Recent Modelling of Antarctica

Recent studies using a variety of dating techniques have indicated that the former extent of the East Antarctic Ice Sheet at the LGM may have been less than previously thought (e.g., Gore et al., 2001; Hodgson et al., 2001; Mackintosh et al., 2007). Past modelling studies however, including both those by Ritz and Huybrechts described above, have placed the margin of the East Antarctic Ice Sheet at the edge of the continental shelf in agreement with the CLIMAP reconstruction (CLIMAP, 1976). With this in mind some new numerical modelling experiments have been undertaken in order to investigate LGM ice sheet configurations compatible with the newly available geological evidence from East Antarctica.

The GLIMMER community ice-sheet model is an implementation of the shallow-ice approximation with full thermodynamic coupling and the inclusion of basal sliding (Payne, 1999). Using this model an ice sheet was created on a slightly modified and isostatically rebounded BEDMAP topography (Lythe et al., 2001) with 20 km grid resolution. Model parameter values were tuned to reproduce as closely as possible the present day extent of the Lambert Graben region of the East Antarctic Ice Sheet (Fig. 12.9A). The accumulation rate was calculated for each grid cell as a function of continentality (proximity to the coastline) following the method of Oerlemans (1982); surface air temperature was calculated using a fixed lapse rate and mass loss at the marine margins as a prescribed calving percentage.

Table 12.1 shows the parameter values supplied to the model for the simulations of both the present day and LGM ice sheets. For the LGM these values are taken from the general consensus of the paleoclimate literature (e.g., Petit et al., 1999; Siegert, 2003; Van Ommen et al., 2004). The extent and thickness of the present day ice sheet produced by the model is a good match with observations in the Lambert Glacier region of East Antarctica (c.f. Fig. 12.9A with Fig. 12.3). The shape of the ice-sheet around the Lambert Graben and the surface height of the interior ice-sheet are reasonably well reproduced; the model used here is not capable of reproducing ice shelves and hence the Amery Ice Shelf does not appear in the results.

A characteristic of this reconstruction of the LGM ice sheet (Fig. 12.9B) is the uneven nature of the glacial advance around the margins. The location of the grounding line at the LGM is predicted by the model to have advanced ~80km towards the middle of the continental shelf on the western flank of the Lambert Graben, by only ~40km on the eastern flank and for no significant advance on the present day margin along much of the rest of the coastline (Fig. 12.9B, C). This pattern is matched by negligible changes in

Figure 12.9: Numerical model results for the Princess Elizabeth - Lambert Graben - Mac.Robertson Land area of the East Antarctic Ice Sheet. (A) Reference experiment created with parameter values tuned for the present day. (B) Ice sheet reconstruction with best-estimate parameter values for the LGM. (C) Predicted ice surface height change (B-A) at the LGM. The locations of the two longitudinal profiles in Fig. 12.10 are marked A-B

Figure 12.9: Numerical model results for the Princess Elizabeth - Lambert Graben - Mac.Robertson Land area of the East Antarctic Ice Sheet. (A) Reference experiment created with parameter values tuned for the present day. (B) Ice sheet reconstruction with best-estimate parameter values for the LGM. (C) Predicted ice surface height change (B-A) at the LGM. The locations of the two longitudinal profiles in Fig. 12.10 are marked A-B

Table 12.1: Values of model parameters used for the simulation of both the present day and LGM East Antarctic Ice Sheet.

Present day (0 ka BP)

LGM (~21 ka BP)

Sea level air temperature (°C)

-15

-25

Air temperature lapse rate (°Cm_1)

-0.008

-0.008

Relative precipitation

100%

50%

Eustatic sea level (m)

0.0

-120.0

Calving fraction at marine margins

90%

90%

Basal traction factor

1.0 x 10-4

1.0 x 10-4

Geothermal heat flux (W m"2)

56 x 10-3

56x10-3

ice-sheet thickness along parts of the coast which are coincident with significant thickening in other areas (Fig. 12.9C). The patterns of advance and thickening seen here are strongly influenced by the bathymetry of the continental shelf. Where the shelf is shallow the 120 m sea-level drop at the LGM allows expansion of the ice sheet. Under this scenario the extent of the ice-sheet is controlled largely by eustatic sea-level and hence by the volume of northern hemisphere ice sheets (e.g., Denton et al., 1986). In contrast the interior of the ice-sheet experiences a general decrease in thickness of between 100 and 150 m under the LGM conditions prescribed here (Fig. 12.9C). This is a result of the decrease in accumulation which is not mitigated by the decrease in air temperature as surface melting is relatively unimportant here under either present day or interglacial conditions.

The locations of the two profiles shown in Fig. 12.10 were chosen to allow direct comparisons to be made with the LGM reconstructions of the icesheet surface from cosmogenic dating of glacially abraded surfaces and transported clasts (Fig. 12.3 and Mackintosh et al., 2007). The AB-profile shown in Fig. 12.10A follows the route of the Fisher Glacier to its confluence with the Lambert Glacier and where the continuation of the latter discharges into the Amery Ice Shelf. Though the model is not capable of reproducing ice shelves, it can be seen (by comparing Figs. 12.3B and 12.10A) that grounded ice in the model extends ~ 100 km past the position of the grounding line inferred from GPS measurements of tidal motion (Fricker et al., 2002). This over-advance under present day conditions occurs possibly because the depth of the topographic low, near the deepest point of which is located the present grounding zone, is underestimated in BEDMAP. The present day model terminus position of the glacier in the Lambert Graben is in fact very similar to the location of the LGM transition zone between steady icesheet flow and ice streaming inferred from geological evidence (Fig. 12.3B).

0 100 200 300 400 500 0 100 200 300 400

Figure 12.10: Longitudinal profiles through the model reconstructed ice sheet for Fisher-Lambert Glacier and Framnes Mountains. These profiles demonstrate the main features of the LGM reconstruction of East Antarctica i.e. surface lowering in the interior with little or no advance of the margins.

0 100 200 300 400 500 0 100 200 300 400

Figure 12.10: Longitudinal profiles through the model reconstructed ice sheet for Fisher-Lambert Glacier and Framnes Mountains. These profiles demonstrate the main features of the LGM reconstruction of East Antarctica i.e. surface lowering in the interior with little or no advance of the margins.

It is therefore perhaps unsurprising that no further terminus advance occurs in this area when the model is run under LGM climate/sea-level conditions (Fig. 12.10A).

The LGM advance of the East Antarctic Ice Sheet margin along the CD profile (Fig. 12.9) has been shown by Mackintosh et al. (2007) to have been only around 5-10km.This is beneath the grid resolution of the present model and hence the unchanged margin position along the CD-profile under LGM conditions (Fig. 12.10B) is in agreement with the field evidence. The model also predicts a slight LGM lowering of the ice-sheet surface along the entire length of this profile as a result of decreased accumulation, counter to the geological evidence which suggests a moderate increase in thickness near the margin (Mackintosh et al., 2007). The spatial extent of this thickening, however, is uncertain and may be too small to be resolved by the current model.

0 0

Post a comment