WEAVER

Weaver felt a rush of euphoria. She only had to enter her password and the laptop gave her access to more information than she'd ever imagined. Under normal circumstances it would have taken her months to gather the kind of data she had here – and even then the military satellites would always have been off-limits. But this was amazing! She could sit on the balcony of her suite, log into NASA's server and immerse herself in the American military's satellite maps.

In the 1980s the US Navy had begun to investigate a remarkable phenomenon. Geosat, a radar-imaging satellite, had been launched into a near-polar orbit. There was no provision or possibility for it to map the ocean floor – radar was incapable of penetrating water. Instead, Geosat's mission was to measure sea-surface heights to within a few centimetres. It was thought that by charting great expanses of water it would be possible to show whether the sea level – tidal fluctuations aside – was the same across the planet.

Geosat's findings exceeded all expectations.

Scientists had suspected that the oceans were never completely smooth, even in conditions of perfect calm, but Geosat's images made the planet look like an enormous, lumpy potato. The oceans were full of dents, humps, bulges and troughs. For a long time scientists had assumed that the water in them was spread evenly across the globe, but the map offered a different picture. Off the south coast of India, for example, the sea level was 170 metres lower than it was in the waters around Iceland. To the north of Australia, on the other hand, it rose up to form a peak eighty-five metres above the mean sea level. The oceans were vast mountainscapes whose topography seemed to follow the lie of the underwater landscape. Towering underwater mountain ranges and deep ocean valleys replicated themselves on the surface with only a few metres' difference in height.

It all came down to variations in gravity. An underwater mountain gave the sea floor additional mass, so its gravitational field was stronger than that of a deep-sea valley. It pulled the surrounding water towards the mountain and made it pile up in a hump. The water surface bulged above a mountain – and dipped above a trench. For a short while, a number of exceptions kept the scientists guessing – for example, when water piled up above a deep sea plain – but in the end it transpired that some of the rock on the seabed was denser and heavier than average, and with that the gravitational topography fell into place.

The slopes of the water's mounds and valleys were too gentle for any sailor to detect. In fact, if it hadn't been for satellite mapping, no one would have stumbled on the phenomenon, but now scientists could use their knowledge of the surface to deduce what was happening in the depths. It was more than just a new method of charting the topography of the seabed: it was a key to understanding ocean dynamics. Geosat had revealed that powerful currents circled in the oceans, forming eddies that measured hundreds of kilometres across. Like coffee being stirred in a mug, the rotating masses of water formed a depression at the centre, while the outer rings rose upwards. It became apparent that these eddies also caused the ocean's surface to rise and fall, independent of gravitational variations, and that they themselves were part of far larger rings of water – oceanic gyres. From the long-distance perspective of satellite mapping, it became clear that all the world's oceans were rotating. In the northern hemisphere, enormous networks of rings spun in a clockwise direction, while in the southern hemisphere the flow was anti-clockwise. The speed of rotation increased with proximity to the poles.

This allowed scientists to prove another fundamental principle of ocean dynamics: the rotation of the planet determined the speed and direction of the gyres.

Logically, therefore, the Gulf Stream wasn't a stream, but the western boundary of an enormous vortex made up of smaller eddies: a gyre rotating slowly, and pushing towards North America in a clockwise direction. Because the whirlpool wasn't in the centre of the Atlantic Ocean but to the west, the Gulf Stream was pushed against the American coast, where the water piled up in a ridge. Strong winds and the poleward flow of the water increased the speed of the swirl, while the immense lateral shear with the coastline slowed it down. As a result, the north Atlantic whirlpool was rotating in a steady circular current, in line with the principle of angular momentum, which ruled that circular movements remained stable unless disrupted by an external force.

And it was the possibility that the current was being disrupted that Bauer had feared. He'd been trying to find proof Water had stopped cascading into the Greenland Sea, which was alarming, but not decisive. Proving the existence of global changes meant obtaining data on a global scale.

In 1995, after the Cold War had ended, the American military had begun to release the Geosat maps and the system had been replaced with a string of new satellites. Karen Weaver had access to all their data, which combined to form a complete history of oceanic mapping from the mid-nineties onwards. She spent hours trying to match up the different readings. There were variations in detail – sometimes a satellite's radar altimeter would mistake a thick bank of mist for the surface and record a measurement that was disputed elsewhere, but in general the results were the same.

The closer she looked, the more her initial excitement gave way to anxiety.

In the end she was certain that Bauer had been right.

His drifting profilers had transmitted data for only a short time, without seeming to follow the path of any current. Then one after another, the floats had fallen silent. Practically no feedback was available from Bauer's expedition. She wondered whether he had sensed how right he'd been. She could feel his knowledge weighing on her shoulders. He had entrusted her with his legacy, and now she could read between the lines. She knew enough to grasp that a catastrophe was looming.

She went back through her calculations and checked for mistakes. She repeated the process again and again. It was worse than she'd feared.

ONLINE

Still in their PVC suits, Johanson, Oliviera, Rubin and Roche stood under the decontamination shower. The vapour from the solution of 1.5 per cent peracetic acid was guaranteed to obliterate every last trace of any lurking biological agent. Once the caustic fluid had been washed away with water and neutralised with sodium hydroxide, the scientists were permitted to leave the sealed chamber.

SHANKAR AND HIS TEAM were working round the clock in an attempt to make sense of the unidentified noises. They'd called in Ford to help them, and were busy playing Scratch and other spectrograms over and over again.

ANAWAK AND FENWICK had gone for a walk and were deep in conversation about possible ways of hijacking an organism's neural system.

DR STANLEY FROST had turned up in Bohrmann's suite. His baseball cap was pulled down over his glasses and his massive figure seemed to fill the room. 'Right, Doc, it's time we talked,' he boomed.

He explained his thoughts on the worms – interesting, all in all. He and Bohrmann clicked right away, drank a few beers at lightning speed and came up with a series of disturbing, yet plausible scenarios to add to the list of possible disasters. Now they were conferring via satellite with Kiel. Since the Internet connection had been restored, the Geomar scientists had been sending a steady stream of simulations. Suess had reconstructed events on the Norwegian slope as accurately as possible, leading them all to the conclusion that a catastrophe of such magnitude should never have occurred. The worms and bacteria had certainly had a dire effect on the slope, but something was missing: a tiny piece of the jigsaw, an additional catalyst.