ABSTRACT The mass balance of marine ice-sheets, such as the West Antarctic Ice Sheet, is mostly c... more ABSTRACT The mass balance of marine ice-sheets, such as the West Antarctic Ice Sheet, is mostly controlled by their grounding line dynamics. Most numerical models simulating marine ice-sheets involve simplifications and do not include all the stress gradients. First results obtained with a 3D full-Stokes model for the grounded ice-sheet / floating ice-shelf transition, using the finite-element code Elmer/Ice, are presented. The initial geometry, which takes into account a dome and a calving front, has been laterally extruded from a previously investigated 2D flowline geometry. The grounding line migration is computed by solving the contact problem between the ice and the rigid downward sloping bedrock, where a non linear friction law is applied in the two horizontal directions. The evolutions of the sea-air and sea-ice interfaces are determined by the solution of a local transport equation. The consistency between the D model and the analogous results of the flowline model is shown by comparing the results in the basic extruded case, with no normal flux through lateral boundaries. Thereafter, spatially non uniform perturbations are introduced, to simulate the grounding line dynamics under fully three-dimensional perturbations.
ABSTRACT Ice discharge and grounding line retreat in West Antarctica have been accelerated during... more ABSTRACT Ice discharge and grounding line retreat in West Antarctica have been accelerated during the last decades. One of the most striking example is Pine Island Glacier (PIG) which accelerated dramatically over the last 30 years. Such rapid changes in this part of Antarctica are due to large modifications of ice dynamics which are nevertheless poorly understood, and badly represented in numerical models, as pointed out by the IPCC fourth assessment report. Here, a 3D full-Stokes model of a marine ice sheet is used to carry out prognostic simulations of PIG over the next two centuries. The flow problem is coupled with the evolution of the upper and lower free surfaces, and the position of the grounding line is determined by solving the contact problem between the ice-shelf/ice-sheet lower surface and the bedrock. The upper and lower surfaces, and the bathymetry provided on a 1 km grid (courtesy of A. Le Brocq) are used to produce the initial geometry of the entire PIG basin. The mesh refinement is a function of the surface velocities (also provided on a 1 km grid by A. Le Brocq) Hessian matrix and the distance to the grounding line. Surface velocities are also used to infer the basal drag through the resolution of an inverse Robin problem. The initial surface is first relaxed and the results are compared to the observed current surface elevation, surface velocity and change in surface elevation. A perturbation experiment is then performed for which the whole ice-shelf is instantaneously removed. This test can be seen as a worst case scenario as all the buttressing induced by the ice shelf is lost instantaneously. The effect of the ice-shelf disintegration for the following two centuries is discussed in terms of grounding line retreat and increase in sea level.
We have used computational fluid dynamics modeling (CFD) to synchronize the flow conditions in th... more We have used computational fluid dynamics modeling (CFD) to synchronize the flow conditions in the flow channels of two complementary surface-sensitive characterization techniques: surface plasmon resonance (SPR) and quartz crystal microbalance (QCM). Since the footprint of the flow channels of the two devices is specified by their function, the flow behavior can only be varied either by altering the height of the flow channel, or altering the volumetric rate of flow (flow rate) through the channel. The relevant quantity that must be calibrated is the shear strain on the measurement surface (center and bottom) of the flow channel. Our CFD modeling shows that the flow behavior is in the Stokes flow regime. We were thus able to generate a scaling expression with parameters for flow rate and flow channel height for each of the two devices: f(QCM)=2.64f(SPR)(h(QCM)/h(SPR)(2), where f(QCM) and f(SPR) are the flow rates in the SPR and QCM flow channels, respectively, and h(QCM)/h(SPR) is the ratio of the heights of the two channels. We demonstrate the success of our calibration procedure through the combined use of commercially available SPR and QCM flow channel devices on both a biomolecular interaction system of surface immobilized biotin and streptavidin and a targeted drug delivery model system of biotinylated liposomes interacting with a streptavidin functionalized surface.
ABSTRACT The Greenland Ice Sheet (GrIS) has been increasingly loosing mass during the last two de... more ABSTRACT The Greenland Ice Sheet (GrIS) has been increasingly loosing mass during the last two decades, enhancing its contribution to sea level rise. The total mass balance of an ice sheet is equal to the surface mass balance minus the discharge (ice flux to the ocean). Both terms are contributing increasingly to the observed GrIS imbalance . Here, we use use the new generation continental scale ice-sheet model Elmer/Ice with (1) an unstructured mesh allowing to model outlet glaciers with a sufficient resolution; (2) a complete resolution of the full system of equations governing the ice flow and (3) inverse methods to better constrain poorly known parameters from observations. As a first result, we show that our model can reproduce the currently observed ice dynamics and the currently observed ice discharge. We then study the sensitivity of the GrIS to surface mass balance. The model is forced by three different Regional Climate Models (RCMs) both for the ERA-Interim period and for prognostic runs 200 years into the future under emission scenarios A1B and E1, following the experimental setups designed for the ice2sea simulations. These imposed surface mass balance forcings significantly differ in terms of total values and trends, and we show that it directly affects the predicted total ice volume. In the absence of perturbation of the basal and seaward boundary conditions, the difference in the predicted ice discharge remains small. We finally investigate the dynamical response of the ice sheet to changes in basal lubrication by linking the basal sliding to surface run-off anomalies, according to the ice2sea experiments.
