Why do hurricanes curve?

With forecasts now available more than a week out, it can be disorienting to see a hurricane’s future track. How abruptly the storm changes direction in the coming days will determine whether it spins back into the Atlantic or hits New York, Florida, New Orleans, Puerto Rico, or Texas.

So why do storms often curve in the first place? The short answer: high and low pressure systems. “If you think of the hurricane as a ball in a pinball machine, high pressure systems are the bumpers and the low systems are the troughs where balls wants to roll towards,” says Neil Jacobs, Panasonic Avionics’ chief atmospheric scientist.

A semi-permanent high pressure system hangs over Bermuda. Known as the Bermuda High, it’s responsible for driving storms that form in the Atlantic westward. “That’s what happened with Irma,” says Jeff Masters, a meteorologist at Weather Underground. The storm caught the Bermuda High’s edge, which hangs above Key West, and swung north, eventually hitting the Florida Keys.

In addition to the Bermuda High, a hurricane’s path is shaped by other pressure systems, including a semi-permanent high pressure system that, while more transient than the Bermuda High, typically hovers over the southwestern U.S., and a low pressure system that can drift from the Middle Atlantic as far south as New Orleans and Florida, putting those regions at risk.

Why are hurricanes in the Atlantic so much more destructive than hurricanes in the Pacific?

The Pacific has its own semi-permanent high pressure system. Called the Pacific High, it’s located off the west coast of the U.S. and spins clockwise, pushing hurricanes that develop off the coast of Mexico further out into the ocean. (Occasionally, however, storms curve far enough North to impact Hawaii.)

Temperature also plays a role: Hurricanes “derive their energy from heat,” says Jacobs. In the Atlantic Ocean, the gulf stream can push water temperatures above 80 degrees. The Pacific is considerably cooler. Even at the height of summer, surface temperatures rarely crack the mid-70s, which means storms in the Pacific “die pretty quickly once they travel northwest.”

How accurate are hurricane forecasts?

Forecasts are twice as accurate today as they were 20 years ago, says Masters. The shorter the timeline the more accurate the reading. For a five, four, three, and two day forecasts, the average error is 225 miles, 175 miles, 120 miles, and 70 miles, respectively.

Today’s statistical models are based on longer term data sets, which provide more accurate forecasts. (Before the 1970s and the advent of reliable satellite imagery, hurricane data for storms that didn’t make landfall was based on ship records.) Computing power has also advanced, and today’s models can incorporate new data points generated by a storm in close to real time. “People don’t remember how bad forecasts used to be,” says Jacobs. When Hurricane Hugo pummeled the Southeast U.S. in 1989, its path remained unclear mere days before it made landfall.

Today, accurate forecasts extend far further into the future. Meteorologists, for example, predicted that Hurricane Sandy would hit New York City more than a week in advance.

Marshall Shepherd, director of the atmospheric sciences program at the University of Georgia, believes that forecast inaccuracies are often overstated. A hurricane’s projected track is displayed as a protruding forecast cone, which accounts for statistical uncertainties and forecast errors. “If it’s in the cone, it’s fair game,” says Shepard, although many people believe that a forecast is only accurate if center of the storm perfectly traces the line at the center of the cone.

“A storm can be 300 miles across,” says Jacobs. If you get too caught up in whether an area aligns with the center of the storm, you miss “who is going to experience hurricane force winds.”

Meanwhile, predicting a hurricane’s intensity is a more inexact science, says Masters. There are more, hard-to-read variables to account for. An emerging storm deep out at sea “is an almost impossible environment to observe.”

What’s the difference between a Category 4 and a Category 5 storm?

Hurricanes are classified on a five-point system known as the Saffir-Simpson scale. First introduced in 1969, the scale’s ratings are based on a hurricane’s sustained wind speed: Category 1 (74-95 mph), Category 2 (96-110 mph), Category 3 (111-129 mph), Category 4 (130-156 mph), and Category 5 (157 mph or higher).

While a hurricane’s rating is connected to the damage it can cause, the relationship is imperfect. The rating system is a good measurement of the impact of wind — anything above a Category three has the power to significantly damage well-built framed houses — but does not account for storm surges or rainfall-induced floods. As we saw with Hurricane Harvey, these factors can play a major role in a storm’s deadliness and destructiveness.

How much worse will this get?

From Harvey, to Irma, to Maria, powerful, damaging storms keep coming.

If you feel like we’ve entered a new age where deadly storms are the new normal, you aren’t alone. Which begs the question: How monstrous will these hurricanes grow?

The short answer is, we don’t really know, says Jacobs. There’s not enough data for meteorologists to draw any firm conclusions (remember, they’ve only tracked storms via satellite for less than half a century, whereas hurricanes have bashed into the Atlantic coast for hundreds of millennium).

However, when scientists factor in projection models for climate change, the outlook gets bleaker. Hurricanes are driven by heat, which means as water temperatures rise, storms will have “more energy available to them,” says Masters. Computer models show that this projected warming will increase storm severity by 2% to 11%, with rainfall increases of up to 20%.

“A 12 mile per hour increase in wind speed can result in three to four times the damage,” says Masters. In other words, a storm that would cause $1 billion in damages today would cause $3 billion in damages by the end of the century. And that doesn’t take into account rising sea levels, which push storm surges further inland.

Bottom line: predictions are imperfect and uncertain. But no matter how shaky the forecast, putting our collective heads probably isn’t the answer. “We have to more resilient in how we design cities and build infrastructure,” says Masters.

“You can either chose to believe the climate models or not,” says Jacobs. “But if we sit around for 50 years and wait to see if they’re verified, it might be too late.”