The 1,000-year event flooding that hit South Carolina means everyone there can forget about flood insurance until the year 3015, right? Not even close, according to scientists who say terms like "1,000-year event" are widely misunderstood and less accurate because of Earth's changing climate.
A "rain event like South Carolina doesn't necessarily ... happen every 1,000 years," tweeted meteorologist Marshall Shepherd of the University of Georgia, who's also a host for the Weather Channel and past president of the American Meteorological Society.
"(It) is apparent to me that many people still do not understand the concept of what 100- or 1,000-year rain event means," he also noted in his blog. "Many people literally assume it means this event 'can only' happen every 1,000 years."
But one extreme event does nothing to prevent another, any more than one flip of a coin influences the next. They are separate, independent events that answer only to the laws of probability.
"(T)he same amount of rain could conceivably occur the very next year, or might not occur until thousands of years later," explained Bob Henson and Jeff Masters of Weather Underground.
"Those ocean temps were amazing underneath Joaquín but the real trouble was brewing on Tuesday last," said Dan Satterfield, a weather blogger and Chief Meteorologist for the CBS affiliate WBOC TV in Salisbury, MD. "When I looked at the water vapor image that morning and saw three channels of moisture flowing up the east coast and from as far away as the tropical Pacific, it was clear that an historic event was likely. No doubt about it though there is a lot of science saying that the frequency of these events will increase."
To understand what "1,000-year event" means, it's better to think like a bookie. It's all about the odds. In this case the odds of a three-day rainfall amount of 17.1 inches for Charleston and 14.2 inches for Columbia are 1 in 1,000 years, explained Henson and Masters. Likewise, the 1-in-1,000-year rainfall amount over 24 hours for Charleston is calculated to be 14.8 inches and for Columbia, 12.5 inches.
The U.S. Geological Survey (USGS) uses these sorts of statistics to understand stream and river flows, which are naturally tied to precipitation. On their water resources website the agency explains that the interval of an event recurring is based on the probability that the event will be equaled or exceeded in any given year. This year there was a 1-in-1,000 chance of the rainfall and flooding seen in South Carolina. Next year there will be the same odds.
Coming up with these odds isn't easy, especially in places without a long history of meteorological or stream flow data. The USGS requires a minimum of 10 years of data to begin analyzing the odds of various streamflow events. Of course, the more years of data to draw from, the better.
At least that would seem to be the case. All the historical data in the world won't help when the climate is in flux.
"The concept of a 1,000-year event is one related to an unchanging climate," explained climate scientist Kevin Trenberth of the National Center for Atmospheric Research. "But it no longer works for a changing climate as the statistics are not 'stationary.'" That is, they're changing over time.
"In many places what used to be a 100-year event is now a 30- or 50-year event," Trenberth said. "The value varies in location. What used to be a 500-year event is now a 70-year event. This is readily seen when you have three 500-year events in 15 years, as occurred along the Mississippi River."
At best, the concept of a 1,000-year event is idealized, said Trenberth, and needs to be interpreted carefully.
"Bottom line yes, for many events anthropogenic global warming (AGW) has already affected our calculations, so that estimate of a X-year event has chances of being based on a record that includes changes due to AGW," added Claudia Tebaldi of the National Center for Atmospheric Research.
And yes, that means you still need flood insurance -- if you can get it.