Page 5 of 8 FirstFirst ... 34567 ... LastLast
Results 121 to 150 of 227

Thread: Gliese 581g -- in the habitable zone (barely)

  1. #121
    Join Date
    Sep 2009
    Posts
    963
    Quote Originally Posted by Rowen View Post
    The press is the worst possible place to get scientific facts. They will always stretch, distort and bend the data given them to sensationalize the story to sell papers, ad space etc. Absolutely the wrong place.
    Quote Originally Posted by Rowen View Post
    ...and refuse to pay any attention to the press. Find the the data at the source. For news about a specific research go to that research group, they will all have a website. They are the horses mouth
    Quote Originally Posted by Rowen View Post
    Never, ever rely on wiki for this kind of information. If you follow this general guideline you'll be much better informed.
    I'll echo this extremely good advice.

  2. #122
    Join Date
    Oct 2010
    Posts
    46
    Ok, sure, but actually wiki seems to be quite reliable. They update it on daily base and always cite the source of the information. I've just read the original paper, and seems most of data is pulled from that document. So these info are not false:

    http://fr.arxiv.org/PS_cache/arxiv/p...009.5733v1.pdf

    Some examples of which I was talking about:

    SRC: The estimated FAP of the _37-day peak is 2.7 10−6.

    SRC: The 37-day signal is clearly visible in the HIRES data set alone
    and manifests at the correct period.

    SRC: Apparently, the HARPS data set alone is not able to reliably sense this planet.

    (Which implies that announcement of non detection by Pepe et al. is not a big deal and could have been expected. Btw, where is the paper of HARPS team describing the counter argument? It was said they need to publish it before the counter claim is being taken seriously.)

    Oh and just as a side not regarding:

    "The GJ 581 system has a somewhat checkered history of habitable planet claims."

    Isn't it about the HARPS team, it was their checkered detection and correction. Wasn't it? I don't know btw, just asking...

  3. #123
    Join Date
    Sep 2009
    Posts
    963
    Quote Originally Posted by Bitula
    Ok, sure, but actually wiki seems to be quite reliable.
    Oh some articles are well written with good information. But the fact that many aren't negates the reliability of Wikipedia altogether. While it's not exactly a good source of info, it's a great starting point.

    SRC: The estimated FAP of the _37-day peak is 2.7 10−6.
    This doesn't necessarily mean that the probability that the planet doesn't exist is 2.7 10-6, but rather that with the way they've done their fit to the data, the RV signal of g has that FAP. Being as they seem to only find the planet when the two datasets are put together, it might imply that the RV signal for g is an artificat of the use of two datasets (maybe there's an error somewhere down the line during the merging of the two). Thus, what they may indeed be finding with such a low FAP is a systematic error. The FAP speaks only to the reality of the signal, not to what it actually represents.

    Which implies that announcement of non detection by Pepe et al. is not a big deal and could have been expected. Btw, where is the paper of HARPS team describing the counter argument?
    Perhaps. Though the HARPS team did claim that by trying to force a 37-day periodicity into the RV data, a 37-day period appears in the residuals. This would suggest that the HARPS data is not only sensitive enough to detect the 37-day periodicity, but finds it does not exist.

    It was said they need to publish it before the counter claim is being taken seriously.
    Agreed. It's still far too early to say one way or another. There's a lot of work to do by both teams.

    "The GJ 581 system has a somewhat checkered history of habitable planet claims." Isn't it about the HARPS team, it was their checkered detection and correction. Wasn't it? I don't know btw, just asking...
    When Gliese 581 c was announced, it was hailed as the first candidate habitable exoplanet. Not entirely sure why, however, as it gets more energy from its star than Venus does from the Sun. Further work clearly showed that Gliese 581 c is an unlikely candidate for anything remotely resembling a Earth-like world (tidal lock aside). When Gliese 581 e was discovered, the orbital period of d was revised down from 80-something days to 60-something days. Turns out the ~80-day period was an alias of the true, ~60-day signal. This brought Gliese 581 d to the outer edge of the habitable zone, and was hailed (though with a quieter trumpet) as the first candidate habitable exoplanet.

    It's an understatement to say that the Gliese 581 system has fallen prey to some rather sensationalist ramblings.

