Established Member
About the formation of Filamentary and Large Scale Structures of the Universe ?
That one based on magnetic field and gravity ?
http://adsabs.harvard.edu/cgi-bin/np...13B&db_key=AST
Or that gravitational (only) drived model based on dark matter ?
http://arxiv.org/abs/astro-ph/0504097
Administrator
This isn't really a Q&A topic. If you already have a formed opinion, you should be taking this someplace else, such as Astronomy if you want a mainstream discussion, or ATM if you have something you want to try out.
Personally I don't like your rhetorical style that implies that exactly one of these is exactly right. Odds are there's some truth in both papers, and neither is perfectly right.
Forming opinions as we speak
Established Member
Ok for a transfer to Astronomy.Originally Posted by antoniseb
Administrator
I have moved it, and left a redirect that will expire in 24 hours.
Forming opinions as we speak
Moderator
IMHO it's not and either/or question, it is just two different approaches, both of which will probably not get the completely correct large scale structures. Most likely both processes will work at the same time and a joint model will have to be created. It's sort of like the solar wind, it is not that we do not know how that is generated, it is that there are too many processes that can generate it and they have to be combined.
All comments made in red are moderator comments. Please, read the rules of the forum here and read the additional rules for ATM, and for conspiracy theories. If you think a post is inappropriate, don't comment on it in thread but report it using the /!\ button in the lower left corner of each message. But most of all, have fun!
Bi-weekly space physics research "blog" at tusenfem.blogspot.co.at
Established Member
Battaner in part 4 claim that CDM based model(s) in (1982) ,(1997) predicted a random distribution of Galaxies on very large scale. This is contradicted by observations which show rather a considerable regularity (Battaner's citation)...
See introduction chapter for details
Part 4
http://adsabs.harvard.edu/abs/1998A%26A...338..383B
Established Member
I wouldn't call Battaner's work mainstream at all. Certainly, it is not ever discussed in modern studies of the matter evolution of the Universe.
Springel et al.'s work on the Millennium Simulation is very close to the observed properties of the matter distribution. The differences can be mostly ascribed to their choices of initial parameters being slightly different.
Established Member
It seem that Battaner model (1997-98)is in reply to the failure of earlier CDM model(s) predictions made in (1982)and (1997) to match observation.
http://adsabs.harvard.edu/cgi-bin/np...13B&db_key=AST
see reference in post 6.
Yep ! But there is a trick...
If you look at Page 19 you will see that the parameters they used for making their model were specifically selected to match surveys results ..."our parameter adopted values are consistent with a combined analysis of the 2dFGRS surveys and first WMAP data".(Citation)
http://arxiv.org/abs/astro-ph/0504097
Established Member
The trick is that they selected the "cosmological parameters of our LCDM-simulation" that were "consistent with a combined analysis of the 2dFGRS surveys and first WMAP data". They want to match the results of their simulation with the actual observations (e.g. the 2dFGRS surveys) so this selection is not surprising.
Established Member
No ,that is not surprising at all. However how they can pretend that they are making predictions?.Are they talking about the evolution of the large scale structures in n billions years?
Because as pointed out in post 6
CDM based model(s) made in (1982) ,(1997) predicted a random distribution of Galaxies on very large scale. This (was) contradicted by observations which show rather a considerable regularity
Established Member
You mean... They are using evidence to refine their model?! What a disgrace!If you look at Page 19 you will see that the parameters they used for making their model were specifically selected to match surveys results ..."our parameter adopted values are consistent with a combined analysis of the 2dFGRS surveys and first WMAP data".(Citation)
Generally with these sorts of models you can count it as a success if you get more out than you put it. If you can explain a lot of things you see with a small number of tuned parameters in a model then it is a useful one. Just look at the Standard Model.
BTW - saying that a 30 year old model made in the early history of LCDM modelling got the answers wrong is not a good argument against modern models.
Established Member
But the point is that they claim that they are making predictions with their model, they don't.
Can you point about these new -'lot of things'- that this model can explain which was not already know at the moment of the simulation ?
