Navigation: Introduction → Preliminary Work → Creating a Model Using Weka → Results and Implications

Results and Implications

The Societal Implications of Hurricane Forecasting

Hurricanes are destructive and deadly storms that directly impact millions of people and indirectly affect hundreds of millions more. Fortunately, humans have recently learned how to predict the formation, location, and intensity of hurricanes with a respectable amount of accuracy. Prior to the installation of telegraph lines, people had no advance warning of an approaching storm. From the 1850s until around the end of WWII, hurricane forecasting relied almost completely upon the rapid communication of surface observations taken in the presence of the storm to locations thousands of miles away. Over the past six decades, the use of radar, satellite, and radiosonde observations along with dynamic numerical weather prediction models has allowed for the detection and highly accurate forecast of tropical systems. In discussing the societal impacts of hurricane forecasting, two case studies are presented. First, the 1900 Galveston Hurricane (with little to no advance warning) killed an estimated 8,000 people and $494 million in damage (adjusted for inflation in 2007). In contrast, Hurricane Rita (with days of advance warning) resulted in 7 deaths but caused $11.94 billion in damage.

As previously mentioned, the only warning that citizens in 1900 had of an approaching storm was through telegraph reports from stations already affected by the hurricane. For example, if locations in Cuba reported hurricane conditions, the western Caribbean Sea, the entire Gulf of Mexico, and the entire Eastern Seaboard of the United States would have a seemingly equal likelihood of being affected by the storm. Since there was essentially no method of forecasting the future location of tropical cyclones, there was little incentive for people to prepare in advance. Logically, there was even less reason for people to evacuate low-lying coastal regions. In fact, given the lack of the automobile and fixed schedules of the railroad, evacuations had to take place with enough to allow for travel time and time spent finding lodging, which could take five days.

Even if people in 1900 were able to predict with perfect accuracy the location of a hurricane five days into the future, a lack of intensity forecasting would have prevented any meaningful preparations from being made. Category Five hurricanes rarely undergo further intensification, and tropical storms can rapidly strengthen to a major hurricane within 24 hours. Thus, people rarely (if ever) prepared for a hurricane hit around the turn of the 20th century.

Reports that a hurricane was affecting the eastern Gulf of Mexico reached Galveston on September 7, 1900. In addition, approaching clouds and high seas indicated the approach of a hurricane; hurricane warnings were issued that afternoon. However, because skies over the city of Galveston were partly cloudy, many residents ignored the warning. Unfortunately, since Galveston was only connected to the mainland by ferry, the residents’ ignorance of warnings proved to be deadly. The next day, the hurricane made landfall at Galveston, with winds over 100 mph. Destruction was so great that it took nearly 24 hours for the news to reach mainland Texas; by that point, very few of the injured could be rescued. With 8,000 dead, the Galveston hurricane of 1900 is an example of the damage a surprise (for all intents and purposes) hurricane can inflict.

A little more than 105 years later, Tropical Storm Rita became a hurricane on September 20, 2005. The next day, Rita strengthened to a Category 5 Hurricane with maximum sustained winds of 180 mph. Meteorologists updated the forecast track of the storm several times a day, and issued a hurricane watch for the western Gulf coast on the 21st – almost four days before Rita made landfall early on September 24th. Given the extreme intensity of Rita and with memories of Hurricane Katrina still fresh in people’s minds, evacuations of Galveston began almost immediately following the posting of the hurricane watch; interstate highways out of Houston switched to their contra-flow plans (both sides of the highway moving in the same direction) the next day. In some ways, the advance warning of the storm was almost too good; traffic slowed to a standstill and more people died during evacuation than from the storm itself. However, Rita was more than 24 times as costly as the 1900 Galveston hurricane, a result of the rapid development of the Gulf coast areas in the latter half of the 20th century.

Currently, hurricane forecasting is accurate enough to allow for people to know the general intensity of the storm with enough time to evacuate to safety. However, the steady migration of people to coastal locations presents an increasingly challenging problem to government officials, who must come up with plans to safely evacuate millions of people. Unlike in 1900, when very few people were aware of an approaching storm, today’s technology and communication advances make it extremely difficult not to know that a hurricane could strike within the next several days. Therefore, the majority of direct hurricane deaths now occur when people are unable to move to safety. This is still a major problem in third-world countries and, unfortunately, in large cities such as New Orleans or Houston. While there may never be a way to prevent any fatalities from hurricanes and resulting complications, modern forecasting abilities have dramatically reduced the threat these storms pose to human safety.

Societal Implications of Our Forecast Model in Particular

Our hurricane forecast model predicts five parameters three hours into the future: latitude, longitude, storm intensity, the radius of gale force winds, and the radius of maximum winds (the parameters were listed in order of decreasing quality). Errors in predicting these values have various impacts to society, as detailed below.

Predicting latitude is fairly straightforward, since storms rarely move southward. However, there are times when storms are forced southward by a strong upper-level ridge or wander aimlessly due to a lack of steering winds. Coastlines that are mainly north-south oriented, such as the western Gulf of Mexico, western Caribbean Sea, and the U.S. east coast, are most dependent on accurate forecasts of latitude. For example, a hurricane with a westward component has the potential to make landfall somewhere between Miami and New York City in the next 60 hours, depending upon the north-south motion of the storm. Given the high population density of the East Coast, it is unfeasible to tell the hundreds of millions of residents between these two cities to prepare for a hurricane landfall; showing the importance of accurate latitude forecasting.

Longitude prediction, while not as simple as latitude, is generally simple in the short-term. The main factor in the east-west storm motion is the direction and strength of upper-level steering winds. Coastlines that are generally oriented west-east stand to benefit most from an accurate longitude forecast. If a storm is moving northward in the Gulf of Mexico, it has the potential to strike an area from Lake Charles, LA to Tallahassee, FL. Evacuation routes along the Gulf Coast are complicated because of limited north-south interstates, and contra-flow plans for Louisiana split their main east-west interstate to diverge from New Orleans. However, if a hurricane were to make a last second turn towards the east or west, the contra-flow plan becomes useless since people can not turn around. Again, since millions of people live along the Gulf coast, hurricane evacuations for the entire region are unfeasible and accurate longitude predictions are desirable.

Intensity prediction is affected by a number of factors, but in the short-term, the most important factor is what the storm did over the previous three hours. Short-term forecasts of hurricane intensity are not of much use, since people who have chosen to stay will not have time to evacuate if rapid intensification occurs just prior to landfall. However, long-term intensity predictions are extremely valuable to any coastline, as well as ships and the oil industry. An inaccurate forecast of intensity leads to a skeptical public (which leads to problems with future storms) and unnecessary economic damage.

Forecasting wind radii are important for ships, whether on the dock or out at sea. Small boat owners will decide whether or not to secure their vessel depending on the forecast of gale-force winds. With regards to the radius of storm-force winds or the radius of maximum strength winds, commercial freight and Navy ships stand to benefit the most. If strong winds are forecast to pass over a popular shipping lane, routes must be diverted. If strong winds are forecast to be present near Newport News, Virginia, the U.S. Navy will move enormous ships a safe distance away. If forecasts of maximum wind radius are inaccurate, millions of dollars are lost and naval vessels could be destroyed.

Previous Page