Солнечная система и ее тайны

Планеты Созвездия НЛО

Challenges of Planet Detection

In this section, we've discussed some of the challenges associated with detecting life outside our planet: the inverse square law and the absorption of light. Despite these factors, we have been able to detect planets - but why have we not yet detected a planet the size and mass of Earth, in a one-year orbit around a star like the Sun? Why have we yet to study the atmosphere of a planet Earth's size? The reason is that the methods of finding and characterizing planets that we've discussed throughout this course all have their limitations, and those limitations do not favor the detection of an Earth-like planet. In this section, we'll discuss the limitations of each method.

Scientists often use the concept of "signal-to-noise ratio" when discussing the significance of a detection. The signal is the strength of your detected event; the noise is everything else that dilutes your signal. The signal-to-noise ratio is a quantitative way of describing how much stronger your detection is that the background level of noise. A higher signal-to-noise ratio means it is more likely that your detection is a real event - and not just noise.


We've already talked in depth about how the probability of a transit affects the likelihood of whether we'll see a planet transit from our vantage point. The transit probability disfavors the detection of planets in long period orbits, because the transit probability is lower. For Earth, the probability of transit is 0.47%. For Jupiter, which orbits 5 times farther out, the probability is 0.0009%.

The depth of a planet's transit is one factor that makes Earth-sized planets more difficult to detect. The fraction of light blocked by a planet during transit is equal to the ratio of the planet's cross-sectional area to the stars'. That means that, around any given star, larger planets will have deeper transits than smaller planets do. For example, the transit of Jupiter across the Sun would have a depth of 1%, while the depth of Earth's transit would be just 0.0084%. Our telescopes are hindered in the detection of small signals by random noise in the number of photons received and by the way in which data is collected. 1% is well-within reach of good amateur astronomy equipment, but 0.0084% pushes the limits of Kepler.

Kepler never expected to identify an Earth-like planet with just one transit event: the signal-to-noise of the transit was expected to be quite small. But, by adding together many transits, one can increase the signal-to-noise of a planet detection. This technique is very important for finding the transits of small planets. But a planet with Earth's orbit takes an entire year to orbit its star: you have to wait one extra year for every additional transit you need to see.

Another source of noise cannot be beaten by time: the variability of the star itself. The brightnesses of even the most boring of stars fluctuates on many timescales. Spots on the surface of the star cause the brightness to slowly vary as the star rotates, moving the spot in and out of view. Flares produce rapid increases in brightness. The fluid motion in the star's atmosphere also results in small-scale, rapid variability. These brightness variations complicate the detection of transits - particularly of small transits - and in some cases may be preventing their detection.


The wobble (radial velocity) method is subject to similar limitations as the transit method. Unlike the transit method, planetary systems don't need to have that narrow range of inclinations that allows them to cross the surface of their star as viewed from Earth. However, the amplitude of the wobble signal that we see for a star does depend on how its planetary system is inclined, and it's easier to detect systems that are edge-on.

As in transits, the strength of the wobble signal depends on the properties of the planet. More massive planets tug more forcefully on their star, producing larger signals. Planets orbiting closer to their star are more rapidly moving, so they also have a larger influence on the star's radial velocity. Again similar to the transit method, signal can be built up by observing the planet for multiple passes around its star.

The last source of noise - stellar variability - is also a serious problem for the radial velocity method. The brightness variations discussed previously also tend to produce radial velocity variations. These must be removed - a sometimes difficult prospect - if one hopes to find the wobble induced by a small planet.


Of all the methods, microlensing is the least biased against small, distant planets. Microlensing is most sensitive to planets with orbital periods falling between about 1 and 10 Astronomical Units (between the orbits of Earth and Saturn), and the strength of the signal can be large even for low-mass planets. In fact, a Super Earth with an orbital period of about 10 years was detected around a distant star in 2006. The main drawbacks of microlensing are that an event is extremely rare, requiring the chance alignment between a foreground planetary system and a background star, and that it is extremely difficult to learn more about the planet once you've found it.

Direct Detection

Once you block the light of the star, what kinds of planets can you see? Planets orbiting too close to the star will be lost in the glare of their star's light, even in the best of circumstances. For this reason, the planets that have been directly imaged are distant from their stars and typically orbit them with periods of hundreds of years.

Brighter planets are also be easier to see with direct detection. There are two ways for a planet to shine: it can reflect the light of its host star, or it can emit its own light. Reflected starlight won't be anywhere near a match for the light of the star -- especially at large separations -- so that leaves the latter option as our best bet. Planets do emit light (energy) of their own. Some of this energy is absorbed and re-radiated starlight. Some is energy left over from the planet's formation, which is radiated away as the planet cools down. The bigger the planet and the younger the planet, the more energy it has available to give off.

Direct imaging is therefore most sensitive to young, massive planets orbiting far from their host stars. These planets are far-removed from the Super Earths we've been discussing in this course.

The intriguing case of Alpha Centauri Bb

It's possible that we've found a very nearby (non-habitable) planet - it might orbit a star just 4 light-years away. Alpha Cen Bb was detected by Xavier Dumusque and the HARPS (High Accuracy Radial Velocity Planet Search) collaboration using the radial velocity technique. It proved to be an extremely difficult measurement: the size of the signal was just half a meter per second. To find the signal, they had to correct for a large variety of other radial velocity variations; only then did the planet make itself known. The primary sources of noise was the star: its rotation and magnetic activity, its natural pulsations, and fluid motion on its surface.

Because the signal is so small, the methods used to extract the signal are important. Dumusque made public all of their data, which allowed other researchers to try their hands at identifying the signal. According to reports published as of August 2014, the follow-up studies failed to recover the signal. These results don't mean the planet isn't there, but that the current data aren't enough to demonstrate the planet's existence.

Another group, led by Debra Fisher of Yale University, also observed Alpha Centauri B and has an independent data set. Their group did not detect the planet, but noted that they expected to be able to only marginally detect such a small signal.

Солнечная система и ее тайны