Showing posts with label predictions. Show all posts
Showing posts with label predictions. Show all posts
Thursday, November 01, 2012
Thank a meteorologist
Where I live, or rather in several of the places where I have lived, meteorology often seems to have about as much predictive accuracy as astrology. The German weather website wetter.com predicts rain for Rostock all afternoon and evening, starting 2 hours from now, while windfinder.com predicts no precipitation at all today. It is currently raining.
Because this kind of obvious failure of retail meteorology is so common, bashing ‘the weatherman’ is a popular and easy past time, and I think this shapes the views of many people on meteorology generally. But when things like Hurricane Sandy do things like this to a densely populated portion of a continent and kill only dozens of people, rather than tens of thousands, that is because of the very healthy state of meteorological science.
Imagine, if you will, the impact of Sandy on the U.S. if meteorology hadn’t progressed in the last hundred years. We would have gotten reports of a hurricane hitting Jamaica, Hispaniola and then Cuba. Folks in New Orleans, hit by Katrina a few years earlier, with even less warning than in the real world, would start evacuating, along with much of Florida and the Gulf Coast. No telling where this thing might hit. New Jersey and New York would scarcely take notice. When Sandy clobbered the Bahamas, people in the Carolinas would need to evacuate fast. Who knows who the monster is coming for? Two days later when the surf in New Jersey started getting really rough, and the wind really strong, the idea that New Jersey and New York could get hit would occur to a lot of people. But what does it mean to get hit? Lots of rain and wind? Flash floods? Would anyone have guessed that there would be a nearly 14 foot storm surge in New York City, flooding houses, lobbies and subway tunnels with seawater? Would lower Manhattan have been evacuated at all? Would there have been time to get out? I doubt it.
My guess is the East Coast would have been far less prepared than New Orleans was for Katrina, because people in real world New Orleans, even if their government let them down, could turn on the TV and know what was coming. In a meteorology-free New York, the idea of large parts of the city being underwater would have seemed ridiculous until walls of water were ripping through tunnels still full of people.
We have come to take for granted that we should be warned, accurately, of major weather events days in advance, and we grumble about the uncertainties and inaccuracies. There is certainly still room for improvement. But the fact that New Jersey was named as one likely landfall while the center of the storm was not yet to Cuba is absolutely amazing. That requires a level of understanding and predictive computing that I find hard to fathom. It frankly seems a bit like magic. The National Hurricane Center nailed this one, and in so doing prevented tens of billions of dollars in damages and tens of thousands of lives. Somebody deserves a medal. Thank you meteorology.
Key Words
current events,
meteorology,
predictions,
weather
Monday, October 22, 2012
Idiozia
I love Italy. Friendly people, good food, great art, nice weather, etc. But there is something in Italian governance traditions that make me feel okay about not living there. Prime examples are Berlusconi and this, the conviction of six seismologists and a government functionary for failing to predict an earthquake in L'Aquila that killed 309 people in 2009. More exactly, they issued a report that said that it wasn't possible to predict whether or not their would be a major earthquake that year, but estimated the probability as fairly low. The earthquake came, and they are the scapegoats.
Imagine if you will that an Italian man goes to the doctor, and asks if he is going to have a stroke that year. The doctor gives him an exam, some blood tests and a questionnaire, then uses the best available methods to estimate that this man has only a 10% chance of having a heart attack that year. The man goes for a hike, has a heart attack, dies. Should the doctor go to jail for failing to know for sure that the heart attack is coming? Of course not. We don't put people in jail for failing to know things that no one could possibly know. We don't, but apparently the Italian courts do.
Now imagine the Italian government comes to you next week and wants you to produce an expert report on an unpredictable topic for them. Would you be eager to help?
Imagine if you will that an Italian man goes to the doctor, and asks if he is going to have a stroke that year. The doctor gives him an exam, some blood tests and a questionnaire, then uses the best available methods to estimate that this man has only a 10% chance of having a heart attack that year. The man goes for a hike, has a heart attack, dies. Should the doctor go to jail for failing to know for sure that the heart attack is coming? Of course not. We don't put people in jail for failing to know things that no one could possibly know. We don't, but apparently the Italian courts do.
Now imagine the Italian government comes to you next week and wants you to produce an expert report on an unpredictable topic for them. Would you be eager to help?
Subscribe to:
Posts (Atom)