43

I have been learning about the solar system from popular science shows. In these shows they suggest that, after having seeing around 2500 other solar systems, astronomers have concluded that our solar system is not the normal one. They see sun hugging hot jupiters and super earths close to their star. They find most systems have most of their matter closer to their star. They draw conclusions about our system, namely that our system is a freak.

We know that the first exoplanets found were hot jupiters. And they found super earths close to their star. We also know that telescope technology is always improving. With increasing telescope quality, we can find dimmer objects. These would naturally be smaller planets orbiting distant stars at further distances from their stars.

My question is this: how do we know that the apparent rare quality of our system is not an artifact of limited observing power that selects for larger objects close to their star?

        

candied_orange
  • 2,287
Ben Sprott
  • 1,420
  • 8
    Relevant question on Astronomy: https://astronomy.stackexchange.com/q/13301/1559. In short, if we'd been looking at the solar system from the outside, we'd have only found Jupiter, Saturn, and if everything lined up just right, maybe Earth or Venus. – Mark May 03 '21 at 22:34
  • 3
    Note that using the Kepler mission’s criterion of three transits to “discover” an exoplanet, the best case for discovering something in a Saturn-like orbit would be transits observed in years 2009, 2039, and 2069. Many other techniques are also biased towards short-period orbits. – rob May 04 '21 at 20:29
  • 2
    @Mark Though it should very much be noted that even those would be very obviously much farther from the Sun than the exoplanets we've observed so far. But there's definitely observation bias at play anyway - closer to the star means shorter orbital periods, which helps most of the ways we try to find exoplanets, among other things that heavily favour such systems. – Luaan May 05 '21 at 08:56

2 Answers2

51

The solar system cannot be said (yet) to be "rare" because we lack the ability to examine the planetary systems around other stars in detail. In particular, the census of low-mass planets and planets that are more than an astronomical unit from their star is very incomplete.

Nevertheless, enough is known to say that the solar system is unusual in some respects. The main oddity about the solar system is that it doesn't contain any "super-Earths" or "sub-Neptunes" at all (i.e. planets intermediate in size between Earth and Neptune), despite them being common in other systems, and the close-in planets are all small and rocky. Most ($\sim 70$%) solar-type stars have at least one exoplanet larger than the Earth orbiting with a period of 100 days or less (e.g. Kunimoto & Matthews 2020). Most of these close-in planets are 2-5 times the size of the Earth.

The presence of a Jupiter-sized planet at 3-7 au is also somewhat unusual - occurring in $<10$% of solar-type stars Wittenmyer 2016).

The lack of a hot Jupiter in the solar system is not unusual, since the occurrence rate of these is only of order 1%.

These frequencies are corrected for the known and well-understood biases in detection sensitivity associated with system geometry, signal-to-noise ratio and observing cadence. These factors can easily be accounted for in a forward modelling approach. This is where you simulate your exoplanet population, then "observe" it, in software, including all the observation biases and detection thresholds, and then adjust the characteristics of the simulated population until the simulated observations match the real observations.

Note for those wanting to discuss the anthropic principle. It seems to me that whether the planetary system around our Sun has an unusual architecture has nothing to do with our presence. What the anthropic principle may have a bearing on is explaining why we live in an unusual solar system or why we might expect the solar system to be unusual.

