Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

B Exoplanets - Suspicious evidence and fantastical conclusions

  1. Jan 14, 2019 #1
    Hi,

    I thought I would come and ask here as something has been bugging me for quite a while.

    Exoplanets.

    I'm not being bugged by the fact we are looking for them or that they exist. I am a bit suspicious of some of the evidence and how it's interpreted. While the adage is "Extraordinary claims require extraordinary evidence" I feel that some of the exoplanet findings are the other way around. It seems to me that the evidence being found amounts to extraordinary and the conclusions gleaned from this evidence are extraordinary.

    Take for example gas giants that have orbital periods of only a few days. Another example is the higher than expected number of very close and very large planets. The astrophysics community seem to be going all out to find explanations for these extra-ordinary examples.

    Is it not more likely that there are issues with the evidence gathering? Is it not scientific that if an experiment results in completely unexpected or extraordinary results that you first question your experimental approach to double and triple check it is not an artefact of your experiemental design... before you start coming up with fantastical conclusions like a gas giant with an orbital period of a few days... which does not seem likely!

    I think the science is too young and the methods used to vague and speculative, open to too much interpretation to be making the kind of claims that are a weekly occurrence in the popular science media.

    Am I the only one?

    Paul
     
  2. jcsd
  3. Jan 14, 2019 #2

    DrClaude

    User Avatar

    Staff: Mentor

    While some scientists jump too quickly on bandwagons (FTL neutrinos come to mind), I donate think that the claims here are extraordinary. We used to have a single data point for the formation of planetary systems, so it is not surprising that the observation of other systems has changed what we thought was possible.

    Also, there are many different ways in which exoplanet are detected, so it would be surprising that all these methods "conspire" to produce the same kind of incorrect data.
     
  4. Jan 14, 2019 #3

    Vanadium 50

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    Can you cite a prediction for this?
     
  5. Jan 14, 2019 #4
    I think OP is referring to the situation as it was in the 1990s. With just one planetary system as a reference, and way lower computing power, scientists had much simpler planet formation models than nowadays. Inner planets were expected to be small and rocky, outer – massive and gassy. It was no surprise to anyone that massive and close-in planets were first to be discovered by the Doppler method, because this is exactly what you need for strong RV signal. But the first class of discovered exoplanets (actually after pulsar planets) unexpectedly combined these two characteristics. People did not predict migration – Jovian planets were expected to form and stay beyond the snow line. That was the time of wild guesses about exoplanets – some supposed that 51 Peg b and its kind were enormous rocky planets (150 Earth masses!). Now we are in much different situation.

    The pool of knowledge gathered over >20 years is enormous. Some hot jupiters are studied to such detail that we observed weather patterns on them. Yes, sometimes scientists jumped to conclusions over weak signals corresponding to interesting orbits and some discoveries were retracted. But almost 4000 are considered confirmed. The papers often start with long discussion about signal analysis, noise reduction, massive unseen companions, starspots and other forms of stellar activity, before claiming discovery of a planet. Thanks to Kepler mission we finally found these small planets that were missing – they were hiding in the noise, and still are easier to spot transiting the stars than to find in RV signal. Large planets on far orbits, on the other hand, are much less likely to transit (it's just geometry) and if observed over a fraction of orbital period, they are hard to confirm.
     
  6. Jan 16, 2019 #5

    stefan r

    User Avatar
    Science Advisor
    Gold Member

    I believe the rapid orbit planets provide more reliable data. Science is done by reproducing results. If it takes many years for a planet to orbit then it takes many years to repeat the observation. The close in planets are easier to see. The close, large planets have a much stronger gravitational effect on their host star. The hot Jupiters will have the loudest, brightest, and most reproducible signal.

    Why pick on exoplanets? The LIGO detectors are looking for stellar mass objects (black holes) that orbit each other hundreds of times per second. There are also stars that orbit each other within days. Algol (beta persei) has been observed for hundreds of years. Beta persei Aa1 and beta persei Aa2 orbit each other in 2.9 days. Both stars have a radius much larger than the Sun. Io orbits Jupiter in 1.8 days. What is wrong with an orbit of a few days?

    What would you consider a likely orbital period for a gas giant?

    I would look at what type of gas giants are observed. Then use that data to determine what I believe to be most likely.
     
  7. Jan 16, 2019 #6
    Used to be the only practicable way of hunting exo-solar planets was photographic astrometry, so-patiently accumulating star positions for many years, measuring their 'wriggle'. Peter van de Kamp did much, much work thus. Sadly, it came to naught, as his reports of planets around Barnard's Star turned out to be due to a tiny offset in telescope alignment after cleaning etc. Kin to Hubble's mirror mis-calibration. Weep...

    Then came the Doppler detectors, and 'Hot Jupiters' whose cyclic tug on their parent stars could be spotted within hours, days then weeks, not decades. Better, this work could be readily duplicated, errors quickly spotted, residuals accumulated.

    Once you know such can be found, other methods get deployed. Eclipsing, lensing etc etc.
    And now there's a ruddy zoo of discoveries...
    Fun thing about eclipsing extra-solar planets is that the orbital plane is tightly constrained so, if you can get Doppler data too, that's seriously synergic for orbital determinations...

    One unsettling thought is that for every star with an 'eclipsing' planet, there must be a dozen or more where the orbital plane does not align with us. May be that finding stars / star-systems *without* any planets becomes unusual...
     
  8. Jan 17, 2019 at 4:38 AM #7
    I can add that the sensitivity limits have moved significantly with all the effort put in this topic in last decades. I remember the chart from 2004, showing prospects for planetary detection in the coming years (now in the past). That one: https://commons.wikimedia.org/wiki/File:Extrasolar_Planets_2004-08-31.png. You may note that Space Interferometry Mission was still in development and TESS has not been conceived yet.
     