ABSTRACT Proper knowledge of bedrock elevation is a crucial input parameter for ice sheet modelin... more ABSTRACT Proper knowledge of bedrock elevation is a crucial input parameter for ice sheet modeling. For obvious reasons of inaccessibility, the knowledge of bedrock elevation is fragmented, and the question of where future radar measurement campaigns should focus on is critical, especially in the context of global warming and related current changes in ice sheet dynamics. By using a full-Stokes finite element code with a proper mechanical solution for the contact problem at the grounding line, this study examines the sensitivity of an ice flow model on the accuracy of the bedrock description. We show that mass exchange between ice-sheet and ocean is little affected by our crude knowledge of the bed elevation in the interior of the continent. Conversely, in the vicinity of coastal regions and in order to avoid unrealistic unstable behavior of outlet glaciers, the bedrock needs to be known with a sampling scale of minimum 2 km.
ABSTRACT To investigate past climate change in the Northwest Pacific region, an ice core was retr... more ABSTRACT To investigate past climate change in the Northwest Pacific region, an ice core was retrieved in June 1998 from the Gorshkov crater glacier at the top of the Ushkovsky volcano, in central Kamchatka. Hydrogen isotope (δD) analysis and past accumulation reconstructions were conducted to a depth of 140.7 m, dated to 1735. Two accumulation reconstruction methods were applied with the Salamatin and the Elmer/Ice ice flow models. Reconstructed accumulation rates and δD were significantly correlated with North Pacific surface temperature. This, and a significant correlation of δD with the North Pacific Gyre Oscillation (NPGO) index implies that NPGO data is contained in this record. Wavelet analysis shows that the ice core records have significant multi-decadal power spectra up to the late 19th century. The multi-decadal periods of reconstructed accumulation rates change at around 1850 in the same way as do Northeast Pacific ice core and tree ring records. The loss of multi-decadal scale power spectra of δD and the 6‰ increase in its average value occurred around 1880. Thus the core record confirms that the periodicity of precipitation for the entire North Pacific changed between the end of the Little Ice Age through the present due to changes in conditions in the North Pacific Ocean.
ABSTRACT The mass balance of marine ice-sheets, such as the West Antarctic Ice Sheet, is mostly c... more ABSTRACT The mass balance of marine ice-sheets, such as the West Antarctic Ice Sheet, is mostly controlled by their grounding line dynamics. Most numerical models simulating marine ice-sheets involve simplifications and do not include all the stress gradients. First results obtained with a 3D full-Stokes model for the grounded ice-sheet / floating ice-shelf transition, using the finite-element code Elmer/Ice, are presented. The initial geometry, which takes into account a dome and a calving front, has been laterally extruded from a previously investigated 2D flowline geometry. The grounding line migration is computed by solving the contact problem between the ice and the rigid downward sloping bedrock, where a non linear friction law is applied in the two horizontal directions. The evolutions of the sea-air and sea-ice interfaces are determined by the solution of a local transport equation. The consistency between the D model and the analogous results of the flowline model is shown by comparing the results in the basic extruded case, with no normal flux through lateral boundaries. Thereafter, spatially non uniform perturbations are introduced, to simulate the grounding line dynamics under fully three-dimensional perturbations.
ABSTRACT Ice discharge and grounding line retreat in West Antarctica have been accelerated during... more ABSTRACT Ice discharge and grounding line retreat in West Antarctica have been accelerated during the last decades. One of the most striking example is Pine Island Glacier (PIG) which accelerated dramatically over the last 30 years. Such rapid changes in this part of Antarctica are due to large modifications of ice dynamics which are nevertheless poorly understood, and badly represented in numerical models, as pointed out by the IPCC fourth assessment report. Here, a 3D full-Stokes model of a marine ice sheet is used to carry out prognostic simulations of PIG over the next two centuries. The flow problem is coupled with the evolution of the upper and lower free surfaces, and the position of the grounding line is determined by solving the contact problem between the ice-shelf/ice-sheet lower surface and the bedrock. The upper and lower surfaces, and the bathymetry provided on a 1 km grid (courtesy of A. Le Brocq) are used to produce the initial geometry of the entire PIG basin. The mesh refinement is a function of the surface velocities (also provided on a 1 km grid by A. Le Brocq) Hessian matrix and the distance to the grounding line. Surface velocities are also used to infer the basal drag through the resolution of an inverse Robin problem. The initial surface is first relaxed and the results are compared to the observed current surface elevation, surface velocity and change in surface elevation. A perturbation experiment is then performed for which the whole ice-shelf is instantaneously removed. This test can be seen as a worst case scenario as all the buttressing induced by the ice shelf is lost instantaneously. The effect of the ice-shelf disintegration for the following two centuries is discussed in terms of grounding line retreat and increase in sea level.