ProfRob
  • 130,455
  • 1
    If we assume, for the sake of argument, that our solar system is the only one within range of our detectors which is capable of supporting life, then we should reasonably expect to find that our solar system has some unusual features. You have catalogued various ways in which the solar system differs from all other observed planetary systems, yet you seem unwilling to make the logical inference that the observed anomolies account for the uniqueness of our situation. As data continues to be collected, the differences seem to multiply: nowhere do we observe a situation paralleling our own. – Ed999 May 03 '21 at 15:12
  • 5
    The question wasn't to "account for the uniqueness" of the solar system (and there is no evidence that the solar system is unique or "differs from all other planetary systems"); it was to address the question of whether the solar system was unusual and how that can be established given the observational selection effects. @Ed999 – ProfRob May 03 '21 at 15:50
  • Re large, close-in planets, it would seem that there's more than a little selection bias there, because it's much easier to detect large close-in planets with the current detection methods. – jamesqf May 03 '21 at 16:24
  • 5
    @jamesqf Of course. That is why one uses a forward modelling approach, as I say in my answer. The numbers I've quoted above account for these biases. – ProfRob May 03 '21 at 17:14
  • It seems to me that close gas planets are hard to come by, since the star would usually literally blow the atmospheres of those planets away... leaving a small chunk of rock like a mercury. If that is true, then "all close-in planets are hard and rocky" should be the norm. – Christopher Schultz May 03 '21 at 22:10
  • 2
    @ChristopherSchultz that's overly simplistic. A sufficiently massive planet will be able to hold on to its gases even in close proximity to the star, just as well as a smaller one can only do further away. – leftaroundabout May 04 '21 at 08:41
  • 3
    @ChristopherSchultz That was much of the thinking that has been essentially disproved by exoplanet observation so far. And plenty of hypotheses even before that lay serious doubt on that logic - including the fact that it seems that our gas giants were much closer to the Sun in the past, when the Sun's activity was much stronger. Mercury is tiny; even without the proximity to the Sun, we wouldn't expect it to hold on to much more than a tenuous atmosphere. In contrast, Venus is very close to the Sun, and has a massive atmosphere (though granted, made out of heavy gases, not hydrogen). – Luaan May 05 '21 at 09:00
  • "The main oddity about the solar system is that it doesn't contain any "super-Earths" or "sub-Neptunes" at all" Isn't that tautological because you define the standard based on the extremes of our own solar system? No one living in my house is smaller than the smallest person living in my house, but is that an interesting conclusion? – Flater May 05 '21 at 11:42
  • @Ed999: "Uniqueness" is a very strict concept that you can't just derive from "unusual". The answer established that our solar system is, based on what we've observed, unusual, but that does not suggest it is unique, not even relatively so. "As data continues to be collected, the differences seem to multiply" Remember that an average doesn't inherently imply close proximity to its members. No one has a "near average" body shape, but that does not make us all unique in body shape either. – Flater May 05 '21 at 11:45
  • @Flater the analogy is that we do a survey of lots of houses and we find that in almost all cases there is someone bigger than you in the front room. Whereas in your house the only person in the front room is smaller than you. That makes your house unusual, and is an interesting conclusion. Super-Earths and sub-Neptunes means a specific class of planet (that appear to be common), which have radii between that of the Earth and Neptune. – ProfRob May 05 '21 at 12:17
  • @ProfRob: The observation remains flawed for our solar system when the two size boundaries are adjacent on an ordered list of planetary radii in our solar system. If I'm 182cm, and my wife is 176cm, and many people exist with a height between 176cm and 182cm, that does not make it odd that there isn't a third person in my house measuring between 176 and 182cm. That doesn't make it an interesting conclusion, unless we're talking about houses with many, many residents, and if the range hadn't been based on the specific (few) residents of my house to begin with. – Flater May 05 '21 at 12:23
  • 1
    @Flater if you are 182cm, the other person in your house is 176cm and everyone else in every other house has a height between 176 cm and 182 cm, then yes, it does make you very odd. – ProfRob May 05 '21 at 12:30
  • @ProfRob: "Many people of this height exist" is not the same as "every other house contains exclusively people between this height range". – Flater May 05 '21 at 12:31
  • 1
    @Flater and neither have I said that the solar system is very rare or unique. – ProfRob May 05 '21 at 12:32
  • 1
    I feel the need to thank you for your high-quality contributions here. Thank you. – Peter - Reinstate Monica May 05 '21 at 12:38
  • The fact that we do not possess observational data on all stellar systems in our galaxy does not invalidate the fact that for all systems on which we do have such data none shows a parallel to the conditions in the solar system. We must conclude that, based on the available data, there is no comparable stellar system in that data, and that therefore the solar system data is unique. – Ed999 May 06 '21 at 02:32
  • @Ed999 Some exoplanetary systems may be very similar to the solar system. So no, we must not conclude that. – ProfRob May 06 '21 at 06:59
  • Sprinkle equal parts red and green confetti in the lawn. Send your students out to pick it up. Have them count what they collect. The students might argue that you sprinkled far more red confetti. The wise student however will realize that red confetti is easier to find in grass than green confetti and so conclude how biased their data is and not over state what they know. – candied_orange Jun 06 '21 at 14:24
  • @candied_orange The numbers I have quoted account for this bias. – ProfRob Jun 06 '21 at 14:33
  • @ProfRob numbers don’t account for this bias. Understanding the methodology and it’s weaknesses might. – candied_orange Jun 06 '21 at 15:46
  • @candied_orange please read my answer and then check the references I have cited. Observational biases are accounted for. – ProfRob Jun 06 '21 at 16:53
10

I know this is something that scientists take into account, but perhaps their words have been twisted a bit by the popular science shows: there is a measurement bias.

We know what we know about exoplanets because of several ways in which we are able to measure them. Namely, by observing the planets' occultation of their stars (the light dims when the planet crosses in front of it) and by observing the gravitational wobble of the stars due to the planets.

Both of these are going to be easier to see for large planets that are closer to their star.

So I don't know if we are really at a point yet where we can make definitive statements about things such as "our solar system is rare".

geshel
  • 410
  • There's also a selection bias going on: The scientists performing the experiment could only do it from the planet on which life appeared. Whether life evolves or not is impacted by the type of planets in your system. – Jeffrey May 03 '21 at 13:07
  • 4
    @Jeffrey that has no bearing on whether the solar system is unusual or not, only on what conclusions you can draw about why we live in an unusual solar system. – ProfRob May 03 '21 at 14:07
  • @ProfRob I'm pretty sure that if OP lived in a different system, he would not care about whether this one is unusual, but I see your point :-) – Jeffrey May 03 '21 at 14:08
  • 7
    This answer seems to be arguing that these biases can't be accounted for. Which is partially untrue. As long as there is any sensitivity to a particular parameter space then that sensitivity can be used to correct the measured frequencies. – ProfRob May 03 '21 at 17:17
  • @ProfRob True, but it's dangerous extrapolation - you can chart what kind of things we shouldn't expect to see even if they exist - i.e. we expect to find such planets, but it's not weird we didn't yet. Jumping to "therefore they are likely to exist" is a leap, especially given how many older hypotheses about what we should expect to find in a star system have been already cast into serious doubt. – Luaan May 05 '21 at 09:07
  • 1
    @Luaan I haven't said what you claim. Biases in detection sensitivity can (and are) corrected for in all respected work on the topic of exoplanet frequency. Results are only quoted where there is some sensitivity to detection. – ProfRob May 05 '21 at 09:30
  • @ProfRob: You can account for a bias in ratios between observations, but not for things you haven't observed to begin with. You can't counterbias the unknown, as you inherently don't know what it is you'd be countering. – Flater May 05 '21 at 11:51
  • @Flater and that's what I said. – ProfRob May 05 '21 at 12:13