Circles of uncertainty
Hurricane forecasting and tracking is now far better than anything in years gone by, yet it is still far from being an exact science. Hurricanes are still a law unto themselves. When avoiding the hurricane is a matter of life or death, the situation is obviously different for a yacht in or within reach of a good hurricane hole, and one in the open sea.
As has been the case since 1495 when Columbus first discovered, to his cost, what native peoples called the “big wind,” the two most important rules are:
1. Don’t be there.
2. If a hurricane is coming, get out of the way as fast and as soon as you can.
The first is easy enough: just don’t be in the well-known danger areas during the hurricane season.
In regard to the second, there are long-standing rules about safe and dangerous sectors and what to do if caught in the path of a hurricane. Unfortunately, there are two major problems to these rules. Nobody knows where a hurricane is or where it is going. Obviously a hurricane is rather difficult to avoid if it suddenly appears from nowhere and comes chasing after you.
Even with satellites and hurricane-hunter aircraft, it is difficult to pinpoint a hurricane’s position because on a small scale a hurricane follows an erratic course, especially in the early oceanic stage. The quoted line of advance is only the average direction about which it is oscillating. The regular six hourly reports issued by the National Hurricane Centre in Miami quote an order of accuracy. This is very important and must be added to any reported position – which in itself will be several hours old. In the open ocean this accuracy is commonly 40 to 60 miles and can be as much as 100 miles. Only when near land and within range of airport radar (cynics say only when close to expensive real estate) does this decrease to 10 to 20 miles. Unless you are tuned into the Hurricane Center directly, the only reliable reports come from Coast Guard shore stations like NMN. Local radio stations ignore the very important factor of accuracy entirely. Every position report is given as if it was Holy Writ, quoted to a tenth of a degree – six miles – when it is more likely 60. Any local station report should be distrusted and assumed to be no better than 50 miles.
Consequently, a hurricane should never be plotted as a position but as a circle of uncertainty with a radius equal to the accuracy. Sixty miles is one degree. This shows how misleading it is to quote a position confidently to a tenth of a degree.
The Hurricane Center will forecast a position 12, 24, 36, 48 or 72 hours ahead. Analysis of these forecasts has shown an average error of:
Hours ahead Error in miles
An easy-to-remember, rough rule of thumb might be 50, 100, 150, 200 miles for the corresponding 12, 24, 36, 48 hours – and believe nothing for the 72-hour entry!
Therefore, the radius of the circle of uncertainty must be accuracy plus error. These are pretty big circles, and they extend in all directions. Just two days out it is a circle of six degrees, which would cover most of the chain of Caribbean islands. No ordinary voyaging yacht has a hope of outrunning a hurricane in that time, especially in the big seas that always build ahead of the storm. Moreover, it must not be overlooked that these are average figures. With luck they would be less. They could be more.
The daytime satellite observations have a smaller and more accurate footprint than the night-time observations that use infra-red. The eye of a hurricane, commonly 10 to 15 miles across, is a very small weather feature to spot from space compared with other ocean-wide weather systems. So the time of a satellite report must be considered in the accuracy too.
It is assumed, of course, that each six-hour position report is based on the latest information. This is not so. If no data is available it may be interpolated forward to where it is thought likely to be, admittedly by expert forecasters. Nevertheless, the hurricane may have other thoughts, and from experience it seems that it often does.
Sadly, it is not unusual for the six-hour report to be 12, even 24, hours out of date. When a hurricane gets near shore the systems can get overloaded. But I have even known them to be three months out of date.
Therefore, do not depend on getting accurate, up-to-date reports. Be prepared to backdate a wider circle of uncertainty. Local radio stations generally close down when a hurricane approaches and staff seek safety at home. (I would not want to be under a high radio mast myself!)
How reliable is the forecast direction? An old book, 2aritime Meteorology – A guide for deck officers, suggested an arc of 40° on either side of the forecast track. Modern technology and understanding of the steering forces has reduced that in the short term, but in the longer term that 40° arc is nearly tangential to the 48-hour circle of uncertainty.
Figure 1 shows how the circles of uncertainty might apply in real life. Suppose hurricane Minerva is in position A, accuracy 60 miles, on a course to hit Martinique in the middle of the Caribbean islands within 24 hours. B and C are the 12-hour and 24-hour circles of uncertainty. The 24-hour circle shows it could hit anywhere from Guadeloupe to the Grenadines. By now its position and likely short-term course are more predictable, but nevertheless the 40° vectors include the whole Caribbean, which, a few days ahead, could bode ill for Puerto Rico or the Virgin islands.
Now move ahead 12 hours to position D in figure 2. The circle of uncertainty, E, now more aptly a circle of panic, is smaller. The target is still Martinique, but even at this late hour Minerva could deviate and devastate Dominica or St. Lucia instead.