Established Member
My point about the lot of things was that predictions are hard in astrophysical terms. We have made a lot of observations and we do not have the luxury of being able to wait a billion years and see if we are right. A predictive success for models like this can include it explaining something you did not give it as a starting condition. So (as a random example) if you just gave it a few parameters relating to density and temperature, tuned it to match the CMBR and it produced galaxies then that would be a predictive success. The model did not know about galaxies but it produced them.
Sometimes you get lucky and a model predicts something you have not looked for - but in general the data has already been collected, now we need models to tie it together.
Established Member
You are in agreement that their so called "prediction" in the introduction chapter relate in fact to the description of the initial conditions in the early Universe which finally leaded to the formation of large scale structures...
Wrong the model was adjusted all along to match galaxies specific characteristics.
see page 7
..."the modelling assumption and parameters are adjusted by trials and errors in order to fit the observed properties of low redshifts galaxies"
Moderator
Ehhh, DUH! That is how scientific research works. You make assumptions about the "initial parameters" of the universe, then see whether the result is in agreement with what we see, if not, then adjust the parameters, as those are basically the unknowns of the universe.
What is it that you expect? Just throw in random stuff and if that does not work, throw away your computer model?
Naturally you want the magnetic model to be the "correct" one, taking your background into cosideration, but like I said before this is not an either/or question, because nature does not make it that simple. You may want to throw in an incorrect paper from 1982, but that argument is laughable, because as said before, that is at the beginnig stage of this kind of science, and that paper can hardly be used as an argument.
All comments made in red are moderator comments. Please, read the rules of the forum here and read the additional rules for ATM, and for conspiracy theories. If you think a post is inappropriate, don't comment on it in thread but report it using the /!\ button in the lower left corner of each message. But most of all, have fun!
Bi-weekly space physics research "blog" at tusenfem.blogspot.co.at
Established Member
Please read my posts more carefully. The bit you bolded is being interpreted out of context.
That is what I said. It was being used as an example, not as a specific. It was showing what predictive power could mean in astrophysical models. It was not a commentary on the model or paper presented.So (as a random example) if you just gave it a few parameters relating to density and temperature, tuned it to match the CMBR and it produced galaxies then that would be a predictive success. The model did not know about galaxies but it produced them.
Established Member
There was no Lambda-CDM model in the 80s: the dark matter models of the time were a variety of warm, hot, and cold dark matter, with several possibilities for the values of omega_M and omega_b. It turns out that if you don't have the correct values of the initial density parameters, you don't get the matter evolution of the Universe correct. Who knew?
Established Member
Right i have understand what you proposed however your specific claim sounds like you are saying that they have only put few basic parameters and let run the model from that point which is totally wrong. see post 19
This is your specicfic claim
So (as a random example) if you just gave it a few parameters relating to density and temperature, tuned it to match the CMBR and it produced galaxies then that would be a predictive success. The model did not know about galaxies but it produced them.
ETA
Can you point out to the predictive power (a prediction) this model as made about something who was not already known at the time of the simulation?
Last edited by Don J; 2012-May-10 at 07:06 PM. Reason: fix a quote
Established Member
The CDM model made in 1997 was not better.
The present CDM model looks more like a staged movie.See page 7 about the specific algorithms used for galaxies formation for example.
http://arxiv.org/abs/astro-ph/0504097
Last edited by Don J; 2012-May-10 at 07:06 PM. Reason: Typos
Established Member
Actually, that's exactly what they did. Why do you think it isn't? If there is a part of the paper you are confused about, perhaps you could quote it and we could discuss it.
Their semi-analytical model for galaxy formation is a model for how galaxies occupy dark matter halos, and was not tuned to match observations. They did not follow the formation of individual stars, gas and dust into galaxies (that's impossible given current computing capabilities), but rather parametrized our understanding of that process with a simpler procedure (the "semi-analytic model"), and applied that procedure to the dark matter distribution.
Established Member
Already covered this. Not going to keep repeating myself.Can you point out to the predictive power (a prediction) this model as made about something who was not already known at the time of the simulation?
Established Member
Your answer in
Post 13... I presume ....
