Exposing PseudoAstronomy

June 12, 2015

Are We on the Verge of Discovering an Earth-Like Exoplanet?


I announced awhile ago that I was on episode 347 of the Canadian, “The Reality Check” poscast where I talked about exoplanets and some hype — deserved or otherwise — about almost but never quite yet discovering Earth-sized exoplanets.

While they post a lot of links and other things on their website, they don’t post transcripts of what we actually talk about. Since I spent a solid many minutes writing and editing my segment’s text, I thought I’d post it here:

There’s lots of ways to talk about exoplanets, but I’m going to take the traditional approach and start with a very broad but brief overview of how we have found the few-thousand known extra-solar planets, or “exoplanets” for short. There are five main ways.

The most obvious is the most difficult: Direct Imaging. This is where you take your telescope and would look at a star and see the planet around it. This is almost impossible with current technology, and we have less than 20 exoplanets found this way. It’s so hard because the star is so bright relative to the planet and because most star systems are so far away. And obviously, if the planet is larger and farther away from the star, it’ll be easier to see.

The second main method has also only produced about 20 planets so far: Gravitational Microlensing. Einstein showed that large masses bend light, and we can see this in space when an object that’s far away passes behind a massive object that’s a lot closer. The light from the background object gets distorted and magnified, much like a lens … a lens caused by gravity. If the foreground object happens to be a star, and that star has a planet, then that planet can make a detectable contribution to the lensing, not only in amount, but in the exact shape of the lensing effect.

The earliest actual successful method was a special form of what’s called the Timing Method, specifically in this case, pulsar timing. Pulsars are incredibly dense stars called neutron stars, and we get a blast of radio waves every time one of its poles sweeps in the direction of Earth. These are so regular that any tiny perturbation can be detected and attributed to something weird, like a tiny planet tugging on it and so changing that regular spinning signal.

This is the same concept as the highly successful method that found the most exoplanets until a few years ago: Radial Velocity. The idea is that we normally think of a planet, like Earth, orbiting the sun. But it doesn’t really. It *and* the sun orbit a mutual gravitational point called the “barycenter” that is between the two. For Earth and the sun, that point is VERY close to the sun’s center, but it’s not quite in the center. That means that over the course of a year, as Earth goes around that point, the sun will, too (on the opposite side of that point). So, it will wobble very very slightly as it orbits the barycenter.

We can’t possibly observe this tiny tiny motion of other stars. BUT, we can use the light that star emits to do it by using the Doppler shift. That’s the phenomenon where if something is moving towards you, the waves it emits become compressed, and if it’s moving away from you, the waves get stretched out. The common example is a train whistle going from high to low pitch, but in astronomy, this is where the light is shifted to blue and then to red.

So, if the planet around another star is at its closest point to us, the star emits light and we see it all normal. As the planet starts to move away from us, the star starts to move very slightly toward Earth, and so its light will be very slightly blue-shifted. Then, the planet gets to its farthest point, and starts to move towards Earth, which means the star starts to move away, and we see its light red-shifted. This is an incredibly tiny effect, and the smaller the planet, the smaller the shift in the light. Or the pulsar timing change.

There was a lot of progress throughout the late 1990s and early 2000s in very high-resolution spectroscopy in order to get better and better at observing smaller and smaller planets. The easiest ones to observe are the largest because they make the biggest shift in the star’s light, and ones that are closest to their star are easier because you don’t have to observe as long. To observe a planet that has a 10-day orbit, you just have to observe that star for about a month from Earth to get decent statistics.

That’s why all the exoplanets discovered early on were what are called “Hot Jupiters,” since they were very large and very close to their stars.

The final method is the Transit Method. If a fly passes in front of a bright light, you can see a slight decrease in the light. If a bird passes in front of a light, you’ll see a larger decrease in the light. Same thing here: A planet passes in front of the star and temporarily blocks part of the light from the star that we would see at Earth. The big issue with this method is that you have to have the fortuitous geometry alignment where the planet’s orbit is just right so that it passes in front of its star as seen from Earth. The first one wasn’t detected until 1999, but a decade later, the dedicated spacecraft COROT and then Kepler were launched to look for these, monitoring the same fields of the sky, tens of thousands of stars, moment after moment, looking for those brief transits. In 2014, Kepler released over 800 planets discovered with this method, more than doubling the total number known, and that was on top of its other releases and, to-date, it’s found over 1000.

The transit method, despite the issue of geometry, is probably the best initial method. If you have the planet going in front of its star, then you know its alignment and you can follow-up with the radial velocity method and get the mass. Otherwise, the radial velocity method can only give you a minimum mass because you don’t know how the system is oriented, you only know that radial component of velocity, hence its name.

