This was before e and b has been confirmed by the HIRES team, so seems like the corrections therein are false?"RADIAL VELOCITY PLANETS DE-ALIASED. A NEW, SHORT PERIOD FOR SUPER-EARTH 55 CNC E"
I have read that. I'm not saying that this paper confirms g, just that it increases the probability of it's existence.I urge you to adopt more of the paper's authors' sense of reserve.
Did they publish anything so far? I don't understand why it takes them so long to publish their counter argument.This paper is encouraging, though I am still curious to find out what the HARPS team's final conclusions are.
look at this link:
Can this be true? Where is the analysis Shostak refers to?
55 Cancri was understood to be a five-planet system when this came out. I'm not really sure where the data for the five-planet fit comes from, but I think most of it is HIRES. The fit to the data with the revised period of planet e is much better (lower χ2) with the shorter, "de-aliased" period. I've not seen a response to that study yet.Originally Posted by Bitula
They may be collecting more data before coming out with an official statement or a paper. I think it's clear that more RV measurements are needed to disentangle the existence of g from the eccentricity of d.Originally Posted by Bitula
I dunno. I have no idea what qualifications Seth Shostak has when it comes to extrasolar planet science. I don't know if he works in that field, or if it's entirely in SETI. I didn't see anything in that link worth taking to heart.Originally Posted by Bitula
Indeed! The fact that two papers related to Gliese 581 g have already appeared on arXiv so soon after the discovery paper was quite a surprise to me.Originally Posted by Trakar
Oh, ok, I thought Gliese 581e and b planets are discussed there whereas it is HD 156668 b and 55 Cancri e.55 Cancri was understood to be a five-planet system when this came out. I'm not really sure where the data for the five-planet fit comes from, but I think most of it is HIRES. The fit to the data with the revised period of planet e is much better (lower χ2) with the shorter, "de-aliased" period. I've not seen a response to that study yet.
Sure, but that will take a lots of time, several months at least. I think, if they are serious about their counter claim, they should publish something before that based on their new data set which was not available for Vogt. If, as you say, forcing g into the model does not work at all, as so it is not just about the fact that it is not clearly detectable in their dataset, they should publish the counter argument. No additional data is needed for that. IMO g fits in their dataset as well, that is why probably they do not publish until new data is available.They may be collecting more data before coming out with an official statement or a paper. I think it's clear that more RV measurements are needed to disentangle the existence of g from the eccentricity of d.
I mean he attributes it to Mayor: new analysis by astronomer Michel Mayor and his Swiss team suggests... But I've just found that wiki today has deleted the entry referencing Shostak, so hopefully it is just a misunderstanding from his side.I dunno. I have no idea what qualifications Seth Shostak has when it comes to extrasolar planet science. I don't know if he works in that field, or if it's entirely in SETI. I didn't see anything in that link worth taking to heart.
Originally Posted by BitulaSpeculation is never a good substitute when it takes little effort to look it up.Originally Posted by Bitula
I don't think anyone here does. All the HARPS team said is that they don't see the planet, and they have the sensitivity to do so if it existed. It's not a proof in any way. The HARPS team may have the signal of g hidden in the eccentricity of d. This uncertainty is why they are collecting more data. I'm sure if they were confident about its non-existence, they'd have a paper out or in the works.Originally Posted by Bitula
Really? Both papers (the two not discussing the planet's atmosphere) would tend to disagree.Originally Posted by Bitula
The HARPS team claim they don't see it. You claim they do and are hiding it.Originally Posted by Bitula
Not sure why you would guess that. Given the recent change in tone of the discussion, I'm more frustrated than anything, and a bit disgusted.Originally Posted by Bitula
Are you sure? Just asking...I don't think anyone here does.
That's not ALL you said. You also said (quoted): "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.". Otherwise I agree with everything else above.All the HARPS team said is that they don't see the planet, and they have the sensitivity to do so if it existed. It's not a proof in any way. The HARPS team may have the signal of g hidden in the eccentricity of d. This uncertainty is why they are collecting more data. I'm sure if they were confident about its non-existence, they'd have a paper out or in the works.
No please read my sentence carefully before arriving at a wrong conclusion. You probably refer to this:The HARPS team claim they don't see it. You claim they do and are hiding it.
"IMO g fits in their dataset as well, that is why probably they do not publish until new data is available."
Let me translate it:
1) G fits in their dataset (with a certain FAP)
2) From which follows that no one is lying.
I'm not referring here to the confirmation process but to the scenario described as " if they were confident about its non-existence". Now since we both think this is not the case we can close this whole issue here.Really? Both papers (the two not discussing the planet's atmosphere) would tend to disagree.
What? Btw I corrected myself before that, so what motivates you to emphasis my mistake? I could have also mentioned yours regarding "Hungarian HARPS"..., but it would have been childish.Not sure why you would guess that. Given the recent change in tone of the discussion, I'm more frustrated than anything, and a bit disgusted.
I have the right to speculate. There is nothing wrong about this claim. It's actually true unless you want to say that the size of the telescope does not matter... So is that what you think?Speculation is never a good substitute when it takes little effort to look it up.