  4. #124
    Join Date
    Oct 2010
    Posts
    64
    I had a lengthy post on this, but in hindsight I feel it was a bit inappropriate and somewhat embarrassing, so I'm simply going to say that I am disappointed in how the discovery was handled and leave it at that.
    Last edited by Rowen; 2010-Oct-23 at 05:25 AM.

  5. #125
    Join Date
    Sep 2009
    Posts
    963
    Gliese 581g as a scaled-up version of Earth: atmospheric circulation simulations
    http://arxiv.org/abs/1010.4719

  6. #126
    Join Date
    Oct 2010
    Posts
    46
    I've read somewhere - from a non-original-scientific source though - that the dispute between the two teams is about whether the 37 day signal is a planet or part of 581 d orbit. But HIRES team claims that two circle orbits is a better fit then one elliptic D. And given that this can be said only with HIRES data involved and that checkered history is not on their side it seems likely that G does exist. However interestingly, there is a serious debate regarding planet D, which as I see it in the original paper lowers the chances for a fit to just about 95%. Now if parameters of D, or D itself turns out to be false then planets from e to f could be invalidated I guess. Or that is how I understood it. So I give existence of G a 95% chance. That's still quite a lot. Btw, I don't see any official site regarding this whole issue, so my only chance is to read the press releases and Wiki or probably someone here at BAUT has info which is closer to the source.

  7. #127
    Join Date
    Sep 2009
    Posts
    963
    An investigation you may be interested in,

    Collect the data from the paper describing the discovery of Gliese 581 g, and plug it into the Systemic console.
    (download link )

    Some have done this, posting interesting results here.

    Of course keep in mind that we don't have the current HARPS data set to work with, so we're a bit limited as to what we can work with. But we do have the HARPS 2009 + HIRES data set on which the discovery is claimed, so we can persue some fairly independent investigation of the issue.

  8. #128
    Join Date
    Oct 2010
    Posts
    46
    According to that thread no combined data-set investigation has been performed. Only HARPS 2009 has been examined, and even so G can be found (with a high FAP I guess), but that is same as the original paper claims: " SRC: Apparently, the HARPS data set alone is not able to reliably sense this planet." Or have I misread something there? And: " SRC: The 37-day signal is clearly visible in the HIRES data set ALONE and manifests at the correct period."

  9. #129
    Join Date
    Oct 2010
    Posts
    64
    Maybe the two data sets are incompatible with each other and should be treated separately?

  10. #130
    Join Date
    Sep 2009
    Posts
    963
    Quote Originally Posted by Bitula View Post
    " SRC: Apparently, the HARPS data set alone is not able to reliably sense this planet." Or have I misread something there?
    Via the Dynamics of Cats blog.
    "HARPS statement is stronger than "we don't see it" - they find that if they force a solution they get a negative signal appearing, implying the planet is not there, not just that they are not sensitive to it."

    SRC: The 37-day signal is clearly visible in the HIRES data set ALONE and manifests at the correct period.
    That, actually, may be a bad thing, if what the HARPS team is claiming is true. Smaller data sets have a hgiher chance of having aliases. If g is an alias of another planet's signal, then while the periodicity may be there, it may not correspond to a real planet (recall that the 3-planet fit had d at 80-something days, but that turned out to be an alias with additional data).

    But indeed the Keck team claims they find it in both the HIRES + HARPS data set (according to the paper), so this is where the independent investigation with the systemic console comes in handy.

  11. #131
    Join Date
    Sep 2009
    Posts
    963
    Quote Originally Posted by Rowen View Post
    Maybe the two data sets are incompatible with each other and should be treated separately?
    Data points comprise of just three attributes. A julian date, the measurement, and the 1σ error. So at a basic level, it's trivial to combine data sets.

    There's a lot of instrumental effects though that can cause the data from one telescope to be shifted slightly relative to the data of another telescope, with the errors arising from anything from the photocentre not being properly aligned with the spectrometer slit (errors larger than 1 m s-1 will occur if the photocentre moves away from the slit by a few thousandths of the slit width!), temperature considerations (a 1 K change in instrument temperature can produce a 90 m s-1 RV drift), change in air pressure near the instrument (a 1 mbar change in pressure can result in a ~300 m s-1 drift!). HARPS keeps their spectrometer in a vacuum chamber to help deal with this, but small wavelength shifts can't be avoided so they typically have to model them out of the data. Motion of Earth (rotation, orbit around the sun, Earth+moon barycentre) and the solar system needs to be accounted for. They do affect the observed radial velocity of the target.

