The results of a major geophysical experiment called SIGHT (South Island GeopHysical Transect) carried out across the South Island from 1995 to 1998 by geophysicists from Victoria University, the University of Southern California, the Massachusetts Institute of Technology and IGNS, are giving new insights into the tectonic drama being played out beneath the Southern Alps. Geologically speaking, New Zealand is a battleground between two tectonic giants: the Pacific Plate to the east and the Australian Plate to the west. The nature of the contest varies along the battlefront, depending on the type of engagement (subduction or collision), the crustal entities that are engaged (oceanic or continental) and the rates of engagement. Right now there is considerable research interest in the nature of the clash between the plates along the Alpine Fault—the collision zone that is responsible for the existence of the Southern Alps. Fundamental to understanding the geological processes in this area is the recognition that there are two types of crust, each characterised by distinct igneous rocks originating in the earth's mantle. Oceanic crust is predominantly basalt, a volcanic rock rich in iron- and magnesium-bearing minerals, and it is only about 7 km thick. Compared with the earth's radius—about 6500 km-7 km is skin-thickness or less; vanishingly thin with respect to the great bulk of the mantle. More than 60 per cent of the earth's crust is oceanic, but determining its nature took a long time because the ocean floor is blanketed in sediment and is effectively hidden from view. Continental crust is dominated by granite, a plutonic rock rich in silicon-and aluminium-bearing minerals. It is less dense than oceanic crust and is therefore more buoyant. Accordingly, continental crust rides higher on the earth's surface compared with oceanic crust. As a rule, the higher it rides the thicker the crust. Below the Tibetan Plateau in Asia and the Altiplano in South America—the highest-standing large areas on earth—the crust is 70-80 km thick. Typical continental crust is 40-50 km thick. Oceanic crust may be thought of as milk and continental crust as cream. Like cream, granite is enriched in lighter materials, and, once at the surface, it is hard to destroy because it is buoyant. Accordingly, the amount of continental crust has slowly increased through time as the mantle has evolved. Indeed, the area of continental crust has probably never been as large as it is now, and it is still slowly growing. The distribution pattern of continental crust keeps changing with plate tectonics, but as a generalisation the area cannot diminish; it can only increase. The mantle is still producing cream. The plate boundary that runs through New Zealand separates areas of both oceanic and continental crust on both the Pacific and Australian Plates. North of the North Island, the oceanic crust of the Pacific Plate is being subducted beneath the oceanic crust of the Australian Plate. In terms of the contest between the plates, in this oceanic– oceanic scrummage the Australian Plate is winning. This struggle is happening out of sight, beneath the sea. The North Island itself consists entirely of continental crust belonging to the Australian Plate. Along the eastern margin of the North Island, oceanic crust of the Pacific Plate is being subducted beneath the continental crust of the Australian Plate—a continental–oceanic collision. Throughout the South Island the situation is quite different. Here we have a continental–continental collision, where continental crust of the Australian Plate is colliding with continental crust of the Pacific Plate. The Australian crust is the stronger, and, though crumpled, the eastern edge of the Australian Plate is behaving like a stop wall to the weaker mountain-forming western margin of the Pacific Plate. South of the South Island, the nature of the plate boundary changes again, and the oceanic crust of the Australian Plate is being subducted below the continental crust of the Pacific Plate. Sight was set up to provide insight into the South Island section of the tectonic battlefront. It was a huge undertaking, involving setting off about 100 explosions along two complete traverses of the South Island. Drill holes to take the explosives were up to 65 m deep, and each contained 600 to 1200 kg of Powergel. More than 400 seismographs were deployed to record echoes from the rocks. The two onshore profiles were extended offshore from both coastlines across the whole continental shelf using a number of ships. At sea, about 40,000 airgun shots substituted for actual explosions. A multitude of workers was involved in the project over a three-year period, and huge volumes of seismic data were recorded. From this research we have now learned that the Alpine Fault dips to the east beneath the Southern Alps at about a 40-degree angle. To put it another way, continental rocks of the Pacific Plate are being pushed up a relatively steep 40-degree ramp that emerges at the surface as a remarkably straight line, the Alpine Fault. But if the Alps are coming up, what is happening at depth to the material that underlies them? The new SIGHT data tell us that the crust is thickened beneath the Mount Cook region to about 45 km. This thickening is referred to as a crustal root, and it has formed as lower crustal rock is pushed down during plate collision while upper crustal rocks are pushed upwards. The root provides the Alps with buoyancy, giving extra impetus to their uplift. To the west and east of the Alpine Fault zone, crustal thickness drops away dramatically to about 25 km, which is the standard crustal thickness below the New Zealand continent (called Zealandia). Sight has not only told us where the base of the crust is across the South Island and how deep it is, it also has revealed a great deal about the nature of the material within the crust. Studies of the electrical properties and seismic wave speeds beneath the Alps have provided compelling evidence that there may be a very substantial zone of high-pressure fluid within the crustal root at depths of between 8 and 35 km below the surface. This zone has unusually high conductivity and unusually low seismic velocity. The full extent of this zone along the length of the Alps remains unknown, but it certainly exists beneath our highest mountains in the area from Mt Cook to Whataroa River. The zone is inferred to be a region of high-pressure fluid that is behaving like a massive cushion supporting the Alps. Think of it as a geological water bed! All this is of more than passing interest because it carries implications for earthquake generation on the Alpine Fault. One of the effects of fluid within the crust is to reduce shear stress. This effect may explain the puzzling yet well documented lack of significant earthquake activity associated with the Alpine Fault. The absence of earthquakes may be due to a lack of tectonic stress, in turn attributable to this inferred high-pressure fluid. Indeed, one study has confirmed that there are no earthquakes deeper than 8 km beneath the high Alps, which coincides with the top of the zone of inferred high-pressure fluid influence. Geologists have recently established that the Alpine Fault has moved at least three times within the last 600 years—the approximate dates are 1717, 1630 and 1460. This information has been gleaned from painstaking investigations at southern and northern ends of the Alpine Fault, but not from the central part close to the highest section of the Alps. Debate among scientists continues: when the Alpine Fault moves, does it do so along its entire length or is it somehow segmented? What we can be sure about is that when end segments of the Alpine Fault move, they do so in eight-metre jumps laterally, and this must involve earthquakes of between magnitude 7 and 8 or more. There has long been a school of thought that suggests the central portion of the Alpine Fault is behaving differently from the end segments. One possibility raised by the work on fluid pressure is that strain release could be occurring by "slow earthquakes" in the central section. Japanese scientists have recently determined that such earthquakes exist. Using an array of 1200 sophisticated instruments recording continuously, these earthquakes have been monitored in Japan operating spasmodically over periods of days, weeks and months. They involve gradual movement of blocks of crust without the sudden failure of locked-up rocks and the ensuing short, sharp earthquakes. The key to such a process may well lie in the action of crustal fluids in reducing shear strength within rock, so allowing gradual slippage as strain along a fault increases. Our Alpine Fault may well be performing in exactly the same way. What is needed to determine what is happening here is Japanese-style heavy-duty continuous‑surveillance monitoring of the Alpine Fault, particularly in the central zone beneath the high Alps. New Zealand's new GeoNet hazard monitoring system will be able to do this, but with a much reduced number of instruments (perhaps 100). In Japan, the stakes are high enough to justify state-of-the-art monitoring instruments in large numbers (each unit costs about $200,000). After all, Japan has eight times the number of earthquakes that New Zealand has and is the world's second-largest economy.