They know they missed it. Forecasters at the National Hurricane Center in Miami say when it comes to the strength of Hurricane Irene as it approached North Carolina, they know they were off. Way off.
“At least in the guidance we were looking at there was no indication of anything that would cause the storm to weaken. So, we thought we would have a Category 3 at landfall,” said Bill Read, the director of the Hurricane Center. Irene came in at a Category 1, the weakest. Read said there’s good reason they were so far off.
The science of forecasting how strong or weak a storm will become is simply not very good. With Irene, forecasters say they weren’t even as good as their five-year average.
“Every storm comes up with a surprise,” Read said. “In this case it was one where it went downhill. Charlie a few years ago is one that went uphill. Neither case did we see that coming, and that’s my measure of the fact that we have a long way to go.”
Hurricane forecasters say they want to get it right all the time. But if you are going to be wrong, they say it's better to be wrong in weakening storms like Irene.
“I’d say a bigger worry than one weakening at landfall is the ’35 hurricane that came through the Keys," Read said. "Charlie if it’s a little bigger. Audrey in 1957. Get the picture?”
In all of these cases, the storms rapidly intensified as they neared the coastline. By then, it’s too late to order massive evacuations.
CNN's severe weather expert Chad Meyers said when Hurricane Irene smashed into the Outer Banks of North Carolina, the contact weakened the storm.
So, forecasters had the path right, but the impact of landfall changed what the amount of destruction would be in some areas. Wind shear helped knock down velocity, and unexpected dry air sucked some of the power out of the storm.
"It literally knocked the stuffing out of the eye," Myers said. "It never got its mojo back."
Meteorologists measured pressure levels inside the storm that could have allowed it to strengthen back into a Category 3 hurricane, Myers said, but Irene's romp over land in North Carolina prevented the eye wall from spinning into a more destructive storm by the time it arrived in New York.
"It never had that opportunity because North Carolina got in the way, dry air came across over Virginia and Maryland and got in the way, and although this was very low pressure, the reason why we could never let the guard down for New York City ... was because the pressure was low enough that at any time, if this storm decided to get its act together, it could have gone from a 60-70-80 miles per hour storm - it easily could have been a 110 (miles per hour) storm like it was in the Caribbean and like it was in the Bahamas."
Add to that the difficulties of having true accurate model data when it comes to hurricane forecasts.
“Real-time observations, like that collected by NOAA’s Hurricane Hunters, are put into these models to hopefully give a more accurate forecast. We are much more accurate on forecasting severe storms and their behavior for this very reason, the availability of observations," CNN meteorologist and weather producer Sarah Dillingham said. "These storms occur over land, allowing scientists to take measurements within those storms and use that data to run computer models. Think about a hurricane, over water, with no way to collect data except from above.
"Makes it difficult to see what is actually going on, right?"
Dillingham said dropsondes, instruments that are dropped from above and into these storms, are used to collect data inside hurricanes as they fall to the surface.
"This is helpful, but you are also trying to view the data it collected, track where it was in the storm, and at the same time determine what that means from a scientific perspective," she said. "Also, you may think, what about trying to send something up into the storm from below. In a hurricane? Good luck with that."
Dillingham said "the sparsity of real-time observations in tropical systems is what makes it so difficult to produce a more correct intensity forecast, and certainly makes it difficult to improve them."
"The track of these storms depends on atmospheric winds and surrounding storm systems, and we understand these factors much better, giving us a better handle on the 'steering' of these systems," she said. "This makes tracking more accurate overall. Things like RI, or rapid intensification, in tropical cyclones is just not fully understood yet, and until we can obtain that vital observational data within these storms - while they are over open, warm waters - we will struggle to model these kinds of processes.”
For that reason, Dillingham said Read's explanation of why they have trouble making these predictions is spot on.
And that's also why Read says the decisions to evacuate made by emergency managers and state and federal officials was the right one.
In every aspect of the storm except for wind speed, Read says, they got it right.
According to their initial analysis, the track forecast of the storm was 20% better than their five-year average. They do very well at predicting the path of a hurricane.
“We had storm surge flooding all the way up from the Carolinas into New England," Read said. We’ve had tremendous and tragic rainfall flooding. We’ve had loss of life from trees down well inland and the power outages.”
Hurricane forecasters admit their ability to foresee a storm’s strength is not much more today than it was 20 or 30 years ago.
According to the Hurricane Center’s initial analysis, “Irene exemplifies the state of the science.” They are pinning their future hopes on programs like the Hurricane Forecast Improvement Project. Scientists say they are seeing some promising results. This program uses high-resolution models and enhanced Doppler radars to measure the core of Hurricanes.
Forecasters say that three out of four times you will likely be asked to evacuate and you’ll coming back saying "Why did I leave?"
But that fourth time, if you don’t, Read says, you’ll wish you had.