    There's a lot of room for error.

  12. #132
    Join Date
    Oct 2010
    Posts
    46
    Quote Originally Posted by Hungry4info View Post
    Via the Dynamics of Cats blog.
    "HARPS statement is stronger than "we don't see it" - they find that if they force a solution they get a negative signal appearing, implying the planet is not there, not just that they are not sensitive to it."
    Yes, but again, they looked into their own data alone. And it should be considered, as well, that the Keck telescope is almost 3 times the size of the one used with HARPS. This should make a difference. Just speculating though.... . I've read somewhere that Pepe received all HIRES data, so now probably HARPS team finally is investigating the combined data set. This dispute should be settled in a few weeks thus, and IMO it will come out that additional data from a third source will be required, which may take several years.

  13. #133
    Join Date
    Sep 2009
    Posts
    963
    And it should be considered, as well, that the Keck telescope is almost 3 times the size of the one used with HARPS.
    Telescope size isn't the only limiting factor. Instrumental sensitivity, exposure time, photon noise, etc is also crucial.

    A quick look at the paper that presents the discovery of g and f makes it clear that between the two data sets, HARPS's is more accurate (generally lower 1σ errors). HARPS has been at the forefront of low-mass extrasolar planet science, discovering more low-mass planets than any other group to date. It's a shame there isn't more spectrometers like this one.

    and IMO it will come out that additional data from a third source will be required
    What do you base that on? (specificially the need for a third source)
    Last edited by Hungry4info; 2010-Oct-26 at 09:51 AM. Reason: Grammar.

  14. #134
    Join Date
    Oct 2010
    Posts
    46
    Telescope size isn't the only limiting factor. Instrumental sensitivity, exposure time, photon noise, etc is also crucial.
    Ok, but this is the only thing I can think of which would justify planet G. Otherwise why would Vogt attempt discoveries with an inferior spectrometer and a smaller set of data?

    What do you base that on? (specificially the need for a third source)
    Both teams seem to agree that G is so small that it is on the margin of detect ability with their instruments. What would happen if they hook HARPS to Keck? - that seems to be a promising solution. There is also the Telescope Gran Telescopio Canarias, but it is just slightly bigger than Keck.

  15. #135
    Join Date
    Sep 2009
    Posts
    963
    Quote Originally Posted by Bitula View Post
    Ok, but this is the only thing I can think of which would justify planet G. Otherwise why would Vogt attempt discoveries with an inferior spectrometer and a smaller set of data?
    It may be an issue of HIRES just being more accessible to him. IIRC, the HARPS time is rather allocated amongst various programs. Note that, according to the paper, Vogt et al. were only able to find g with the combined HARPS + HIRES data set. Lower precision can be made up for with a higher abundance of data.

    Quote Originally Posted by Bitula View Post
    What would happen if they hook HARPS to Keck? - that seems to be a promising solution. There is also the Telescope Gran Telescopio Canarias, but it is just slightly bigger than Keck.
    Oh I'd give my left leg to see them do that. But I imagine there's difficulties involved. HARPS is a Hungarian-built spectrometer, and I don't know how easy it is to just hook up to another telescope (political and technical reasons).

  16. #136
    Join Date
    Sep 2009
    Posts
    963
    There is, in development, a "laser-comb" spectrometer that will deliver precisions of 1 cm s-1, far greater than either HARPS or HIRES. It will be installed on ESO's E-ELT.
    http://www.eso.org/public/news/eso0826/

    This will be fun.

  17. #137
    Join Date
    Oct 2010
    Posts
    46
    Oh I'd give my left leg to see them do that. But I imagine there's difficulties involved. HARPS is a Hungarian-built spectrometer, and I don't know how easy it is to just hook up to another telescope (political and technical reasons).
    You sure? Interesting to know (I'm Hungarian...), I thought it is Swiss. Are you not confusing it with HATP?

    Other, from wiki:

    "HIRES has detected more extrasolar planets than any other in the world. The radial velocity precision is up to one meter per second (1.0 m/s)
    Vogt is known worldwide for designing and building HIRES."

