You can copy and paste this URL.
This URL will permanently link back to this page.
Imagine the following scenario. A strong hurricane originating southeast of Florida accelerates toward New York. The Hudson River is swollen due to a rainy period. Computers indicate that the hurricane’s eye will hit Manhattan at maximum tide. Under these circumstances, the water surge could reach 22 feet. What follows looks very much like what recently happened in Japan, just that the number of victims could be much higher.
Though this apocalyptic narration seems to be taken from a Hollywood script, the matter is not about if, but about when, it will happen. The Big Apple has experienced high levels of the tide, a swollen Hudson River, and devastating winds, luckily not all at once. But, as probability theory teaches us, simultaneity is assured in the long run. Unfortunately nobody knows when. It may happen next summer or in 800 years. With our current knowledge, we could foretell it only 2 or 3 days in advance. This time interval seems far from satisfying. Still, a short prediction is better than being taken by surprise, as it usually happens with earthquakes, and is no worse than the few-minute warning the Japanese had with their recent tsunami.
This catastrophe raises questions many people around the globe must have asked these days: How good is science at predicting natural disasters? If a highly developed country like Japan, with its knowledge about earthquakes and tsunamis, experienced so many victims recently, how safe is everybody else?
The answers to these questions depend on each event. Earthquakes and pandemics, for example, are among the hardest to pin down. The experts have difficulties even with probabilistic forecasts, which are not very helpful anyway. But other catastrophes, such as hurricane hits and volcanic eruptions, lie at the other extreme. These disasters can be usually predicted days before they occur.
So far, seismologists had little success with useful predictions. Their most remarkable forecast took place on February 4, 1975, when an earthquake of magnitude 7.3 struck the city of Haicheng in northeast China, killing more than 2,000 people, injuring almost 28,000, and damaging 90 percent of the buildings. But in spite of this tragic outcome, the local authorities claimed victory. Six hours before the event, some experts had warned about the disaster, and most inhabitants stayed outdoors. Had this prediction not been made, some 150,000 people would have been crushed under the rubble.
The Haicheng success, however, turned out to be a lucky overlap of events. In the late 1960s the State Seismological Bureau in China had identified the city’s area as a seismic-prone zone for the next few years. Several tectonic movements, including a land shift in the nearby Bohai Sea, led in 1974 to a forecast for a major earthquake within 2 years. An intense educational campaign kept the population on alert, so when a warning for an imminent shock was issued, a combination of organized and spontaneous activity drove the people outside their homes. At 7:36 PM, when the earthquake struck, most citizens were neither at work nor asleep. They had stayed outdoors, though the prediction had given no precise time and magnitude. No wonder that most alarms issued since have been false.
Earthquakes are hard to predict because we know little about where the tectonic plates meet. Using the GPS, seismologists are now trying to improve this aspect of their work. Currently, however, they are also focusing on real time warnings. A program of this kind will be implemented next year in San Francisco. From the time a sharp tectonic shift takes place until the shock waves, which propagate with the speed of sound, reach the city, some 30 to 50 seconds may pass. Instant warnings would give people time to take cover or run outside the buildings, trains could stop to avoid derailment, and gas supply could be cut off, thus diminishing the risk of explosions and fires. Though far from ideal, real time warnings are going to save lives.
Unlike with earthquakes, we are much better at predicting the path of hurricanes. A century ago, the only warning our grand-grand parents could hope for was a sudden and dramatic drop of the atmospheric pressure an hour or so before the disaster would hit. Today, in an age when we can watch these giant whirlwinds live on our TV or computer screens, scientists can issue timely warnings. What they still find hard to forecast is the strength of the hurricane, but their estimates are improving.
The sad consequences of Hurricane Katrina were not triggered by lack of prediction. Scientists and engineers had sounded the alarm well in advance, but those who could act didn’t listen. A repeat of those mistakes would be much more catastrophic in an event similar to the one described in the beginning.
