In August 2017, Hurricane Harvey made landfall on the coast of Texas.

It was not only one of the most powerful storms to ever hit the United States, it was also one of the most destructive: Days of historic rain, and devastating floods.

How do we comprehend a storm like this?

[INTRO] Hurricane Harvey is hard to put in perspective.

It forces us to make some pretty strange comparisons.

Harvey dropped an estimated 27 trillion gallons of rain, and that's been compared to a cube with 2.8 mile sides, a giant raindrop floating over Manhattan, four Great Salt Lakes, or 65 weeks of Niagara's Horseshoe Falls pouring over Texas.

Or roughly the equivalent to the volume of 40,882,455 olympic swimming pools.

I don't know about you, but I can't picture that many swimming pools.

None of that really means anything to me.

According to the hurricane scale, Harvey made landfall as a Category 4, and was later downgraded to a tropical storm.

But this scale is based on wind speeds, and doesn't really quantify risks like flooding.

The government designates certain areas as 100-year or 500-year or even 1000-year floodplains, but that's confusing too.

If you live in a 500-year floodplain, it doesn't mean floods happen there once every 500 years, it means a flood has a 1-in-500 chance of occurring *every * year.

That might sound like the same thing, but it's not.

These numbers are probabilities.

Houston's had three 500-year floods in the last decade, because that's how probability works.

If you've ever flipped a coin and gotten a string of tails, you know despite the 50/50 chance of getting heads each time, events sometimes cluster together.

But with a coin, at least we know one side is heads and the other is tails and that they're evenly weighted.

With weather events like storms, probabilities are trickier to determine in the first place, let alone keep them accurate as the world's climate keeps changing.

As cities like Houston and elsewhere grow and replace open land with roads and houses, they use whatever flood plan is in place when those things are built even though future changes in storms or city development can make those old models out of date.

That means what might have had a 0.2% chance of happening even 10 years ago may be more likely now.

There's also the puzzling effects climate change is having on big storms.

There's not statistically *more * hurricanes these days, but warmer oceans mean already-strong storms are getting stronger, dumping more rain, and costing more in damage.

So looking at one number, it might seem like nothing is changing, but looking at another... things are definitely changing.

Knowing all that though, the hurricane scale or the likelihood of a flood in your area or how global climate is affecting your local weather, it's still incredibly difficult to understand what a superstorm might bring, because people suck at understanding risk.

Especially long-term.

Folks who live in areas of frequent flooding tend to underestimate the risks they face.

They're more likely to think they can hunker down and wait out the rising water.

In general, we tend to rely too much on historical data.

What's really alarming is that we underestimate longer term risks like climate change the same way.

According to a survey done by Yale, people overwhelmingly agree that climate change is happening and will affect other people in the country.

But if you look at who thinks climate change will affect them personally, pretty much nobody.

Which doesn't make a lot of sense because every year, more and more people are experiencing the worst effects of a changing climate, millions of them.

Scientists are learning more and more about what the risks of climate change are, from storms to floods or a hundred other things.

If we learn something from Harvey, it's that we're still very bad at understanding those risks, and we have a lot of work and research to do, because we really can't wait until it affects us personally.

It's the only way we'll wrap our brains around what might happen next.

Stay curious