    Again, not sure, but In 15 years of exoplanet hunting, over hundreds of planets were detected by his team without an error.
    And now he makes a single false claim (actually two g and f)? Highly unlikely to me. Seems he knows what he is doing.

  18. #138
    Join Date
    Sep 2009
    Posts
    963
    Quote Originally Posted by Bitula View Post
    You sure? Interesting to know (I'm Hungarian...), I thought it is Swiss. Are you not confusing it with HATP?
    Ah! Yes you're right.

    Quote Originally Posted by Bitula View Post
    Highly unlikely to me. Seems he knows what he is doing.
    Oh I agree he knows what he's doing. I'm just saying it doesn't make him immune to fault. Between the photons escaping the stellar surface, and the paper being published, there's a lot of room for human error.

    Everyone has a first mishap. This may be Vogt's. I'm not saying it is, just that a pro-hominem is just that, a pro-hominem.

  19. #139
    Join Date
    Mar 2004
    Posts
    2,501
    Quote Originally Posted by Hungry4info View Post
    Gliese 581g as a scaled-up version of Earth: atmospheric circulation simulations
    http://arxiv.org/abs/1010.4719
    (apologies, real-life occassionally interrupts even the most engaging cyber-discussions! trying to catch up on the discussion and will probably respond to additional posts up and down thread over the next 24 or so. Forgive me if my posts seem to jump around a bit.)

    Awful lot of initial presumptions here.

    The Earth is strongly a product of the conditions of its birth, nothing about this putative planet (Gliese 581g) speaks of even generally the same conditions of formation.

    Did the planet form before the former giants in tighter inner orbits migrated in from the outer solar system? or did it form post migration?

    One of many issues where either response would seem to lead to drastically different conditions than those that gave rise to the inner-system rocky worlds as we recognise them in our own system. I guess, if you're limiting yourself to real general conditions, yeah, you can stab in the dark a bit, but this study seems more like aerodynamic studies per alia porci. (IMO)
    Last edited by Trakar; 2010-Oct-29 at 10:33 PM. Reason: clarification qualifications = bolded

  20. #140
    Join Date
    Mar 2004
    Posts
    2,501
    Ahh! I remembered reading this and then had a dickens of a time relocating it (of course I spent most of my time looking in the wrong threads, but that is becoming a bit more frequent type of event as the years go by!).

    Quote Originally Posted by Hungry4info View Post
    Regarding the activity level of Gliese 581 though, the star is quiet. It was studied by the Canadian MOST satellite to search for transits of c, but none were found,
    I'll provide a good citation once I find it.

    In the mean time, there's at least a Centauri Dreams blog on the subject.
    http://www.centauri-dreams.org/?p=1298

    Edit:
    In this paper, the activity of the star (as well as MOST results) are considered.
    interesting, but I don't see this as contradicting, but rather supporting my earlier statement

    for reference:
    Ok, but what you wrote is: "with lots of major flare activity". Which is not true, isn't it?
    No, it is accurate, but relative considerations are needed to properly qualify the terms. For example, our own Sun is a relatively quite variable yellow dwarf star (dG2V) and on average it produces a flare about once a day and a major flare about once a century. dM class stars in general experience flares on the average of every hour or so and a major flare every few weeks to months. Compared to the average dM star, Gl-581 is quiescent with (IIRC) maybe a handful of flares a day and only a few major flares per year, but compared to our own Sun, this is approximately two orders of magnitude greater activity.
    I'd have to go back to the paper to confirm with certainty, but I believe the study you mention stated relative quiescence with a measured variation in output over the six week observation period of being on the order of a few tenths of a percent. Our own Sun is an active dG2V (yellow dwarf -variable) star only varying in output by around one tenth of one percent over an eleven year solar cycle. Gl581, according to this study, varies significantly more, over a dramatically shorter time frame,...apparently. Not a flare star (currently, but there are some studies indicating that many if not most dM stars are flare stars for the first 1-2 billion years of their existence, in which case the effects and impacts of those earlier more active periods upon any close-in planets should be considered,...or am I mistaken in this understanding?), but still several orders of magnitude more active than our own active and variable star.

