NO ONE WILL EVER know exactly what Guard and his crew went through on that stormy October night in 1998, but demonic possession would never stand up in court. Given the stakes involved, APL subsidized an expensive forensic investigation led by Willa France.
France hired three meteorological consultants to hindcast the sea conditions with computers. Next, she hired oceanographers to computer-animate what would happen to a C-11 container ship under the conditions the China encountered, and the Maritime Research Institute Netherlands to conduct model tests in a wave tank three times as long as an Olympic pool.
If Parvez Guard was to be believed, the China, already hove to, couldn't have fallen prey to synchronous rolling. Furthermore, the waves were too close together to sync with the ship's roll period. But in 1973, experiments conducted with a scale model revealed that the hull shape of container ships made them vulnerable to a kind of rolling rarer and quite possibly more dangerous than synchronous rolling: "parametric rolling," so called because it occurs not when a ship's roll period is in sync with the waves but when the waves come exactly twice as fast as the ship rolls. Those model tests proved that parametric rolling could be excited by stern-quartering seasthat is, when the waves hit the ship askance on either corner of the stern. France set out to prove that it could also occur in bow-quartering seas.
Months after my trip aboard the Ottawa, she invited me to her apartment in an East Harlem brownstone, where she still lives with her wife of 35 years. While we ate sandwiches and pickles off silver plastic trays, France played me footage of the 1973 model tests. I watched as the toy boat first yawed a few degrees off its bearing, then suddenly began rolling hardso hard the damn thing keeled right over. Next came the computer animations of a C-11 container ship hove to in bow-quartering seas. Away the China's avatar went, pitching merrily along, up and down, over a red grid of giant waves until, suddenly, for no perceptible reason, something changed. The digital ship rolled. At first a little. Then more. Then more, until the bridge wings were tick-tocking like a metronome.
Then came the wave-tank experiment. Here again was a model boat, a replica of the China, which also went pitching merrily along in confused seas, actual ones this time, created by hydraulic paddles. For a minute it seemed as though the experiment would fail. "Nothing is happening," France said, narrating, remembering. "I'm beginning to bite my nails, because we've invested so much money in this and we're not seeing what the computer program has predicted." She stood next to the screen now to deliver her closing argument, pointing out details, like a weatherman delivering the forecast, a gray-haired weatherman in drag. "Now watch," she said. As in the simulations, something subtly changed. The toy boat rolled a little to starboard as the bow pitched down, a little to port as the stern rocked back. Steadily the amplitudes steepened. Finally there it was: full-on demonic possession, green water over the rails, waves swiping at the stacks.
Would any of these waves have qualified as genuine monsterwellen? Probably not. The wave the officers described was monstrously large (taller than France's brownstone) but, computerized hindcasts showed, not statistically improbable. The ship was rolling so heavily, dipping its bridge wings so close to the water as it pitched into the troughs, that even a 50-foot wave could have splashed them. Officers reported rolls as steep as 40 degreessteeper and more violent than the steel lashings had been designed to endure. At 40 degrees, the stacks of containers would have been almost as horizontal as vertical, and a mariner standing on a bridge wing would have been staring into the abyss.
The experiment proved that if Parvez Guard had done everything he'd been trained to do, if he'd heaved to and decelerated and the engines had not yet failed, the accident still would have occurred. And ironically, if Guard had not done what he'd been trained to doif he'd maintained his speed, for instancedisaster might never have struck.
According to documents I examined at a Manhattan courthouse, lawyers for the cargo owners had presented evidence that cast doubt upon France's explanation. There was, above all, the engine failure. Since some of the evidence was undisclosed, France was not permitted to discuss details, but she could speak hypothetically. If such a ship in such conditions had lost power, it would have drifted into beam seas and, because the waves were out of sync with it, rolled no more than 15 degreesnot enough to snap the lashings.
No judge ever decided whether France's explanation solved the mystery. APL settled out of court for an undisclosed amount. But France's discoveries did help limit APL's liability. "We got a very favorable settlement," France told me.
Not all of France's colleagues understood her fascination with the China disaster any better than they understood her poetry, or the mysticism of Buber, or, later, her metamorphosis, but in 2003, after the legal proceedings had all been settled, France published her findings in the journal of the Society of Naval Architects and Marine Engineers. A few months later, under the headline "Parametric Rolling Will Rock Insurers," Lloyd's List warned the marine-insurance industry of this "alarming new danger," which appears to be an unintended consequence of the flared bow and U-shape of the oversize, post-Panamax hull. France's findings not only helped explain the mystery of the China; they would later help explain what had happened to the Maersk Carolina, the P&O Nedlloyd Genoa, the CMA CGM Otello, and an unknown number of other maritime mysteries. Last July, the American Bureau of Shipping and the Polish Registry of Shipping announced that they were embarking on a "multi-year, joint research and development program" to find technology that would help prevent parametric rolling in "extreme wave conditions," hoping to exorcise once and for all the devil that possessed the China.