you use a computer.

But what if you couldn't?

What if it was, say, the 1950s

and you needed to work out if a bold but questionable plan

to dam the San Francisco Bay was a good idea?

The answer is this: The US Army Corps of Engineers Bay Model.

The Bay Model is one and a half acres or more.

What you're looking at is one of our former scientific, hydrodynamic,

engineering testing facilities.

And this was the tool, the instrument, that they used to see

what the unforeseen consequences of the John Reber Plan was going to be.

In the 1940s and 50s a man named John Reber

had a plan to completely change the San Francisco Bay.

Enormous dams would create freshwater lakes.

There'd be brand-new reclaimed land for industry and for air and naval bases.

Reber said it would make the bay a defensible military fortification,

move people safely inland

and the newly dammed rivers would provide huge amounts of drinking water.

Reber was not a professional engineer.

He was a theatrical producer who had done a lot of research,

but because he worked in showbiz, he knew how to promote something.

And by most accounts he was a friendly, sincere, convincing man.

So unlike other ideas for giant engineering projects, the Reber Plan actually caught on.

Debate went back and forth for years but eventually

Reber's plan seemed realistic enough, at least to politicians,

that the US Army Corps of Engineers were tasked to see if it was practical

and they were given $2.5m, that's about $25m today, to find out.

And with it, they built this.

This model when it was built was the next level, the next generation.

It was extremely accurate as an instrument.

At the time there really wasn't anything better.

There's 250,000 strategically placed little copper tabs in the bottom of the model

to keep the saltwater and the freshwater from going in and out too quickly,

but also to duplicate any little protrusions sticking up out of the bottom of the bay.

The model is 1:1000 scale horizontally,

1:100 scale vertically and 1:100 scale in time.

That means about two hundred times a day,

the tide comes in and the tide goes out,

because San Francisco Bay has tides and therefore, so does the model.

Everything was hand operated for the first thirty years.

It took probably anywhere between twelve and fifteen people to operate it,

sometimes it was as many as sixty people here at one time

working on various different experiments

and everybody had to be totally in sync.

It was like an orchestra.

And the interns were out there in chairs, in lab coats, in the water

measuring the various different ebb and flow of the tide.

It took three years for Reber's plan to be tested here

and in that time Reber passed away.

And for his supporters, the results from the model were devastating.

It looked great on paper, convinced a lot of people,

but when it was tested they found that it was only good on paper.

The end result was it failed on ninety-nine different accounts.

Wild and unpredictable catastrophic flooding was just one of them.

The dams wouldn't create lakes, they'd create evaporation ponds.

The tide would create dangerous currents and waves.

In short, not only would the Reber Plan have been a disaster,

it would have been a billion dollar disaster.

The ecosystem would have been devastated too,

but it was the 50s, so no-one was really thinking about that.

The Corps of Engineers, their job done,

figured the model would come in useful again someday.

And it did,

helping to test smaller and more practical schemes across the bay for decades.

These days, of course, computers can do all of that

for a fraction of the time and the money,

but it's still a good educational resource.

Last time I talked about a grand civil engineering scheme like this,

about Herman Sörgel's Atlantropa,

I said it was a testament to how big we can dream,

but this model is a testament to something else.

To science.

To having a hypothesis.

To testing it.

And then, when it fails, admitting that it's wrong.

There shouldn't be any shame in that.

Sometimes we follow bad ideas,

and changing your mind based on new evidence

and allowing others to do the same

is something our world should be built on,

and it's exactly what this model made happen.