    Edit2: The HO Lib designation was a mistake if I recall right. The star was misidentified as variable for some reason.
    I've seen no official statement of such, and it is still listed as such in SIMBAD and as an active variable (of one or several differring varieties - according to local convention) in all the registries I've checked. If you have or find reference to this re-designation please post it so that I can update my understandings and considerations.

    http://simbad.u-strasbg.fr/simbad/sim-id?Ident=GJ+581

    Addendum - MOST (Microvariability and Oscillations of STars) link
    Last edited by Trakar; 2010-Oct-29 at 10:17 PM. Reason: addendum

  21. #141
    Join Date
    Sep 2009
    Posts
    963
    Regarding the reference to the misidentification of HO Librae, apparently Weis et al claim a photometric variability of < 6 mmag, consistent with MOST results. I don't seem to find a reference that specifically states that Gliese 581 was misidentified as variable (alas, it was something I had heard a while back), but I can't find any reason to believe it is.

    As regards to stability of Gliese 581 for RV studies, some of the most definitive statements I could find are in the discovery paper of the fourth planet (this paper),
    The strength of the Ca II H&K emission lines in addition classifies GJ 581 in the lowest ~5% magnetic activity bin of our M dwarfs sample.
    The strongest photometric stability constraint is provided by 6 weeks of continuous monitoring of GJ 581 by the MOST satellite. Preliminary results ... [snip!] ... indicate that the peak-to-peak photometric variability during these six weeks was less than 5 mmag.
    Last edited by Hungry4info; 2010-Oct-31 at 01:50 AM. Reason: Spelling "misidentified"

  22. #142
    Join Date
    Mar 2004
    Posts
    2,501
    Quote Originally Posted by Hungry4info View Post
    Regarding the reference to the misidentification of HO Librae, apparently Weis et al claim a photometric variability of < 6 mmag, consistent with MOST results. I don't seem to find a reference that specifically states that Gliese 581 was misidentififed as variable (alas, it was something I had heard a while back), but I can't find any reason to believe it is.

    As regards to stability of Gliese 581 for RV studies, some of the most definitive statements I could find are in the discovery paper of the fourth planet (this paper),
    Weekends aren't good research times for me, but I'm down in Oakland today and its rainy, my wife is on a shopping trip and its either the laptop or the hotel bar. Since its not yet noon, I'm tapping away on the laptop.

    All of the stellar catalogs and discussion papers I have access to are listing Gl581 as a variable star (HO librae, BY Draconis, etc.). By dM standards, it is admittedly a rather sedate and mild varient of the species, at least currently. It is impossible to say for certain how representative the 6 week observation period by MOST is, but if it presents a typical variation, and is looked at in comparison to our own variable Sun, it is vastly more active with a varience of "a few tenths of a percent" over a 6 week period of observation by the MOST team where our own Sun varies by a tenth of a percent over an ~572 week period. For a Red Dwarf star it seems relatively calm and stable, but compared to our Sun, Gl581 is what, roughly 3 to 4+ orders of magnitude more active? If I'm missing or misunderstanding something here I'd appreciate any efforts any can supply to help me to a more complete and accurate understanding.

    http://www.astro.ubc.ca/MOST/papers/Gliese.pdf
    and the associated collection
    http://www.astro.ubc.ca/MOST/papers/

  23. #143
    Join Date
    Sep 2009
    Posts
    963
    Aliases of the first eccentric harmonic : Is GJ 581g a genuine planet candidate?
    http://arxiv.org/abs/1011.0186

    The radial velocity (RV) method to detect extrasolar planets has been the most successful to the date. The RV signal imprinted by a few Earth-mass planet around a cool star is at the limit of the typical single measurement uncertainty on state-of-the-art spectrographs. As a consequence, one has to rely on statistics to unearth signals buried below the noise. Observationally related issues such as unquantified systematic errors or artifacts introduced by the observing cadence can produce spurious signals. Also, they can mask genuine signals that should be easily detected otherwise. This is the case for the planetary system around the low mass star GJ 581 for which there has been recently announced a 3.1 Earth mass planet in its habitable zone. Even though it is a very stable star, the combination of the observing cadence and the presence of multiple planets has already caused a number of period misinterpretations in the past. We discuss here a particularly devious statistical degeneracy that derives from the aliasing of the first eccentric harmonic of an already detected planet with the characteristic one year sampling frequency. Such a degeneracy can prevent the detection of the true signal and correlate the eccentricities of known planets with the mass determinations of additional low amplitude companions. By performing a number of statistical tests, we conclude that even though the statistical degeneracy is clearly present, the existence of GJ 581g remains well supported by the available data.

