FRISP Report No. 16 (2005) Model investigations of Larsen B Ice Shelf dynamics prior to the breakup H. Sandhäger, W. Rack, and D. Jansen Alfred-Wegener-Institut für Polar- und Meeresforschung, Postfach 120161, D-27515 Bremerhaven Introduction The Larsen B Ice Shelf at the northeast coast of the Antarctic Peninsula (Fig. 1 a) has undergone substantial changes during the last two decades due to regional atmospheric warming (Vaughan and Doake, 1996; Rott et al., 1996; 1998; Skvarca et al., 1998; 1999; Rack et al., 1999; 2000; Rack, 2000; Scambos et al., 2000; 2004 ; Rack and Rott, 2004). The comprehensive documentation of the Larsen B evolution provides important aspects for ice shelf modelling. Observations indicate that all evolution stages from initially stable configuration over gradual retreat to rapid collapse of the northern and central ice shelf parts (Fig. 1 b) were decisively affected by fracture processes. This accentuates the need for expanding conventional ice shelf model approaches to consider the relevance of fracturing. Furthermore, retreat and disintegration of Larsen B were found to be preceded by noticeable changes in the flow and stress regimes of the ice body (Rott et al., 1998; Skvarca et al., 1999; Rack, 2000; Rack and Rott, 2004). Thus, model experiments on these changes in ice dynamics could contribute to gain better insight into the mechanisms of climate-induced ice shelf break-up. One of the key processes appears to be fracturing enhanced by abundant surface melt due to increasing air temperatures (Skvarca et al., 1999; Scambos et al., 2000; 2004; Rack and Rott, 2004). Another aspect to be mentioned is the potential of the Larsen B database to supply reliable model input data and enable detailed validation of model results. Larsen A Seal Nunataks Robertson Island Jason Peninsula Crane Glacier Hektoria Glacier Leppard Glacier Flask Glacier Weddell Sea Cape Disappointment 5 Au g. 19 99 3 0 J a n . 1 9 9 5 Larsen B Ice Shelf 25 km 12000 10000 8000 6000 4000 2000 0 Area (km ) 2 Year 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 1999 1995b 1995a 1986 (b) (a) Larsen B Figure 1: (a) Landsat MSS image showing the 1986 configuration of Larsen B Ice Shelf and parts of its catchment area with distinct glacier systems. Indicated ice front positions of the 1995 post-calving stage and the gradually retreating ice shelf in 1999 are also considered in the numerical experiments. These do not yet include the collapse of the northern and central Larsen B sections in February / March 2002. (b) Time series of Larsen B Ice Shelf area. Dates assigned to the numerical experiments are marked. Our investigations aim at expanding an ice flow model based on conventional approaches and attain- ing an adequate simulation of the ice dynamics of Larsen B in its former stable configuration. The corresponding Configuration Run for the 1986 ice shelf stage yielded considerably more appropriate results than our Reference Run with the ordinary model approach. This gave us reason to perform a series of numerical experiments for the subsequent period of successive ice shelf retreat. Sensitivity Runs were carried out for the Larsen B stages directly prior to and after the major calving event in January 1995, as well as for the August 1999 stage with largely retreated ice front position (Fig. 1).