With the transit method, you can see how much light is blocked by the planet. Knowing the star’s type, you can get a pretty good estimate for the star’s size, and knowing how much light is blocked means you can get the cross-sectional area of the planet and hence its diameter. For example, Jupiter would block 1% of the sun’s light, and since area is the square of length, that means Jupiter is about 10% the sun’s diameter. Since the sun is a type G V star, we have a good model for its radius, though of course we know its radius very well because we’re in orbit of it. But that means not only can we get mass, but we can get size and density.

The transit method also lets us see if there’s a large atmosphere. If the light from the star instantly blinks down to the level when the planet passes in front of it, then any atmosphere really thin or nonexistent. If there’s a gradual decrease, then it’s extended. If its extended, we can follow-up with something like the Hubble Space Telescope and actually figure out what that atmosphere is made of by looking at what colors of light from the star are absorbed as it passes through the planet’s atmosphere.

And as with the radial velocity and timing methods, we know how long it takes to go around its parent star, and along with the star’s mass from what kind of star it is, we can get the distance of the planet from the star.

Okay, so much for a brief overview. But for me, I’ve left out a lot.

Moving on, it should be somewhat apparent that the bigger the planet, and the closer to its star, the easier it is to observe with pretty much ANY of these techniques, except direct imaging or microlensing where you want a big planet that’s far from its star. Big means big effect. Fast orbit means you don’t have to observe it for very long to show that it’s a regular, repeating signal best explained by a planet.

So, the question is then, can we detect an Earth-sized planet, and can we detect an Earth-like orbit? These are really two different questions and they depend on the technique you’re using. If we want to focus on a the two main methods – radial velocity and transit – then the unsatisfying answer to the second is that we do finally have good enough technology, it is just a matter of finding it. With the 2014 Kepler data release, there were over 100 exoplanets that are less than 1.25 Earth’s size. With the 2015 release, there are a total of 5 planets smaller than Earth or Venus, but they orbit their 11.2-billion-year-old star in just 3.6 to 9.7 days.

Even if we have observations for more than a year or two, for something as small as Earth, the level of signal relative to noise in the experiment is still pretty small, and you want a big signal relative to the noise. It’s best to build up multiple years’ worth of data to average out the noise to be able to really say that we have an Earth-like planet. For something like Jupiter, which orbits our sun in about 12 years, we’d need to observe at least two transits, meaning we’re just now approaching the time when we would have a long enough baseline of data with some ground-based surveys, but that’s also assuming we catch that planet for the few hours or days when it goes in front of its star versus the years and years that it doesn’t, and that we do this repeatedly and don’t chalk it up to sunspots.

This is why we really need long-term, dedicated surveys to just stare at the same place in space, constantly, measuring the light output of these stars to see if we can detect any sort of dimming, that’s repeated, from a likely planet.

But, even if we find an Earth-like planet in terms of mass and diameter and location in its solar system, that’s not enough to say it’s Earth-like in terms of atmosphere and surface gravity and overall long-term habitability. It’s just a first step. A first step we have yet to truly reach, but one that is reasonably within our grasp at this point.

But it’s from the existing planets we know of that we get some of the hype that hits the headlines every few months, like “astronomers estimate billions of Earth-like planets exist in our galaxy alone.” I’m not going to say that’s fantasy, but it’s loosely informed speculation based on extrapolating from a few thousand examples we now have from a very, VERY young field of astronomy.

Or, we’ll get articles where the first sentence says, “Astronomers have discovered two new alien worlds a bit larger than Earth circling a nearby star.” It’s in the next paragraph that we learn that “a bit larger than Earth” means 6.4 and 7.9 times our mass, and they orbit their star in just a few days.

So as always, this is a case where, when we see headlines, we need to be skeptical, NOT susceptible to the hype, and read deeper. But that said, it is entirely possible that any day now we will find an exoplanet that is at least like Earth in mass, size, and distance from its host star.

2 Comments »

  1. Would it be possible to detect a planet if we were looking “down” on the system? The star would still wobble but from side to side from our viewpoint. Would this be possible at present or are we going to have to wait for some new technique or technology?

    Comment by nwsingledad — July 10, 2015 @ 11:04 am | Reply

    • Astrometry is another method of detecting a planet around a star, but it is incredibly difficult. So far, no exoplanets have been detected by astrometry that I can find.

      Comment by Stuart Robbins — July 15, 2015 @ 6:05 pm | Reply


RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Create a free website or blog at WordPress.com.

%d bloggers like this: