Our telescope could give everyone their own galaxy to study

Interview with Steve Kahn, professor of physics at Stanford University in California and director of the Large Synoptic Survey Telescope (LSST) project. Now under construction in Chile, it is expected to be operational by 2019

Why do we need another huge telescope?
The wide-angle LSST will map the entire southern hemisphere every few nights. For the first time in human history, we will have detected more galaxies than there are people on Earth. Everyone could have their very own galaxies and stars to study. We’ll find 250,000 supernovae a year, track millions of asteroids every day. Among many other things, it will also teach us about dark matter, clumps of which should distort our images of distant galaxies.

What are the project’s biggest challenges?
LSST is 10 times bigger than anything anyone is even planning, in terms of how much light it will collect and its field of view. The only way to build it is with three mirror bounces to cancel optical distortions, but nobody has ever made a big telescope fold light in this way. Such a large-field view needs the largest camera ever built – the size of a small car. It will have over 3 billion pixels, made up of 189 16-megapixel units. The effort will create the biggest public data set in the world.

How much data will it generate?
Fifteen terabytes every night! Over a decade, the LSST will generate hundreds of petabytes of data, far more than everything ever written in history. Within 60 seconds we will get the data from the camera to a processing facility, detect everything in the image that has changed, and produce an automated alert stream. This will produce about 10 million alerts per night.

https://www.newscientist.com/article/mg22730363-400-our-telescope-could-give-everyone-their-own-galaxy-to-study/

So how will you cope with the data deluge?
The big challenge is not acquiring and storing the data but how you find anything in a database that big. We’ll have roughly 100 measurements (location, brightness, colour and so on) for each of 40 billion stars and galaxies, and we’ll look at every object around 1000 times over the 10 years of operation. You can think of that as a 100,000-dimensional research space, containing 40 billion objects. There’s a scale to the problem that’s frightening.

It’s a huge culture change for astronomy: most of the science will come from thinking of clever database queries. The best contributions probably won’t come from astronomers but from mathematicians, computer scientists and statisticians.


And non-scientists can get involved. What was the thinking behind that?
Some of the LSST’s donors, particularly in Silicon Valley, want to change the world, and on a personal scale. Science is the last bastion of having to go to a university and get a PhD, and there’s a lot of influential people in the Valley saying: screw that, why can’t smart, ordinary people play with cool things? You’ll be able to log on and view the southern sky at any time, or join citizen science projects to measure gravitational lensing.


This is so cool. Totally signing up for it when it's operational.