To be sincere Trakar and Hungry, I'm only concerned with the probabilty of the existence of planet G and anything in connection with this planet given currently available scientifical information. I have no problems with science or scientist. I like science and I like scientists. If you don't have anything new to say in this topic or you do not have any personal oppinion please do not divert the topic into a general discussion about the meaning of science. Or feel free to do so, but don't address me with this, I'm not interested in this, I think I know what science is about and I don't want to discuss it here. All I care about is the topic of the thread: Gliese 581 g!
Yes.Originally Posted by Bitula
You seem to be grasping at straws or splitting hairs here. The part you quoted is effectively saying the same thing as what I wrote: "... and they have the sensitivity to do so if it existed."Originally Posted by Bitula
Are you aware that the HARPS team claim to not see g? Read this sentence carefully:Originally Posted by Bitula
HARPS team does not see g.
Now read your sentence carefully:
"IMO g fits in their dataset as well, that is why probably they do not publish until new data is available."
Do you see the disagreement here?
I already wrote what I think here.Originally Posted by Bitula
The difference is I am not relying on misinformation to support my argument. If the belonging of HARPS to Hungary was something that supported or went against my argument, I would indeed check my facts.Originally Posted by Bitula
I strongly encourage you to read Trakar's post. The way science works is coming up a lot because you seem to have a broken understanding of it. You expect the debate to be settled quickly, and that the HARPS team will immediately publish data, and what-not. It's going to be hard to keep science out of this discussion, especially with your pushing for the HARPS team to take a certain action.
If all you want to discuss is Gliese 581 g, this topic will shortly dry up. There's not a lot actually known about it aside from its period and minimum mass.
Look, to be sincere at least one of us has an elementary problem with basics of logic and interpretation of simple English sentences. Although I am not English, I would still like to point out to the same short statement in point 1:Are you aware that the HARPS team claim to not see g? Read this sentence carefully:
HARPS team does not see g.
1) IMO g fits in their dataset as well
Fitting g into the dataset. Fitting. You do not need to see it to fit it. I agree that they do not see it below a certain FAP. But they can fit it. Fit a theoretical G into their model. Fitting can be done by forcing a solution for example. So at least a forced solution must exist. You said that not even a forced solution can be done, that is why this is not just "grasping at straws or splitting hairs".
Well then let's wait until new info is available.If all you want to discuss is Gliese 581 g, this topic will shortly dry up. There's not a lot actually known about it aside from its period and minimum mass.
No, not again, please stop. This is not the problem here. The problem is that we do not understand what we are saying to each other. Just misinterpretation of simple sentences.The way science works is coming up a lot because you seem...
Have you actually worked with radial velocity data? Fitting curves to RV points to try and find planets out of the mess of points on a plot? I ask because I am not sure what this statement means: "You do not need to see it to fit it."Originally Posted by Bitula
The HARPS team would seem to disagree:Originally Posted by Bitula
Do you know what this means? (again, circling back to RV signals and residuals).Originally Posted by Dynamics of Cats blog
Taking a "better safe than sorry" approach, I will illustrate it here.
I have taken the available HARPS data set of HD 69830 and recovered the three known planets. I'm going with this system because it's quite honestly awesome, and I don't have the entire HARPS data set on Gliese 581 so I won't attempt to do it for that system.
The periodogram (that is, a plot of orbital periods on the x-axis, and how strongly they are found in the data on the y-axis), yields several interesting peaks.
Here are the most significant periods:
Note that several of them have really low FAP. Does HD 69830 have several of planets crammed at a period of ~1 d? Surely no. Alas, the FAP is only a measure of how easy it is to fit a signal at the given period, not a measure of how real a planet corresponding to that signal is. Several signals are aliases of other planets, or the sampling period. Since the star is observed only at night, a strong 1-day period shows up.
So after fitting the planets, we find this really pretty assemblage of RV curves and points. The bottom graphic shows the "residuals." These are a measure of the data point's deviation from the RV curve we have layed on the graph. If these residuals are fairly scattered, we might consider adding additional planets to lower the residuals.
Let's say someone claims a fourth planet with a period of about a year. Well let's look at the periodogram of the residuals we have, to see if there's another signal. After subtracing away the three planets, the periodogram looks like this:
And indeed, there are no obvious peaks. Note the high FAP on the highest peaks it could find.
This is what the HARPS team means when they say "We don't see the planet."
But let's try to force it in. We can fold the residuals to a one-year period. The top graphic shows just this, before we force in a planet. After forcing in a one-year signal to the data, we find such a signal now appears in the residuals. This is what the HARPS team is going on about regarding a negative signal. Alas, not only does this data set have the sensitivity to detect the fourth planet (which I've set at 0.07 MJ), but forcing it in produces this mark in the residuals. I could say "it doesn't exist." (the negative signal here appears as a curve to the data in the lower graphic)
Now I would go ahead and write a paper saying "I don't see it."
But Gliese 581 is much different. The signal of g is kinda mixed in with the signal of d. So HARPS's not seeing it or failure to force it in doesn't necessarily mean it's not there. They may be gathering more data to be confident that an eccentric orbit for d is a better fit than a circular orbit for d, and another planet, g.
