Where's the storm going? How strong will it be? What's causing it to intensify?
These are all questions that mystify weather experts because of the inexact science of hurricane forecasting. But, that's slowly changing.
"Over the past 10 years, we've improved our track predictions on average by about 40%. Our cone that we have is about 40% smaller now than it was at the time of Katrina," according to James Franklin with the National Hurricane Center.
Typically, forecasters are more accurate at predicting the track of a storm than its intensity, but the opposite has been true this week with Hurricane Joaquin
, CNN senior meteorologist Dave Hennen said. Early in the week, he said, no model had the storm hitting the Bahamas, but Joaquin's eye passed over the archipelago's Samana Cays on Thursday.
On Wednesday, U.S. models were predicting landfall in the Carolinas, while the European models had the storm hitting Bermuda -- a difference of roughly 700 miles, Hennen noted. On Thursday, however, the U.S. models had Joaquin hitting the Canadian Maritimes, composed of New Brunswick, Nova Scotia and Prince Edward Island, some 1,300 miles northeast of the Carolinas. By Friday, most models suggested the storm would not make landfall in the United States and would instead churn out to sea - though it could cause severe flooding in South Carolina and North Carolina.
So why is it so challenging to forecast a hurricane's intensity and track?
"It's so complex. There are so many factors, it's impossible to model," Hennen said, explaining that "millions of factors" go into these predictions.
And when it comes to predicting the strength of a storm, Frankin said, "Intensity prediction has been pretty much flat."
That's where the hurricane lab at University of Miami's Rosenstiel School of Marine and Atmospheric Science comes in. Dubbed SUSTAIN
(Surge Structure Atmosphere Interaction Facility), the lab allows scientists to use an enormous water tank to recreate massive storms -- some as strong as a Category 5 hurricane, with wind speeds of 157 miles per hour.
The tank sits in a room the size of a small warehouse. Researchers use huge paddles to make waves that are then intensified by powerful wind turbines. A shoreline and scaled-down home sit on the other end of those waves to help scientists figure out the real-life impact on structures and the environment.
"This is the first time that we've been able to get to these really intense winds of Category 5 equivalent in the laboratory and study how they interact with the ocean surface," Brian Haus, professor of ocean sciences at the University of Miami, said. "We want to know how they get from less intense to more intense and all different steps along the way."
The steps are part of the challenge because of the many factors that go into predicting a hurricane's strength.
"Track is easy because hurricanes are steered by large features in the atmosphere that are easy to measure and easy to model. The hurricane intensity depends on things going on much smaller scales. Waves, the organizations of individual clouds and thunderstorms, complexes to rain bands -- everything is happening on a smaller scale, and that's part of what makes it difficult," says Franklin.
Leaps and bounds
The technology used to measure such things has improved immensely in recent decades, but there's still plenty of room for advancement, Hennen said.
Forecasters have been flying planes into storms since the 1940s, but today, not only do you have "core punchers" that can fly into the eye of a storm, but the National Oceanic and Atmospheric Administration also employs modified G-4 Gulfstreams that fly over and around storms for up to 12 hours at a time, dropping instruments in hopes of narrowing down a storm's path and predicting its intensity.
With Joaquin, as with any storm that threatens the eastern seaboard, forecasters are also launching balloons along the East Coast to measure the upper atmosphere.
All this takes a beast of a computer, and Hennen said some of the fastest computers in the world are used to crunch the available data. The government has recently earmarked significant funding -- most notably via the Weather Forecasting Improvement Act of 2013 -- to make sure NOAA has the computing and modeling capabilities necessary to make the most effective predictions.
As an example of how the technology has evolved, about 30 years ago, forecasters used computers that measured two levels of atmosphere, Hennen said. The advanced computers measured four levels, he said. Now, you could run those programs with the computing power of today's mobile phones, and modern programs measure dozens of layers of atmosphere, Hennen said.
If forecasters are able to more accurately figure out a storm's intensity -- and more importantly, when it will intensify -- Franklin says it could save countless lives.
He points out examples such as Hurricane Charley in 2004, which rapidly strengthened just before making landfall along Florida's southwest coast. The storm ended up killing 31 people in the United States and cost $15 billion in damages across the country, according to the U.S. government.
"If we could get a better forecast that helps in evacuation planning -- particularly in large cities. The sooner you can get good information on how strong a storm is going to be, the more safely you can plan to evacuate people that are at risk," Haus said.
It's important to keep in mind, Hennen said, that in some of the most at-risk locales -- such as the Florida Keys -- it could take two or three days to fully evacuate the endangered area.
Adding to the challenge for scientists is climate change. Because hurricanes are powered by the heat in the ocean, researchers concede it's growing more difficult to predict where or when storms will pop up and strengthen.
Haus points to another issue.
"I think what is clear for coastal areas is that sea levels are rising. There's no way we're going to get around that," he said. "So the risk that coastal communities are facing is just going to be getting higher and higher, particularly in low-lying areas -- Miami, Tampa, Houston, Washington D.C., the Northeast. These areas, like you saw with Sandy, are very vulnerable."