Waterside News
Site under construction
Site under construction
Hurricane Forecast
| Another Hurricane Season is here. Are we doomed? |
| By Chuck Watson Director of Research and Development, Kinetic Analysis Corporation With June comes the usual round of predictions and warnings about the upcoming hurricane season. University, government, and private sector researchers rush to the microphones to announce their latest forecasts for hurricane activity. So how do these forecasts work, and what do they really tell us? There are two main factors that control how active a hurricane season will be. The first is a 30- to 40-year cycle known as the Atlantic Multidecadal Oscillation, or AMO. The AMO is part of a global cycle of changes in the thermohaline (temperature/salinity) circulation in the oceans. During warm phases, the average temperature of the Atlantic is as much as one degree above the long term average; during cold phases, it can dip a degree below average. That may not seem like much, but think about how much energy it takes to change the average temperature of a body of water as big as the Atlantic! During the warm phase, there is more energy for hurricane development and growth. The AMO sets the “big picture” for activity. The second major factor controlling hurricane activity is El Nino/Southern Oscillation, or ENSO cycle. The ENSO cycle is a 2- to 3-year cycle most well known for changes in temperature of the Pacific Ocean off the coast of South America. There are three phases to the ENSO: warm (El Nino), neutral, and cold (La Nina). While the exact process is a complicated interaction between the ocean and atmosphere, the results are fairly simple. When the Pacific is warm, winds over the Atlantic are unfavorable for hurricanes. When Pacific temperatures are near average, hurricane activity is also near average, and when the Pacific is cold, winds are quite favorable for hurricane activity, especially for storms striking the United States. In addition, the ENSO cycle controls many of the details of where storms go during the season. So, in short for the AMO, it's "warm bad, cold good," and for El Nino, "warm good, cold bad." Unfortunately, the forecast for this year is for the warm AMO phase that started in the mid-1990's to continue, and for a La Nina to develop starting in July. This is the worst possible combination of the two; for this and other reasons, it seems likely that hurricane activity will be well above normal this year. There are basically three ways of doing seasonal forecasts. The first, pioneered by Dr. William Gray and his team at Colorado State University, uses statistical correlations between various climate signals to estimate the number of storms. The second uses some statistical analysis, supplemented with computer models, to compute activity levels. The National Weather Service uses this technique. A third method, developed by me and Dr. Mark Johnson at the University of Central Florida, starts with the second method, but expands it by running over 100 computer simulations of the current year, using the weather between January and May as a starting point. Each run diverges from previous runs due to the complex nature of weather combined with the uncertainty of observations (especially over the oceans), so we get 100 “possible” 2010 seasons to analyze. The advantage to our approach is that we can compute the chances of a strike on any given location, not just the number of storms. For what it's worth, my team is predicting that there is nearly twice the normal chances for a hurricane to strike the Georgia coast this year as opposed to the long term average. However, knowing the number of storms, the activity level, or  even the probabilities, may not be much comfort to the individual living on the coast. By virtually any measure of seasonal activity, 1992 was a very quiet year. Except for Hurricane Andrew, which caused $40 billion in damage (in 2009 dollars). Other years, with twice as many storms, caused zero damage in the US. Another colleague of mine, Dr. Steve Lyons, calculated the correlation between the number of storms and the number landfalls, as well as the number of storms and losses. There was virtually no correlation – in other words, even if you know exactly the number of storms that will happen in a year, that tells you nothing about your chances of getting hit. So, given that the fact that seasonal predictions are almost worthless even if they were 100% accurate, and probabilities are mainly of use to insurance companies, governments, and stock speculators, what should you do? My advice is to treat every year as if it was “the year.” Make sure your insurance records are up to date. Make a sweep through your house with a video camera, jot down model and serial numbers for high value items, shoot video or pictures of your cars, boats, and other items as well to prove their pre-storm condition. If your boat policy covers portable navigation equipment like GPS, make sure you have that recorded. Think through the logistics of your plan – under what conditions will you shelter in place, when will you evacuate, if you will relocate your boat will you need to move it then come back for the family? Will your family meet you somewhere? How far in advance of a storm do you need to start acting to get things done in time to reach safety? These are the sorts of things to decide in advance. Don't forget pets – some shelters and hotels still will not take them during evacuations. Living on the Georgia coast has a lot of perks. But the down side is that in any given year there is about a 1 in 10 chance of at least a tropical storm, a 1 in 20 chance of a storm bad enough to cause significant damage, and a bit over a 1 in 120 chance of near total devastation. Those are not great odds, but they are not terrible either, and other parts of the world have risks as well. The bottom line is don't be scared, be prepared. |