You raise a good point. I consider myself fluent in English but not perfect. If I misinterpret something you say, I do sincerely apologise.Originally Posted by Bitula
Last edited by Hungry4info; 2010-Nov-05 at 02:19 PM. Reason: Elabourating, Grammar
Ok, I see. Again, I think in general there is no disagreement between us accept for some semantic or just mundane interpretation of for example this statement: "The HARPS team would seem to disagree". This statement cannot be true in a circumstance where the number of RV measurements are insufficient to favor an elliptic model over a circular one given their potentially incomplete dataset. So it is actually not that they disagree, but that they cannot make conclusions from their own dataset alone. So at most what can be said is that they favor their own theory over the other teams theory and nothing can be done before additional data is collected. But regarding Trakar, as I remember, he claimed that refining the method of analysis - for example refining the modeling process - alone can settle this dispute whereas I think that the method itself is fine and well established and additional measurements will be needed to settle the dispute unless probably if HARPS teem works with the combined dataset, but that is not good enough for a confirmation, because it is not independent. And best of all would be to refine the sensibility of current RV spectrographs by for example using larger telescopes (this is just an example), because it is still disputable that the number of measurements alone can unambiguously confirm G.
It must be realized that things would be much less emotionally "charged" if the initial announcement had taken more the form of "...analysis of data from the Gliese 581 system seems to favor the addition of two planets in short period orbits..." rather than "...Nearby Earthlike planet found! Lead researcher claims life 100% guaranteed!..." And I don't blame the astronomic teams for this as much as the abysmal quality of what passes for science reporting (and education). That said, however, all researchers dealing with the press need to understand how their words are going to be characterized and presented when they are talking to people unused to (or uncaring about) automatically adding in the appropriate qualifications, reservations and conditions as is more the situation when they are talking to peers and collegues. The worst I would accuse the discovery teams of is perhaps a bit of over-zealousness and the potential for some selection bias borne out of that zeal for what they are searching for,...hardly onerous or egregious offenses, in fact it is a rather common problem that is vigorously scrutinized and often one of the focuses of many peer review analyses.
Group think within a research project and the echo-chamber such creates can be subtle and insiduous in its effects and impacts and this applies to all teams. It may well be that the discovery team's choices were appropriate and accurate, and that the confirmation teams are making improper or incorrect assumptions in their reviews and studies, but it is all in the peer-review stage and within a few years I expect that it will be well studied, lead us to better understandings and provide us with some answers we don't currently possess.
I think everyone just got so excited about it, that they have been more excited from the idea than the actually data. It is pretty cool to think that on Gliese 581 g there might be some aliens arguing about if Earth exists or not right now. I think this analysis is interesting but it doesn't matter much if the planet is in the habitable zone since we have no way to get there or communicate with the planet. Even if we did send a signal it would surely break up from all the radiation and particles floating around in the universe.
Funny Facebook statuses
Last edited by apples; 2011-Oct-21 at 03:06 AM.
I'm a casual observer but an infrequent poster, so I can flirt with a modslap. I get more enjoyment out of reading than posting anyway.
I've really enjoyed the discussion between Trakar and Hungry4Info. Very informative - thank you!
On the issue of Bitula, however, between:
- threatening to leave the forum
- employing fallacious logic (credibility means people are more likely to believe you, not that you're right)
- exhibiting a lack of understanding in peer review
- asserting that a confirmational publication 'increases the probability' of something existing (it does no such thing)
- keeping score and acting like that has any relevence to the topic
- being overly defensive, and likely young (no great crime, but unfortunate)
- simply not understanding the science
...I find his/her "contribution" absolutely useless. The poster acts as a catalyst for continued discussion, but significantly lowers its quality. I'd recommend that serious contributors take Bitula for an opportunity to continue the discussion, but do not address the nonsense directly.
Thanks for the distraction! The quantity of fallacious insults with no other purpose than just to insult someone is quite a lot within a single post. If you want to be banned, I guess you are on the right way to achieve this. Btw, I'm not young. I will ignore you (what was your contribution so far btw?), hopefully others will ignore you as well.
Just found this link, which seems to be a good summary of relevant press releases as well:
I would like to note the following, which also supports my current understanding of the state of the issue:
1) The signal apparently shows up in the HIRES data alone.
2) The signal is not present in the totality of HARPS data.
3) As the complete HARPS analysis hasn't been published in detail yet, the HIRES community sticks with its positive result.
So IMO, if it were the case that the find is a mathematical artifact the paper with the detailed HARPS analysis should be released soon. And then at least a theoretical discussion can start. Btw, anyone played with the combined dataset? Or has any opinion based on amateur analysis? Is the original data available at full length? Anyway I personally think that G does exists. Would be nice to hear other's opinion. Just opinion, since I don't think that anyone is about to conduct hardcore scientific analysis here.
The answer to both of these have been established already in this thread.Originally Posted by Bitula
Personally, I don't think my opinion on the subject is relevant. This is science.Originally Posted by Bitula
I mean anyone of you.The answer to both of these have been established already in this thread.