Earth Science.

Nuclear-monitoring system passes test

A global surveillance system set up to detect clandestine nuclear blasts performed well in an impromptu test, according to a team of British seismologists.

As part of the Comprehensive Test Ban Treaty, adopted in 1996, cooperating nations have linked up more than 140 seismic stations to form the International Monitoring System (IMS). On Aug. 22, 1998, 10 of these stations picked up vibrations emanating from a site in Kazakhstan where the former Soviet Union conducted nuclear tests.

The seismic waves recorded at these stations looked highly suspicious, report scientists from the Atomic Weapons Establishment (AWE) in Brimpton, England. The Kazakh event had the strength of a magnitude 3.8 earthquake, but the vibrations did not resemble earthquake waves, says AWE's Alan Douglas.

The stations recorded seismic waves that had passed through the planet but no waves that had rippled along the surface. This pattern is more typical of explosions than quakes, the scientists report in the April 8 NATURE. What's more, the seismic recordings indicate that the waves originated from a point extremely close to the surface, again consistent with an explosion.

That's what it was, in fact. On Aug. 22, engineers from Kazakhstan and the United States set off a small nonnuclear blast to close off tunnels at the test site. The yield of the explosion equaled a 200-ton nuclear shot, says Douglas.

The goal for IMS is to have at least three stations picking up vibrations from events down to magnitude 4.0 (SN: 5/11/96, p. 298). The system should be able to pinpoint these to within a 1,000 [km.sup.2] area, about the size of a circle with a radius of 18 km. In the Kazakh case, IMS picked up a smaller event and located it to within 12 km of the actual site.

Douglas and his colleagues note that this test case is artificial because they knew about the blast before it happened. It shows, nonetheless, that IMS can alert scientists to irregular vibrations, even from quite small explosions. --R.M.

Disappearing ice down south

The glacial shelves surrounding the Antarctic Peninsula lost large chunks of ice last year--an area roughly equal to the size of Rhode Island--accelerating a decades-long trend caused by increasing temperatures in the region.

The peninsula reaches out like an arm toward the tip of South America. Along its coast, glaciers flow into the ocean and form thick floating sheets, called ice shelves. Since the 1950s, several of the smaller ice shelves have disintegrated, and now larger shelves are starting to retreat.

Satellite images show that the Larsen B Ice Shelf shrank by 1,700 square kilometers during the past Antarctic summer, which runs from December through March. Recent analysis of satellite images taken on the other side of the peninsula has revealed that the Wilkins Ice Shelf lost 1,100 [km.sup.2] during the early part of 1998, according to Ted Scambos of the University of Colorado's National Snow and Ice Data Center in Boulder.

Although the loss of ice represents a significant change, some researchers are surprised that so much shelf area remains. A team of scientists last year predicted that the entire Larsen B shelf would rapidly disintegrate once its front edge retreated behind a theoretical line of stability--an event that happened in February 1998 (SN: 5/9/98, p. 303). "We're slightly surprised it hasn't collapsed entirely already," says David Vaughan of the British Antarctic Survey in Cambridge.

With winter now returning to the Antarctic, researchers expect no more ice loss until the next southern summer, when they will watch closely to see if Larsen B splinters. The collapse of ice shelves does not alter global sea level because these sheets are already floating in the ocean. --R.M.