Media Advisory
5 January 1995

Events in the Northern Larsen Ice Shelf and Their Importance

Dr. Ted Scambos
National Snow and Ice Data Center
University of Colorado, Boulder CO
Christina Hulbe
Department of Geophysical Sciences
University of Chicago, Chicago IL


The northernmost Larsen Ice Shelf, located near the tip of the Antarctic Peninsula, has been in retreat for the last few decades, and much of it is now gone. In late January of 1995, a large area, about 2000 square kilometers (770 square miles) disintegrated into small icebergs during a storm. At the same time, farther south, a large iceberg broke off the ice shelf front. While large iceberg calving events are routine for ice shelves, disintegration is not. It is hypothesized that the unusual breakup may be a consequence of weakening caused by extreme surface melting during several consecutive warm summer seasons in the 1990's, and by a regional warming over the last few decades. It is unclear whether the observed warming in the Antarctic Peninsula climate, about 2.5 degrees Celsius since the 1940's, is part of a global warming trend or is simply a normal fluctuation in regional climate. The events in the Antarctic Peninsula are unlikely to have an effect on sea level or climate themselves; however, monitoring the area provides insight into ongoing climate change.

before/after image of 1995 larsen b ice shelf breakup

Before and after image of the 1995 Larsen B breakup. The left side shows the Larsen B Ice Shelf on December 26, 1993. The right side shows the ice shelf after the 1995 breakup event on February 13, 2005. High resolution


Two dramatic events occurred in the Larsen Ice Shelf in late January of 1995. A large iceberg 70 kilometers by 25 kilometers (43 miles by 16 miles) calved from the shelf between Jason Peninsula and Robertson Island (see map), and the northernmost part of the shelf, north of Seal Nunataks, disintegrated. The large iceberg calving event received the most notice in popular media coverage but such events are part of the normal mass balance cycle for ice shelves. The northern disintegration, on the other hand, occurred in an unprecedented manner, by the sudden break-up of a region approximately 2000 square kilometers (770 square miles) into many small icebergs (typically 2 kilometers or smaller). It is the more likely of the two events to be related to climate change.

Ice shelves are thick (usually hundreds of meters) floating platforms of ice. They are connected to land at coastal grounding lines and where they flow around islands, and they are exposed to the ocean at seaward calving fronts. Ice shelves gain mass by flow from grounded ice sheets and glaciers and new snow accumulation on their surfaces. They lose mass primarily by iceberg calving and secondarily by melting. Ice shelves balance between gravity-driven horizontal spreading and stresses at grounding lines and the calving front. Changing the volume of an ice shelf (for example, by losing ice to calving) does not change sea level because the floating ice already displaces a volume of water equal to the volume of water it contains. However, if ice currently resting on the continental surface were to flow into the ocean more rapidly as a result of the removal of the fringing ice shelves, then sea level would rise.

The small, thin ice shelves that fringe the Antarctic Peninsula are sustained primarily by new snow accumulation (unlike the larger Ross and Ronne-Filchner ice shelves, which are fed mainly by flow from inland ice sheets). Thickness of these ice shelves is approximately 200 meters (656 feet). Changes in winter accumulation or summer melting are of fundamental importance to the health of fringing ice shelves. While little accumulation or temperature data exist for the Larsen Ice Shelf, anecdotal evidence indicates substantial recent summer melting. Field expeditions conducted in the 1990's have routinely found extensive surface melting (1), summer melt ponds were observed in satellite images for several years prior to the breakup (see figure), and retreat of the far northern shelf front is well-documented in satellite images (2). It is probable that thinning and weakening due to extreme summer surface melting led to the disintegration of the far northern part of the shelf.