    Last edited: Jan 17, 2019 at 5:00 AM
  9. Jan 20, 2019 at 5:23 PM #8

    Drakkith

    User Avatar
    Staff Emeritus
    Science Advisor
    2018 Award

    No. The evidence itself, for the transit detection method, is simply a light curve. Or, more specifically, a measurement of the light of a star over many successive images. This is extremely easy to do, and any problems with the sensor or other equipment would be easy to see since it would affect the entire image. Once you take the measurements, you plot the results on a graph and you get the light curve. The dip in the light curve that results from the transit of an exoplanet is different than other solar effects such as the formation of sunspots. They also repeat themselves, and these repeated, identical dips over time are what we use to make absolutely sure we're looking at an exoplanet transit and not something else. There just aren't any other phenomena that could generate these identical repeated dips.

    Now, you could claim that maybe there's some other unknown phenomenon that's generating these dips, but the problem is that these dips are exactly what we would expect in an exoplanet transit. So if it looks like a duck, quacks like a duck, and there isn't another animal that better explains it, then it's probably a duck.

    The same applies for other detection methods.

    They are definitely not artifacts of the experiment's setup or design. I have personally taken a light curve of a hot jupiter using my own telescope and imaging equipment. The dip in the light curve is quite unmistakable. If they were artifacts, I would expect to see them in the light curves of other stars in the same image. But I don't see them.

    I'm afraid you have things backwards. Given amplitude of the dip in the light curve, the duration of the dip, and the shape/duration of the entry and exit slopes, the simplest explanation is that a large object, approximately the size of Jupiter, is orbiting very close to its host star. ALL other possibilities are more speculative and more complicated. A hot-Jupiter type planet may seem like an extraordinary and exotic thing, but it is actually fairly mundane in the context of astrophysics. It's just a big planet with a small orbital radius.

    Also note that hot-Jupiters are MUCH easier to find than any other exoplanet. Their large size and mass combined with their small orbital radius and quick orbital period causes very large observable effects relative to smaller planets located further from their host stars. There are almost certainly a far greater number of more 'normal' planets than hot-Jupiters.

    I don't agree. I can't think of any other interpretation that would reasonably explain the light curves seen during transit events, and I certainly don't see any part of it that is vague and unreasonably speculative. Could be more specific about what you think is vague and speculative? Essentially all of the smaller pieces of the theory/model (measuring the light, orbital mechanics, geometry of the transit, stellar properties, etc) are the result of well understood science and math that is decades if not centuries old.
     
  10. Jan 22, 2019 at 12:36 PM #9
    Surely a hot jupiter would have it's atmosphere stripped off it. I mean I know I could calculate the distance from a star for a 5 day orbit (or even faster) but it's certainly less than half of mercury, so the gaseous atmosphere must be exposed to massive temperatures and solar winds.

    That's why I felt, intuitively, it sounded wrong.
     
  11. Jan 22, 2019 at 12:44 PM #10

    Bandersnatch

    User Avatar
    Science Advisor

    A gas giant also has sufficiently high gravity to hold onto its atmosphere.
     
  12. Jan 22, 2019 at 3:32 PM #11

    Drakkith

    User Avatar
    Staff Emeritus
    Science Advisor
    2018 Award

    Even if it doesn't, there's a loooot of atmosphere to burn off before you make an appreciable dent in the planet.
     
  13. Jan 22, 2019 at 6:48 PM #12

    Vanadium 50

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor

    Your objection seems to be that everyone is wrong because their results do not agree with your preconceptions. If you have some calculations that shows "a hot jupiter would have it's atmosphere stripped off it", please show them. But "intuitively, it sounded wrong" just means "I don't like the answer". And unless you are a planetary scientist by profession, where did you develop this intuition?

    Furthermore, you do know there are directly imaged exoplanets, don't you?
     
    Last edited: Jan 23, 2019 at 4:33 AM
  14. Jan 22, 2019 at 9:18 PM #13

    jim mcnamara

    User Avatar

    Staff: Mentor

  15. Jan 22, 2019 at 11:52 PM #14

    stefan r

    User Avatar
    Science Advisor
    Gold Member

    You could rough it out. The Sun has around 13x the surface gravity vs Jupiter. Within √13 solar radii (around 4) that would be a serious problem for Jupiter. Expansion from heating and rotation would increase the distance.
    Orbiting right at the sun's surface would be around 2.5 hours period. So a Jupiter with 2.5 day orbit is a bit safer. You can also have Jupiters with 13 times Jupiter's mass. Their radius does not increase much and usually decreases with mass. A cold 15 Jupiter mass object would take gas from the Sun not the other way around.
     
  16. Jan 23, 2019 at 2:32 AM #15
    Evaporation of hot jupiter atmospheres has been confirmed with observations of first transiting planet of the kind, HD 209458 b. With a mass slightly higher than Jupiter and orbit eight times tighter than Mercury's around star a bit hottter than Sun, it is surrounded by giant cloud of escaping hydrogen and can be considered quite extreme. However spectacular it may look, the process is not strong enough to strip more than a few percent of its mass throughout the lifetime of the system.
    A planet must have even tighter orbit and lower mass (like Saturn or Neptune) to be completely stripped of its atmosphere, if and only if its star fries it with a very high flux of extreme ultraviolet radiation for a prolonged period. There are several papers comparing theoretical models of the phenomenon with the observed population of hot jupiters and hot superearths.
     
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Have something to add?