Live from TED

Deep and faint in the night sky: Andrew Connolly at TED2014

Posted by:
Andrew Connolly. Photo: James Duncan Davidson

Andrew Connolly. Photo: James Duncan Davidson

Astronomer Andrew Connolly begins by telling us that in 1781, English composer, technologist and composer  Sir William Herschel noticed something unusual, a little bit of data that was wrong. This was the discovery of a new planet, Uranus. (“A name that has entertained countelss generations of children.”) Just last week NASA announced 517 new planets, almost doubled the number of planets known. We’re in the beginning of an era where astronomy is defined by massive amounts of data.

As we move into this era of big data, Connolly says, we’re going to start seeing the “difference between more data being just better and more data being different.” That is, it’s not about how much data we collect, but about this: “Does that new data open new windows into our universe?” And if so, what’s the next window?

Connolly thinks time is a fantastic possible new window. Time is about the origins and evolution of our solar system and our universe. It is the key to questions like: “Is our solar system unusual? Why is our universe continuing to expand? What is this mysterious dark energy?”

New technology will enable a view of this changing universe. We have beautiful views of the night sky, but they are static, and that affects the way we view the universe — we see it as eternal and unchanging. But in reality the universe is constantly changing. The universe, Connolly says, “is constantly changing on a timescale of seconds to billions of years. Galaxies collide, stars are born, they die, they explode in these extravagant displays.” In that seemingly tranquil sky, there are actually many supernova. In fact, there are 10 supernova per second explode somewhere in our universe. The nearby sky is also moving, swarming with asteroids orbiting the sun. “The motion,” says Connolly, “is what allows us to build models of the universe to predict it’s future and explain it’s past.”

To see why this is hard, consider taking a picture of the whole sky with the Hubble Space Telescope. It would take 13 million pictures to do it just once. So they need to build new telescopes that can go wide as well as deep. Because of that a large group of astronomers is building the LSST, the Large Synoptic Survey Telescope. (“Proof you should never allow a scientist or an engineer to name anything, including your children.”)

The new telescope will take high-resolution images that cover an area 7 times the width of the full moon. To do that they build a phone camera the size of a VW Beetle, containing 3.2 billion pixels. To view one image at full resolution would require 1500 HD TV screens The telescope will take an image every 20 seconds, constantly scanning the sky. Because of that, we’ll have a complete view of the night sky over Chile every 3 nights. By it’s end, “for first time, we’ll have detected more objects in universe than people on Earth.” And doing that requires an absolutely massive amount of data analysis. Doing that requires sophisticated algorithms, which are as important to the science as the hardware.

Connolly says that thousands of discoveries will come from this project. Here’s two.

We can find planets around nearby stars — in the last few years we’ve discovered thousands, but they are all unlike our solar system. And that raises the question, “Are we not looking hard enough, or is there something special about ours?” To understand that we need to understand the history of our system. That’s where the asteroids come in — “They’re the debris of our solar system’s formation.” As the planets were forming they scattered the asteroids in their wake. So, “Studying the asteroids is like forensics for teh solar system.” It’s also helpful for predicting the future — in particular for predicting, hopefully averting, asteroid impacts. The LSST will be able to see asteroids in more detail, and at further distances, than current telescopes.

Distance also translates to time, Connolly said: “Every foot of distance away is a billionth of a second back in time.” This notion of looking back has revolutionized our view multiple times. In 1929 Edwin Hubble noticed that the universe was expanding. “The observations were simple, just 24 galaxies and a hand-drawn picture.” That was enough to start the birth of modern cosmology.

Seventy years later two groups of astronomers showed that the universe wasn’t just expanding, it was accelerating. They did this using the brightness of supernova. It took three years to find the 42 supernova used in the studies, but that changed our view of the universe. Connolly shows the data, a very unassuming data plot with a few straight lines drawn on it. He says, “This may not look impressive. This is what a revolution in physics looks like. A line predicting the brightness of supernova, and a handful of lines that don’t quite fit that line. Small changes turn our understanding of the universe on its head.”

What is this dark energy? No one knows. There are many different theories, and each produces a slightly different prediction — all of them consistent with our 42 points. That’s the second area where the LSST will shine. They will find 10 times more supernova in the first night than were used in the discovery of dark energy, and 1.45 million supernovae by the end of its survey. Astronomers and cosmologists will be able to progressively rule out different theories. “This,” says Connolly, “is when data becomes more than just better — more data becomes different.”

But these are the simplest questions. As Connolly points out: “We may not know the answers, but we know how to ask the questions…. How many more times are we going to find 42 points that don’t quite match what we expect?”