Established Member
I was replying to the specific claim made by Shaula in post 13...
You are confirming that Shaula is wrong when he claim that The model did not know about galaxies but it produced them._if you just gave it a few parameters relating to density and temperature, tuned it to match the CMBR and it produced galaxies then that would be a predictive success. The model did not know about galaxies but it produced them.
Established Member
And I have already covered that. It was not a specific claim it was a demonstrative example of, hypothetically, how a model could be shown to have predictive power if it was shown to match observations not used to 'train' it.You are confirming that Shaula is wrong when he claim that The model did not know about galaxies but it produced them.
Is it really that hard to understand? You were questioning how an astrophysical model could be predictive, I was explaining. I was not referring to any particular model.
Established Member
Thanks, for the clarification...!No, that is not really hard to understand that this model as no predictive power because that contrary to your demonstrative example the present model was trained to match observations. see post 19
http://arxiv.org/abs/astro-ph/0504097
Established Member
Yes they they are making predictions!.
The evolution of the large scale structures in billions of years is an output of the model. It is not part of the input. It is a prediction of the model.
Which has nothing to do with Springel et al.'s work on the Millennium Simulation which was done in 2005.
Established Member
He is confirming what anyone who knows about the simulation knows: That the model did not know about galaxies but it produced them. and
My emphasis added.To track the formation of galaxies and quasars in the simulation, we implement a semianalytic model to follow gas, star and supermassive black hole processes within the merger history trees of dark matter halos and their substructures (see Supplementary Information). The trees contain a total of about 800 million nodes, each corresponding to a dark matter subhalo and its associated galaxies. This methodology allows us to test, during postprocessing, many different phenomenological treatments of gas cooling, star formation, AGN growth,
feedback, chemical enrichment, etc.
Each node in the simulation contained multiple galaxies. So they used other models to simulat the galaxies for the simple reason that the nodes did not go down to the atoms in the universe!
Established Member
Right ,but as I pointed in the OP there is another mainstream model based on magnetic field which predict the same thing.
here
http://adsabs.harvard.edu/cgi-bin/np...13B&db_key=AST
Wrong, their CDM model known about Galaxies formation
As i pointed out numerous time...
See page 7of the paper about the specific -semi-analytic model- used by the model to account for galaxies formation and the adjustments by trial and error they were forced to made in their CDM model to match observations.
http://arxiv.org/abs/astro-ph/0504097
Established Member
So, I feel like I need to clear up some confusion on the part of several people here about what Springel et al. (and most modern cosmological simulations) did.
A semi-analytic model does not "produce galaxies" in the sense of generating them ab initio from the gravitational collapse of gas, but rather assigns galaxies to dark matter halos following some prescription. The overall matter distribution is set by the cosmological parameters and however they generate the initial Gaussian random field. The small to medium scale galaxy clustering (correlation function or power spectrum) at a given redshift is thus determined by both how they assign galaxies to halos and the clustering of the dark matter particles (and thus halos) at that redshift. Because of the nature of the gravitational collapse of structure, the clustering of matter on large scales (~>70Mpc) is also determined by both, because of effects like the Baryon Acoustic Oscillation (BAO) which shows up at scales around 150Mpc.
So, a simulation like the Millennium run makes direct predictions about both the small scale and large scale clustering of galaxies of various types (whatever types the semi-analytic model produces) throughout the range of redshift that the simulation covers. Those predictions can be verified or rejected by observations of those types of galaxies at the appropriate redshift. By providing initial cosmological parameters that match the results of, e.g., the WMAP data, we can test whether our understanding of cosmology from WMAP matches with what we can observe from the large scale distribution of galaxies (it does). And by the same token, by running a suite of simulations with different cosmological parameters, we can see what other choices of cosmological parameters also produce a similar large scale galaxy distribution (very few).
Established Member
That is not a mainstream model, and that paper does not provide any comparison with observations. In addition, the understanding of the large scale structure of galaxies was quite limited during the 90s, so any such comparison (though again, that paper provides none) would have to be revisited with more recent data from e.g. SDSS or 2dF.
Posting Permissions
Reply With Quote