  24. #144
    Join Date
    Jun 2010
    Posts
    652
    Quote Originally Posted by Trakar View Post
    By dM standards, it is admittedly a rather sedate and mild varient of the species, at least currently. It is impossible to say for certain how representative the 6 week observation period by MOST is, but if it presents a typical variation, and is looked at in comparison to our own variable Sun, it is vastly more active with a varience of "a few tenths of a percent" over a 6 week period of observation by the MOST team where our own Sun varies by a tenth of a percent over an ~572 week period. For a Red Dwarf star it seems relatively calm and stable, but compared to our Sun, Gl581 is what, roughly 3 to 4+ orders of magnitude more active? If I'm missing or misunderstanding something here
    The facts you quote don't constrain the amplitude of 581's variability to be greater than the Sun's. Is the frequency of particular significance? "Activity" is problematic in more than one way and I suspect the variation in the star's output usually isn't the main problem.

  25. #145
    Join Date
    Oct 2010
    Posts
    46
    Quote Originally Posted by Hungry4info View Post
    Aliases of the first eccentric harmonic : Is GJ 581g a genuine planet candidate? http://arxiv.org/abs/1011.0186
    Great!
    Also in this paper:
    Still, given the available data, the FAP for planet g is low enough (0.36%) to
    justify the claim of detection of the potantially habitable planet candidate GJ 581g by V10.

    Now chances are even higher, that this planet indeed exists! So far there are zero counter arguments and one pro (regarding official papers).

  26. #146
    Join Date
    Mar 2004
    Posts
    2,501
    Quote Originally Posted by Hungry4info View Post
    Aliases of the first eccentric harmonic : Is GJ 581g a genuine planet candidate?
    http://arxiv.org/abs/1011.0186
    I suppose this is in response to the earlier Science piece?:
    "First Goldilocks Exoplanet May Not Exist"
    Kerr
    Science 22 October 2010: 433
    DOI: 10.1126/science.330.6003.433
    http://www.sciencemag.org/cgi/content/full/330/6003/433
    "Astronomers have begun debating whether the first claimed habitable planet circling another star is a figment of their data analysis. At a meeting last week in Turin, Italy, a group of exoplanet hunters announced that its observations show no sign of Gliese 581g, the Goldilocks planet on which conditions are just right for harboring life. Only a couple of weeks earlier, a rival group had announced with much fanfare its discovery of the long-sought Earth-like habitable planet..."

  27. #147
    Join Date
    Mar 2004
    Posts
    2,501
    Quote Originally Posted by Hungry4info View Post
    Aliases of the first eccentric harmonic : Is GJ 581g a genuine planet candidate?
    http://arxiv.org/abs/1011.0186
    An additional paper which helped my understanding of the nature of aliases and the methodologies which generate them and the various approaches applied (with relative success and failure) to distinguish such aliases from true signals (and which also actually includes some information on how these have been used with regards to other planets in the Gliese 581 system):

    "RADIAL VELOCITY PLANETS DE-ALIASED. A NEW, SHORT PERIOD FOR SUPER-EARTH 55 CNC E"
    http://arxiv.org/PS_cache/arxiv/pdf/...005.4050v3.pdf

    ...Many problems with aliases are the result of unwarranted assumptions. We describe some common misconceptions about aliases and how they cause confusion.
    1. Assuming that the largest peak in the periodogram
    is the physical frequency. In fact, noise may add coherently to an alias or incoherently to the physi-
    cal frequency, causing the alias to appear larger. This is what happened for Gl 581 (Udry et al.
    2007; Mayor et al. 2009). In multi-planet systems, aliases from several planets could add to make the
    highest peak a spurious signal (Foster 1995).
    2. Assuming that the frequency that yields the best Keplerian or Newtonian planet fit is the true fre-
    quency. As we saw for Gl 581 d, this is not always the case, due to noise.
    (...)
    just a sampling, but the paper is a good explanation of many of the types of issues which need to be addressed in exoplanet research. Excellent educational paper even though it came out prior to the most recent nomination of Gl581g, it does lend a much better explanation of the types of issues which are now being explored and re-examined in the efforts to confirm or reject its discovery.