Other Antarctic Peninsula ice shelves are also retreating, or have disappeared, over the last decade. For example, the Miller Ice Shelf has been receding since about 1974 (3) and the Wordie Ice Shelf broke apart in the 1980's (4). Both ice shelves are on the western side of the Peninsula, where local warming has been documented at British Antarctic Survey bases (4). Temperatures have increased an average of 2.5 C (4.5 degrees Fahrenheit) since the 1940's. In all likelihood, these events are linked. In the late 1970's, Mercer (5) noted the correspondence between the distribution of ice shelves and mean annual air temperature isotherms around Antarctica. He suggested that the -4 degrees Celsius (25 degrees Fahrenheit) isotherm provides an upper limit for ice shelf survival. Both the Northern Larsen and Wordie ice shelves were near this limit.

Ice shelves respond to climate change more rapidly than grounded ice sheets or glaciers. The ice shelves in question were particularly susceptible, being small fringing shelves at the limit of ice shelf stability for the recent global climate (the Northern Larsen Ice Shelf is the northernmost ice shelf on the east side of the Antarctic Peninsula; the Wordie Ice Shelf was the northernmost on the west side). Peninsula shelves have experienced cycles of growth and decay throughout the Holocene (the last 10,000 years) (3). The question remains whether or not the observed warming and the ice shelf breakups are manifestations of global climate warming. Beyond that question still further, lies the question of whether current changes in regional climates are anthropogenic or part of a cycle that would have occurred at this time in a non-industrial world.

What does the future hold? Numerical models designed to investigate the influence of the calving events on flow of the remaining Northern Larsen Ice Shelf suggest that the current configuration is stable (6). The key to stability is support provided by the Seal Nunataks and Robertson Island, along its new northern front. If the shelf becomes detached from those islands, rapid retreat of the front is likely. Unless there is a change in the observed warming trend, further retreats of fringing ice shelves along the Antarctic Peninsula are the most likely scenario in the near future. The current consensus of the ice mechanics community is that the loss of an ice shelf will not cause a drastic speed-up in the glaciers feeding it. However, a study of the velocity of ice motion over the disintegrated area(7) showed a slight (10 to 15%) increase in ice speed, interpreted as possibly the result of reduced back pressure on the source glaciers from the receding shelf. This has important implications for sea level change due to the loss of these ice shelves, should the observed speed increase continue.

It is important to monitor changes in the extent of ice shelves, along the Antarctic Peninsula and throughout Antarctica, to better understand the nature and extent of the climate signal they provide. Some such monitoring programs are already underway within the British Antarctic Survey and among US researchers under NASA and NSF OPP grants.


Pedro Skvarca, Instituto Antartico Argentino, personal communication.

Skvarca, P. 1993. Fast recession of the Northern Larsen Ice Shelf monitored by space images. Annals of Glaciology 17:317-321.

Domack, E. W, Ishman, S. E., Stein, A. B., McClennen, C. E., and Jull, A. J. T. 1995. Late Holocene Miller ice shelf, Antarctic Peninsula, sedimentological, geochemical, and palaeontological evidence. Antarctic Science 7 (2):159-170.

Doake, C. S. M., and Vaughn, D. G. 1991. Rapid disintegration of the Wordie ice shelf in response to atmospheric warming. Nature 350:328-330.

Mercer, J. H. 1978. West Antarctic Ice Sheet and CO2 greenhouse effect: A threat of disaster. Nature 271:321-325.

A finite-element numerical model constructed by one of the authors (CLH). The Northern Larsen Ice Shelf is modeled as a slab of uniform thickness flowing in an embayment with no-slip boundaries and an ice front that experiences the stress of sea water pressure only. The model uses a well-established flow law and stress-equilibrium equations. A series of experiments are conducted with pre-calving/pre-disintegration and present (both with and without Seal Nunataks and Robertson Island) ice shelf geometries.

Bindschadler, R. A., Fahnestock, M. A., Skvarka, P., and Scambos, T. A. 1994. Surface-velocity field of the northern Larsen Ice Shelf, Antarctica. Annals of Glaciology 20:319-326.