Infrared image of Beta Pictoris b from GPI. Photo processing by Christian Marois, NRC Canada.

We’ve had the capability to find planets by taking pictures of them directly for a while now, usually in the infrared, but sometimes in visible light. But now we have the capability to do a whole lot more of it. GPI (pronounced GEE-pie) is the Gemini Planet Imager, the latest tool in the hunt for extrasolar planets, and it just came online this month.

GPI uses a telescope with a mirror the size of a studio apartment and a camera the size of a compact car to take pictures of planets not unlike Jupiter. GPI is expected to find dozens of these planets, at least, and better yet, take good enough pictures to study the infrared light they emit and find out what their atmospheres are made of.

So is GPI any good for finding Earth-like planets? Sadly not. We’ve got a long way to go to get there.

The first problem is that Earth is old. It’s five billion years old, which means it’s cooled down too much to give off much infrared light. To find old planets, we have to rely on reflected starlight in the visible range, and stars are much brighter and produce much more glare in visible light than in infrared.

The second problem is that Earth is small. That means it doesn’t reflect much light. A Jupiter-like planet in the same orbit would reflect over 100 times as much sunlight as Earth, making it that much easier to see.

The third problem is that Earth orbits close to the Sun. GPI is designed to see planets 5 times as far as Earth is from the Sun, or more. The glare from the star gets exponentially worse the closer in you go, so finding an Earth-like planet is that much harder.

Put all these things together, and you have what sounds like an impossible task. In fact, the task of spotting an Earth-like planet around a nearby sun-like star is roughly equivalent to–and I am not making this up–to spotting an ordinary flood lamp, just on the other side of town from an atomic bomb…from Mars.

And yet, people are working on it. There are fancy ways of using interferometry to combine two picture of a star so that the starlight all cancels out, and you only see the planet. Of course, you have to put some rather large telescopes in space to do it, and all the potential NASA missions for the job have been scrapped. But who knows, maybe one of them will be resurrected. It hasn’t stopped the proponents from trying. So maybe in twenty years or so, we’ll have a real life photograph of another Pale Blue Dot.