Volcanic eruptions offer other examples of successful prediction, thanks to a science that didn’t even exist at the end of the 19th century. A relevant case is Mount Saint Helens near Seattle. On March 12, 1982, the experts warned of an eruption within the next 10 days. On March 15 they narrowed the window to 4 days, and on March 18 to 48 hours. The volcano exploded the next day. At that time, the areas around the mountain had been safely evacuated. Even though not all volcanoes are as well understood as Mount Saint Helens, the science studying them is growing fast.
As for tsunamis, nothing could be more edifying than the example of Tilly Smith, a 10-year-old English girl, who saved the lives of more than 100 people on the Mai Khao beach in Thailand on the morning of December 26, 2004, while vacationing there. When Tilly came down from the hotel with her mother, the sea looked bubbly and frothy like on the top of a beer. This image triggered some memories in her mind. Two weeks earlier, during her geography class, Tilly saw a movie about the Hawaii tsunami of 1946. She recognized the same warning signs: just minutes before the deadly wave hit, the water had begun to form bubbles and turn foamy, so very much as it did now. Tilly’s strong reaction to this phenomenon convinced the lifeguard to evacuate the beach. Minutes later the killer wave showed up. Unfortunately, on the many other beaches of the Indian Ocean, nobody recognized the danger signs that morning, an ignorance for which some 220,000 people paid with their lives.
Tilly Smith’s episode shows that some scientific conclusions related to natural disasters are easy to understand and apply. But reaching most of them, such as providing an estimate for the height of the water surge in New York, would be impossible without mathematical models. As an expert in celestial mechanics, I became interested in predicting catastrophes while computing the odds of potential collisions between Earth and asteroids or comets. Though such events are very rare, taking place every few million years, such an impact would decimate life on our planet and throw our civilization back to the stone age. Luckily we can now track rogue cosmic objects, determine their orbits and, thanks to international space programmes, destroy them or change their collision path. More earthly phenomena, however, such as pandemics, are still eluding us, but hopefully not for long. Several groups of mathematical biologists, for instance, are using now the theory of differential equations trying to understand how viruses mutate, in order to prevent another Spanish flu. Scientific prediction offers a striking example of the power of mathematical results, many of which appeared to have no applications at the time they were produced.
Florin Diacu, a professor of mathematics at the University of Victoria in Canada, is the author of Megadisasters: The Science of Predicting the Next Catastrophe.
This new Article is not yet ready for syndication. Please check back in a few minutes.
This Article is not available for syndication. Contact BestThinking for details.
Enjoy high quality content through BestThinking's syndication program. Learn more and register as a publisher today!
Enhance your publication, blog or journal with high quality content from BestThinking. Whether you are looking for a single feature article, a stream of dynamic content or just a few pieces each month, BestThinking's unique, customizable syndication feeds provide rights-verified material from identity verified Thinkers.
To syndicate a Blog or Article, you’ll need to start by setting up a feed. Creating a feed is a 3-step process:
Notice to Thinker Media Ebook Authors
European VAT collectors have changed their tax policy regarding ebooks. Previously being subject to VAT was based on seller's country, so none of our ebooks were taxable. As of 01 January 2015 VAT is based on buyer's country and will be deducted from our royalties. This is being handled automatically by Amazon, Apple, and Barnes & Noble. Details will be in your Royalty report.Close
About the Author
Florin Diacu is Professor of Mathematics and a former director of PIMS at the University of Victoria in Canada. He authored several books, i
About an hour’s drive from my native city in Romania is Copșa Mică—once the most polluted town in Europe. In the mid 1980s, when I often passed through that region by train, the place looked like hell on earth. Everything, from streets and houses to trees and hills, was black from smoke. Grass and...
Kenneth J. Arrow was in his late twenties when he had an insight that would change his life. It all started after the end of World War II while he was working on his doctorate in economics at Columbia University in New York and as part-time research associate at the University of Chicago. He was...
The Presidency Race Soon after the 2000 presidential elections had finally been settled, I heard the following exchange during a comedy show on TV: “What do you think about democracy in America?” “I think it would be a good idea.” Like millions of people, those comedians wondered about the...
The recent earthquake and tsunami in Japan, with all its consequences, raises awareness about natural megadisasters. Can we predict them and alleviate the loss of life and goods? This article presents some ideas about the state of the art of scientific prediction.