“With sloping masts and dipping prow
As who pursued with yell and blow
Still treads the shadow of his foe,
And forward bends his head,
The ship drove fast, loud roar’d the blast
And southward aye we fled”.
The Rime of the Ancient Mariner – Samuel Taylor Coleridge
At 1 pm sharp on 26 December 1998, the starters gun boomed across Sydney Harbour and 115 yachts began the Sydney to Hobart Yacht Race in sparkling summer conditions and a gusty north-east sea-breeze. Followed by a large spectator fleet that traditionally accompanies the racing yachts as far as the Heads, some 1135 sailors began the 1000 km ocean race to Hobart, recognised as one of the world’s blue-water classics.
The race follows the NSW coastline down to the Victorian border near Gabo Island, then out across the notoriously changeable waters of Bass Strait, followed by the run down the east coast of Tasmania and finally the last leg up the Derwent River into Hobart.
The weather at this time of the year across the race area can be highly variable. North-east sea-breezes are common during the afternoon along the NSW coast, and on some occasions these have held for much of the race, allowing for a prolonged fast spinnaker run southwards. But this is also the time of the famous “Southerly Buster” which is a squally southerly change that originates over the east coast of Victoria and then flies northwards up the NSW coastline, often generating winds of around 30 to 40 knots. Busters have played a key role in deciding many past race results, with tactics that best incorporated the timing and intensity of the change often proving decisive. As another quirk, races have sometimes ended with yachts being becalmed in the Derwent River, almost within sight of the finishing line.
The race fleet’s demand for weather forecast information is therefore hardly surprising. Traditionally a close relationship has existed between the Bureau of Meteorology and the Race Organising Committee at the Cruising Yacht Club of Australia (CYCA). For around the previous twenty years or so, a meteorologist from the Bureau would conduct a pre race weather briefing for all the crews around the 24th of December, normally held at the Club’s premises at Rushcutters Bay, Sydney. Special meteorological support was also arranged for the race in which daily or twice daily weather forecasts and warnings were issued and transmitted to the fleet.
This arrangement was still in place on 24th December 1998, when some 250 yachtsmen converged on the CYCA clubrooms at 9 am to hear the latest on the weather. This was not the issue of a formal race forecast as such – the first of these would be prepared two days later, on the morning of race day itself. Rather it was a general “weather outlook” and this indicated that the race should begin Saturday in light conditions, but that a strong southerly change was possible late in the day.
Much of this information was based on what are called numerical weather simulations, and these have been one of the major areas of progress in meteorology over the last 30 years or so. They utilise mathematical equations that describe the motion of the atmosphere and these are combined with thousands of weather observations from around the world and fed into supercomputers to produce a simulation of the weather, normally out to a week in advance. The accuracy of these simulations is normally highest when looking at the period out to 24 hours ahead, but this then tends to drop off out towards seven days.
Many countries have produced their own national weather simulations and these are freely exchanged in the interests of improving weather forecasts on a global basis. The Bureau of Meteorology in Australia has developed its own excellent weather simulation that is routinely used in all weather forecasts, but, in addition, has access to simulations from other international sources that include the US, UK, Japan and Europe. Each simulation has its own biases and individual “quirks” and comparing the output from each is very valuable for the meteorologist. If all the simulations are indicating similar outcomes, then the meteorologist becomes more confident of the forecast. If not, confidence is reduced and forecasts can be constructed to reflect this.
Towards the end of the race briefing, the Bureau spokesman mentioned that one of the simulations operated by the European Centre for Medium Range Weather Forecasts (ECMWF), was indicating the possible development of a low pressure cell to the south-east of Gabo Island sometime on day two of the race. This would be worth watching in the run up to the start.
Race Day dawned, and soon after 9 am, the official race forecast was issued to the fleet. The southerly change mentioned at the briefing two days before was still in the forecast but was now expected to be stronger than initially indicated and was expected to reach Jervis Bay around midnight to 2 am on Sunday morning. A gale warning had been issued for all NSW coastal waters south from Broken Bay, to cover the passage of this change, and it was with this knowledge the fleet finally set to sea.
However, soon after the start, high drama was emerging back at the Bureau Offices in Sydney’s Elizabeth Street. Some of the simulations, including the Bureau’s high-resolution version that became available soon after race start, were now indicating that a strong low-pressure cell was likely to develop to the south over the next 24 hours. In particular the Bureau’s simulation was indicating explosive development of an intense low pressure cell virtually on the race track, just to the east of Bass Strait during Sunday.
Poring over this new data, the race meteorologists became increasingly concerned, and took the unprecedented step of preparing a storm warning, the first time in race history that such an action had been taken. A storm warning is the highest category warning issued for waters in these latitudes and is only surpassed by a hurricane warning, which is used in the event of a strong tropical cyclone. The storm warning was then issued with the race just over an hour old and distributed to a prepared schedule of recipients.
As the yachts moved southwards down the coast the weather progressively deteriorated, with the wind and sea steadily rising. The leading yachts began to encounter storm conditions about 18 hours into the race, with winds averaging around 50 knots and gusting to as high as 75. The conditions peaked across eastern Bass Strait during Sunday, as the low pressure cell predicted by the simulation intensified and tracked right through the fleet, creating incredible havoc and destruction.
Left: The synoptic chart at 3PM on 27th December 1998. An intense low pressure cell had developed explosively right on race track. The low is also plainly visible on the satellite image taken at the same time (below).
Images courtesy of the Bureau of Meteorology
For many competitors, racing was forgotten and it became a straight out fight for survival in the mountainous seas and shrieking tempest. Many yachts were forced to retire and several crewmen were injured after being flung about their vessels.
Ultimately 55 sailors were saved from the mountainous seas through a huge rescue operation involving the Australian Maritime Safety Authority, the Navy and the Air Force. Particularly heroic efforts were performed by helicopter pilots in winching sailors to safety in the incredibly dangerous flying conditions. In the end, five boats had sunk, sixty six retired and only forty four made it to the finish line. Most tragically of all, six crewmen had died in the maelstrom.
A lengthy Coronial Enquiry was held and the findings finally released
in December 2000. The Coroner, John Abernathy was critical of certain aspects of the conduct of the Cruising Yacht Club’s Race Committee and also recommended changes to a range of safety gear to be carried aboard in future races.
As far as the weather services were concerned, he recommended
"That weather forecasts which are specifically provided for yacht fleets contain:-
(a) As well as the average winds expected, the maximum gusts of winds that are likely to occur.
(b) As well as the significant wave heights expected, the maximum wave heights that are likely to be encountered
This increased level of detail is now part of all the forecasts issued by the Bureau in support of the race.
Reference: Australia's Natural Disasters, Richard Whitaker
New Holland Publishers, ISBN 1877069043