Hurricane forecasting methods and accuracy

¶ … scientists predict the number of hurricanes expected during a season and how they predict the development and motion of a single hurricane. Hurricane forecasting is a difficult process, but it has developed into a science that is usually capable of judging the velocity and trajectory of a hurricane in order to help save lives and property. The National Weather Service oversees the forecasting process via the National Hurricane Center and the Tropical Prediction Center. These centers are the country's first line of defense in hurricane forecasting, and they serve a vital purpose for anyone who could possibly suffer from a tropical storm or hurricane.

The National Weather Service (NWS) has a Tropical Prediction Center (TPC) that specializes in predicting and tracking tropical storms, including hurricanes. The National Hurricane Center Web site notes, "The TPC is comprised of the National Hurricane Center (NHC), the Tropical Analysis and Forecast Branch (TAFB), and the Technical Support Branch (TSB). During hurricane season, the latter two provide support to the NHC" ("Forecasting Process"). There are several stages of prediction, from initial observation to warnings released to the public.

Observation is the first element of hurricane forecasting, and it includes several methods of observing and tracking hurricanes and tropical storms that could develop into hurricanes. Observation begins when the predicted storms are far away from land and still far out to sea (all hurricanes form over oceans). The first method of observation is satellites, or the Geostationary Operational Environmental Satellites (GOES). These satellites are used to provide a wide variety of data on the forming hurricanes, such as "location, size, movement, and intensity of a storm, and analyze its surrounding environment" ("Observations"). In addition, the satellites emit radiation that reflects from the atmosphere, which allows the satellite to measure moisture, winds, atmospheric temperature, and cloud cover of the storm. All of this data is transmitted to the NHC, where it is analyzed.

While satellites are the main form of observation, the NHC also relies on ships and buoys, which determine wind speed, pressure, wave conditions, and temperature. These ships and buoys are vital as the storm moves closer to land, because they are the only way to measure waves that are unobstructed by land, and so, they provide real-time measurements to back up readings from satellites and radar ("Observations"). The next form of observation is reconnaissance aircraft, which deploy as the storm moves closer to land. They measure the pressure, winds, humidity, and temperature at the center (or "eye") of the storm, and they are crucial for location of the actual eye. These measurements are extremely important in forecasting the velocity of the storm, but they are not continuous, so the storm can change, throwing the forecasters off guard. The NHC Web site states, "In addition, the measurements are not taken continuously or throughout the storm, so what we have is a snapshot of small parts of the hurricane. Nonetheless, that information is critical in analyzing the current characteristics needed to forecast the future behavior of the storm" ("Observation"). The National Oceanic and Atmospheric Administration (NOAA) also fly aircraft into developing storms to help develop long-term forecasting solutions and more knowledge about hurricanes.

There are also some measures used to record and measure data once the hurricane hits land. Radiosondes are instruments carried aloft by balloons that measure pressure, humidity, and air temperature. Dropsoundes are dropped from aircraft by parachute, and they measure the same items. Both of these instruments are helping scientists better understand hurricanes, and this can help them predict with more accuracy in the future ("Observations"). Finally, radar is used to monitor the storm and its intensity as the storm gets nearer to the coast. Radar is one of the most dependable of the observation tools, and it can tell scientists just where a storm is headed, how much rainfall can be expected, and other data. Unfortunately, radar can only pick up these signals about 200 miles from land, so the storm has to be close to landfall to utilize this data. Finally, there are surface monitoring stations that collect data as the hurricane passes them, but these are the last forms of observation, as they are limited because they can only measure the hurricane for a relatively brief period. As the observation process shows, there are numerous ways to track and measure a hurricane as it forms and moves toward land, but many of them are not available in the initial stages of formation, and so, it is difficult to create long-range forecasts with these observation methods.

After observation of the single, developing hurricane, scientists use several methods of analysis to chart the storm's intensity and progress. They use complicated computer models from a number of different sources to analyze and predict the hurricane's path and intensity. Unfortunately, there are several different models, each with specific purposes, and often the models do not agree. In addition, the models are slow to produce results. The NHC Web site states, "It is important to know these models are only run a few times a day and cannot, therefore, take into account all the short-term changes the atmosphere is constantly undergoing. Models cannot produce forecasts more frequently because they require huge amounts of data and long computational times" ("Hurricane Models"). All of these problems can lead to forecast error and that can lead to dangerous situations if a hurricane hits land in an area where it was not predicted to hit.

Throughout hurricane season in the Atlantic (June through November), as a storm intensifies, there is a six-hour forecast cycle that follows a set procedure that monitors the storm and releases public data periodically. Hurricanes begin as tropical cyclones, and are classified as hurricanes once their winds reach 74 MPH or higher. During the six-hour cycle, the storm's intensity and location are tracked, and the cycle begins again every six hours until the storm's end. During a hurricane watch or warning, updates are issued immediately as they are released.

While predicting the movement and intensity of a single hurricane is difficult, it is even more difficult to predict the number of hurricanes that can be expected during a hurricane season. Just as with predicting individual hurricanes, there are several factors that help scientists to predict these numbers ahead of the upcoming hurricane season. Climate conditions and predicted climate situations such as La Nina, which is a weather pattern of increased air circulation in the area, leading to the formation of more clouds, rain clouds, and tropical storms. In addition, temperatures in the area have been increasing, and this adds to the probability of more hurricanes in the area. These situations are analyzed every year to come up with predictions of how many storms will form, how many will turn into hurricanes, and how many will turn into major hurricanes that hit land.

These forecasters also use prior history as a large portion of their predictions, which means that they rely on their own experience combined with prior years' prediction and actual hurricanes to help them decide how many hurricanes will occur during the current season. Currently, this predictor, called the Hindcast Period, goes back to 1950, and it is one of major components of the current prediction model. Two professors at Colorado State University have developed a model to predict the upcoming season's storms as early as the December prior to the upcoming hurricane season, and their models have been extremely accurate in forecasting the number of hurricanes that will occur. Predictions are updated in June and August during the hurricane season, as well. The forecast for the 2008 hurricane season, issued in December 2007, indicates a "somewhat above-average Atlantic basin tropical cyclone season in 2008, along with an above-average probability of United States major hurricane landfall" (Klotzbach and Gray 1). Prior data shows that this early prediction by these two Colorado State professors has a high reliability factor, which means the U.S. hurricane season may be greater than average this year, with a higher probability of at least one hurricane touching land.