consequently the accuracy of shelfy stream models as approximations of full-Stokes models remains to be determined in detail, especially during transients.

Download Full-text

Modelling the response of the Lambert Glacier–Amery Ice Shelf system, East Antarctica, to uncertain climate forcing over the 21st and 22nd centuries

The Cryosphere ◽

10.5194/tc-8-1057-2014 ◽

2014 ◽

Vol 8 (3) ◽

pp. 1057-1068 ◽

Cited By ~ 20

Author(s):

Y. Gong ◽

S. L. Cornford ◽

A. J. Payne

Keyword(s):

Climate Model ◽

Global Environmental Change ◽

Ice Sheet ◽

Adaptive Mesh ◽

Ocean Model ◽

East Antarctica ◽

Ice Shelf ◽

Amery Ice Shelf ◽

Grounding Line ◽

Lambert Glacier

Abstract. The interaction between the climate system and the large polar ice sheet regions is a key process in global environmental change. We carried out dynamic ice simulations of one of the largest drainage systems in East Antarctica: the Lambert Glacier–Amery Ice Shelf system, with an adaptive mesh ice sheet model. The ice sheet model is driven by surface accumulation and basal melt rates computed by the FESOM (Finite-Element Sea-Ice Ocean Model) ocean model and the RACMO2 (Regional Atmospheric Climate Model) and LMDZ4 (Laboratoire de Météorologie Dynamique Zoom) atmosphere models. The change of ice thickness and velocity in the ice shelf is mainly influenced by the basal melt distribution, but, although the ice shelf thins in most of the simulations, there is little grounding line retreat. We find that the Lambert Glacier grounding line can retreat as much as 40 km if there is sufficient thinning of the ice shelf south of Clemence Massif, but the ocean model does not provide sufficiently high melt rates in that region. Overall, the increased accumulation computed by the atmosphere models outweighs ice stream acceleration so that the net contribution to sea level rise is negative.

Download Full-text

Projecting Antarctic ice discharge using response functions from SeaRISE ice-sheet models

Earth System Dynamics ◽

10.5194/esd-5-271-2014 ◽

2014 ◽

Vol 5 (2) ◽

pp. 271-293 ◽

Cited By ~ 62

Author(s):

A. Levermann ◽

R. Winkelmann ◽

S. Nowicki ◽

J. L. Fastook ◽

K. Frieler ◽

...

Keyword(s):

Sea Level ◽

Ice Sheet ◽

Response Functions ◽

Ice Shelf ◽

Uncertainty Range ◽

Basal Ice ◽

Grounding Line ◽

Rcp 8.5 ◽

Intercomparison Project ◽

Global Sea Level

Abstract. The largest uncertainty in projections of future sea-level change results from the potentially changing dynamical ice discharge from Antarctica. Basal ice-shelf melting induced by a warming ocean has been identified as a major cause for additional ice flow across the grounding line. Here we attempt to estimate the uncertainty range of future ice discharge from Antarctica by combining uncertainty in the climatic forcing, the oceanic response and the ice-sheet model response. The uncertainty in the global mean temperature increase is obtained from historically constrained emulations with the MAGICC-6.0 (Model for the Assessment of Greenhouse gas Induced Climate Change) model. The oceanic forcing is derived from scaling of the subsurface with the atmospheric warming from 19 comprehensive climate models of the Coupled Model Intercomparison Project (CMIP-5) and two ocean models from the EU-project Ice2Sea. The dynamic ice-sheet response is derived from linear response functions for basal ice-shelf melting for four different Antarctic drainage regions using experiments from the Sea-level Response to Ice Sheet Evolution (SeaRISE) intercomparison project with five different Antarctic ice-sheet models. The resulting uncertainty range for the historic Antarctic contribution to global sea-level rise from 1992 to 2011 agrees with the observed contribution for this period if we use the three ice-sheet models with an explicit representation of ice-shelf dynamics and account for the time-delayed warming of the oceanic subsurface compared to the surface air temperature. The median of the additional ice loss for the 21st century is computed to 0.07 m (66% range: 0.02–0.14 m; 90% range: 0.0–0.23 m) of global sea-level equivalent for the low-emission RCP-2.6 (Representative Concentration Pathway) scenario and 0.09 m (66% range: 0.04–0.21 m; 90% range: 0.01–0.37 m) for the strongest RCP-8.5. Assuming no time delay between the atmospheric warming and the oceanic subsurface, these values increase to 0.09 m (66% range: 0.04–0.17 m; 90% range: 0.02–0.25 m) for RCP-2.6 and 0.15 m (66% range: 0.07–0.28 m; 90% range: 0.04–0.43 m) for RCP-8.5. All probability distributions are highly skewed towards high values. The applied ice-sheet models are coarse resolution with limitations in the representation of grounding-line motion. Within the constraints of the applied methods, the uncertainty induced from different ice-sheet models is smaller than that induced by the external forcing to the ice sheets.

Download Full-text