  28. #148
    Join Date
    Mar 2004
    Posts
    2,501
    Quote Originally Posted by Bitula View Post
    Great!
    Also in this paper:
    Still, given the available data, the FAP for planet g is low enough (0.36%) to
    justify the claim of detection of the potantially habitable planet candidate GJ 581g by V10.

    Now chances are even higher, that this planet indeed exists! So far there are zero counter arguments and one pro (regarding official papers).
    I urge you to adopt more of the paper's authors' sense of reserve. As they themselves state:
    ...As a warning, we want to remark that statistical significance tests are very sensitive to
    choices made by different authors (including this work). Therefore, the ultimate confirmation
    will require additional RV measurements and the combination of all the available data in very
    convincing way. Also, since we are exploring signals with amplitudes below the typical single
    measurement precission, one can still be fooled by unnacounted low amplitude systematic
    effects. In addition to obtain more RV measurements, we encourage observing the star at
    more extreme parallax factors, this is, at times when the star is only visible a few hours at
    the begining and at the end of each night. This will result in a significant mitigation the
    yearly aliases that plague the current data.

    We end with a cautionary note to the detection of low amplitude planets in multiple
    systems. The aliasing peaks can appear very far from the nominal period or its eccentric
    harmonic. The eccentric harmonic of a long period giant planet coupled with the aforemen-
    tioned sampling frequencies may cause spurious low amplitude signals also mimicking an
    Earth mass planet in the habitable zone of a sun-like star. Let’s take Jupiter as an example.
    Its eccentric harmonic would have a period of 1/5.5 years−1(or 0.00049779 days−1). The
    corresponding yearly aliases would then be f0.0027378 days −1, giving candidate signals at
    446 and 309 days periods. Both fake planets would be within the habitable zone of a G2V
    star; mimicking a system rich in habitable planets. Also a genuine habitable planetary signal
    can be missed if the eccentricity of an outer planet is adjusted prematurely. In agreement
    with Dawson & Fabrycky (2010), we think that aliasing tests on the eccentric harmonics of
    detected planets will be necessary in future claims of very low amplitude signals.
    IOW - "your actual mileage may vary"

  29. #149
    Join Date
    Sep 2009
    Posts
    963
    @ Trakar,
    A good page on aliases in RV data is this page, which discusses the paper you referenced as well.

    Quote Originally Posted by Bitula
    Still, given the available data, the FAP for planet g is low enough (0.36%) to justify the claim of detection of the potantially habitable planet candidate GJ 581g by V10.
    Perhaps. While the FAP may be low, what it represents is what is being debated here, I think. This paper is encouraging, though I am still curious to find out what the HARPS team's final conclusions are.

    I would write more, however I see Trakar has already posted it.

  30. #150
    Join Date
    Mar 2004
    Posts
    2,501
    Quote Originally Posted by Hungry4info View Post
    @ Trakar,
    A good page on aliases in RV data is this page, which discusses the paper you referenced as well.


    Perhaps. While the FAP may be low, what it represents is what is being debated here, I think. This paper is encouraging, though I am still curious to find out what the HARPS team's final conclusions are.

    I would write more, however I see Trakar has already posted it.
    interesting discussion, one of the issues I have been attempting to get up to speed on, and in line with some of the discussions that others informed me were very much in play at the the IAU conference last month, thank you for sharing!

    PS, please don't let the fact that I have touched upon the subject shy you from the topic, your thoughts and comments are valued.

Similar Threads

  1. Could Chance for Life on Gliese 581g Actually Be 100%?
    By Fraser in forum Universe Today
    Replies: 19
    Last Post: 2014-Sep-09, 10:09 PM
  2. Replies: 0
    Last Post: 2010-Oct-12, 10:50 PM
  3. Earth, Barely Habitable?
    By Fraser in forum Universe Today
    Replies: 30
    Last Post: 2008-Jan-17, 04:13 PM
  4. Replies: 2
    Last Post: 2007-Dec-14, 10:46 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •