Predicting Hurricanes in High Definition

[Narrator] Ten years ago, Katrina began life

as a little tropical depression over the bahamas.

Over the next few days, it got stronger,

then weaker, then stronger again, then weaker again.

When it finally made landfall in New Orleans,

it was a Category 3 hurricane.

Hurricanes are fickle like this.

But a decade later, scientists

have gotten a whole lot better at modeling hurricanes.

Here's how Katrina would have looked

in weather models from 2005.

And this is what a hurricane looks like today.

It's no longer an 8-bit video game hurricane.

So how did we get from there to here?

First, more data.

NASA has formed airplanes right into

the middle of a hurricane to drop sensors.

These sensors collect data on temperature,

rainfall, dust, winds and more.

And NASA gets information from space

using both its own and other countries' satellites.

Take Japan's new powerhouse Himawari 8 weather satellite,

which captured a recent typhoon in amazing detail.

Scientists also just understand hurricanes better.

A classic sign of a bad hurricane are hot towers.

Hot towers are these tall clouds

that carry heat high up into the air

that may indicate a strengthening storm.

Not all hot towers are created equal, however.

Scientists now know that hot towers concentrated

in a specific region behind

a storm's strongest winds are the worst.

All that data is useless if you don't know what

to do with it.

NASA now has powerful supercomputers

that crunch these climate models

at a resolution 10 times better than they did 10 years ago.

This is good news because scientists predict

that because of climate change, we can expect bigger,

and more violent storms in the decades to come.

(humming and vibration background music)