We have used computational fluid dynamics modeling (CFD) to synchronize the flow conditions in th... more We have used computational fluid dynamics modeling (CFD) to synchronize the flow conditions in the flow channels of two complementary surface-sensitive characterization techniques: surface plasmon resonance (SPR) and quartz crystal microbalance (QCM). Since the footprint of the flow channels of the two devices is specified by their function, the flow behavior can only be varied either by altering the height of the flow channel, or altering the volumetric rate of flow (flow rate) through the channel. The relevant quantity that must be calibrated is the shear strain on the measurement surface (center and bottom) of the flow channel. Our CFD modeling shows that the flow behavior is in the Stokes flow regime. We were thus able to generate a scaling expression with parameters for flow rate and flow channel height for each of the two devices: f(QCM)=2.64f(SPR)(h(QCM)/h(SPR)(2), where f(QCM) and f(SPR) are the flow rates in the SPR and QCM flow channels, respectively, and h(QCM)/h(SPR) is the ratio of the heights of the two channels. We demonstrate the success of our calibration procedure through the combined use of commercially available SPR and QCM flow channel devices on both a biomolecular interaction system of surface immobilized biotin and streptavidin and a targeted drug delivery model system of biotinylated liposomes interacting with a streptavidin functionalized surface.
ABSTRACT The Greenland Ice Sheet (GrIS) has been increasingly loosing mass during the last two de... more ABSTRACT The Greenland Ice Sheet (GrIS) has been increasingly loosing mass during the last two decades, enhancing its contribution to sea level rise. The total mass balance of an ice sheet is equal to the surface mass balance minus the discharge (ice flux to the ocean). Both terms are contributing increasingly to the observed GrIS imbalance . Here, we use use the new generation continental scale ice-sheet model Elmer/Ice with (1) an unstructured mesh allowing to model outlet glaciers with a sufficient resolution; (2) a complete resolution of the full system of equations governing the ice flow and (3) inverse methods to better constrain poorly known parameters from observations. As a first result, we show that our model can reproduce the currently observed ice dynamics and the currently observed ice discharge. We then study the sensitivity of the GrIS to surface mass balance. The model is forced by three different Regional Climate Models (RCMs) both for the ERA-Interim period and for prognostic runs 200 years into the future under emission scenarios A1B and E1, following the experimental setups designed for the ice2sea simulations. These imposed surface mass balance forcings significantly differ in terms of total values and trends, and we show that it directly affects the predicted total ice volume. In the absence of perturbation of the basal and seaward boundary conditions, the difference in the predicted ice discharge remains small. We finally investigate the dynamical response of the ice sheet to changes in basal lubrication by linking the basal sliding to surface run-off anomalies, according to the ice2sea experiments.
ABSTRACT Proper knowledge of bedrock elevation is a crucial input parameter for ice sheet modelin... more ABSTRACT Proper knowledge of bedrock elevation is a crucial input parameter for ice sheet modeling. For obvious reasons of inaccessibility, the knowledge of bedrock elevation is fragmented, and the question of where future radar measurement campaigns should focus on is critical, especially in the context of global warming and related current changes in ice sheet dynamics. By using a full-Stokes finite element code with a proper mechanical solution for the contact problem at the grounding line, this study examines the sensitivity of an ice flow model on the accuracy of the bedrock description. We show that mass exchange between ice-sheet and ocean is little affected by our crude knowledge of the bed elevation in the interior of the continent. Conversely, in the vicinity of coastal regions and in order to avoid unrealistic unstable behavior of outlet glaciers, the bedrock needs to be known with a sampling scale of minimum 2 km.
ABSTRACT To investigate past climate change in the Northwest Pacific region, an ice core was retr... more ABSTRACT To investigate past climate change in the Northwest Pacific region, an ice core was retrieved in June 1998 from the Gorshkov crater glacier at the top of the Ushkovsky volcano, in central Kamchatka. Hydrogen isotope (δD) analysis and past accumulation reconstructions were conducted to a depth of 140.7 m, dated to 1735. Two accumulation reconstruction methods were applied with the Salamatin and the Elmer/Ice ice flow models. Reconstructed accumulation rates and δD were significantly correlated with North Pacific surface temperature. This, and a significant correlation of δD with the North Pacific Gyre Oscillation (NPGO) index implies that NPGO data is contained in this record. Wavelet analysis shows that the ice core records have significant multi-decadal power spectra up to the late 19th century. The multi-decadal periods of reconstructed accumulation rates change at around 1850 in the same way as do Northeast Pacific ice core and tree ring records. The loss of multi-decadal scale power spectra of δD and the 6‰ increase in its average value occurred around 1880. Thus the core record confirms that the periodicity of precipitation for the entire North Pacific changed between the end of the Little Ice Age through the present due to changes in conditions in the North Pacific Ocean.
Uploads
Papers by Thomas Zwinger