How close to a hurricane is it safe to go? What is the circle of where not to be? The Hurricane Center reports the distance from the center of 64-, 50- and 35-knot winds. They also quote 12-foot seas. This is more realistic than the dramatic strength of wind at the center. Depending on the strength and size of the storm these distances will vary. They could be 10 miles or more than 100. There is no safe rule. Each hurricane will be different. Obviously, the hurricane-force, 64-knot wind range is to be avoided at all costs since few yachts are likely to survive unscathed. In a strong hurricane, to be even closer to the center means almost certain death.
The 35-knot wind range is gale force 8. Not many voyagers can keep sailing in those conditions, bearing in mind that the quoted 12-foot seas may extend even further. Moreover, 12 feet is average height for what is certain to be a particularly rough, confused, and dangerous sea. That extends still further the circle of where not to be.
If the boat does get damaged and is drifting helplessly or is even Iying-to on purpose, the winds of a hurricane spiral inwards. The boat could, therefore, be drawn in toward the center and into ever greater danger. That is another risk to be considered in planning to sail clear, and the smaller the boat, the greater the likelihood of becoming helpless.
Conventional wisdom is that hurricanes always curve away to the northwest. Most do sooner or later. The critical factor is just how soon are they going to do so and where. Even the Hurricane Center often guesses wrong and its predictions vary from day to day. Never rely on a hurricane doing so before it reaches you. As Hurricane Lennie showed in 1999, hurricanes can go the “wrong way,” and there have been plenty of other cases in recent years to emphasize that, despite being steered by pressure systems, land masses, and other theoretical factors, hurricanes have a wild will of their own.
What does all this mean? In practical terms, there is a circle of uncertainty as to where the hurricane is now; considerably larger circles of uncertainty about where it will be in 12, 24, 48, or72 hours; and around all these must be even wider circles of where you do not want to be. The eye of the hurricane may be only 10 or 20 miles across, but these circles of safety are hundreds of miles.
The classical safe and dangerous semicircles of a storm are where the wind is intensified or reduced by the hurricane’s speed of advance. Remember that both semicircles are very dangerous.
For the average voyaging yacht, whether at sea or seeking a hurricane hole, these circles of uncertainty will be too large to cross in time. That means you must move well before the storm approaches, while you still have a chance to get clear. It must be emphasized that if you are in a circle of uncertainty, you are in serious danger. The hurricane may not actually move in that direction, but you just do not know. It is a gamble and you cannot take chances when the stakes are survival.
Figure 3 shows the situation in the open ocean. A yacht at position A hears that it is in the projected path of Hugo, a big, bad hurricane, in position 8, accurate within 60 miles, speed 15 knots or 300 miles a day. At that speed and course it will reach the yacht’s position in three days. Being somewhere south of Bermuda bound for Europe and reluctant to retrace their steps, they press on NE at an average rate of 120 miles a day. This is a fair average, because if the storm does come their way the rough seas will slow them.
The solid lines show the circles of uncertainty within which Hugo might be 24 and 48 hours later. The dotted lines are the circles of uncertainty of storm force winds, or circles of where not to be, which have to be added to the circles of uncertainty of the hurricane’s position. The radius is therefore: accuracy + average error + distance of storm winds from center. The circle of dangerous 12-foot seas may be even wider. This shows that on the second day the yacht could be in storm force winds and seas, and almost certainly it will be getting pretty rough.
After 72 hours the yacht is now in position C. The circle of uncertainty is very large, too large and uncertain to draw, and could be 500 to 1,000 miles across, but note how the yacht is still inside the 40 degree angle. The critical factor now is how close is Hugo going to be by then, bearing in mind that conditions may be so severe that survival, not avoiding action, will be the overriding consideration. If Hugo comes within the inner dashed circle, they are dead. No uncertainty about that. If anywhere within the middle circle they will get storm winds, from gale Force 8 at the outer edge, to lethal Force 12 nearer the center. The outer circle is the limit of 12-foot seas. These will be bad enough but note that this is the outer limit. Inside that circle they will be progressively more dangerous. The sober official description is mountainous
It is not just the center of the hurricane that must be avoided. The circles of uncertainty must also show the storm-force winds and seas that surround it. The National Hurricane Center quotes what it considers are the distances from the center that such conditions extend, and these are relayed by NMN but not local stations. Actually they are usually quadrants, not circles, but the principle is the same. Keep out.
Near land these circles will also warn you of places where you could possibly be trapped in the classic disaster situation, embayed and without searoom, like the poor Fantome was.
Conventional wisdom, if one can use that expression for conventional ignorance, is to take all information about the position and direction of a hurricane as accurate and precise. This is misguided, foolish, and very dangerous. Everything must be planned on circles of uncertainty, and beyond these, considerably wider circles of where you do not want to be. Even then you should bear in mind that the hurricane may not follow the rules.