Results 1 to 24 of 24

Thread: It's full of...not enough galaxies?

  1. #1
    Join Date
    Jun 2005
    Posts
    3,221

    It's full of...not enough galaxies?

    Recently, the BA presented a beautiful deep IR image taken with the VISTA telescope. He reports that the image shown (I looked at the complete 250 MB JPEG and it's mind-blowing) contains more than 200,000 galaxies yet covers only 0.004 % of the sky. Wow.

    Although...

    Let's assume the image is representative of the whole sky. Then when we divide 200,000 galaxies by 0.00004, we get 5 billion. I understand this to mean that IF this survey could be extended to the whole sky (never mind the Milky Way getting in the way) at constant image depth, the resulting whole sky image should show 5 billion galaxies.

    Now I seem to remember (and the Wikipedia article puts it in the same ballpark) that the observable Universe contains roughly 100 billion galaxies, and my question is - Where is the other 95 % of them?

    Is my math crap? Is my cosmology crap? Is the region of the sky that was imaged especially empty of galaxies? Or do we really only see 5 % of the galaxies (never mind dark matter or some such) that are there even if we take such beautiful, ultradeep images?

    Help appreciated...

  2. #2
    Two factors - I'm not sure this image has comparable depth to the Hubble Deep Fields or long Keck exposures, when the differences in filter bandwidth and sky background fold in. Second, estimates of the total number of galaxies may or may not include the extrapolation for galaxies too faint for us to observe now, which becomes a bigger and bigger issue in the distant Universe as larger fractions of galaxies fall below our detection threshold. (However, a quick check suggests that the number of galaxies actually detected in the HDF, catalogued at about 4000 in 0.002 square degree, extrapolates to about 80 billion over the whole sky, depending on how you count galaxies when they may contain separate clumps, so that last factor probably isn't germane to this estimate).

  3. #3
    Join Date
    Mar 2010
    Location
    United Kingdom
    Posts
    4,154
    Plus satellite galaxies - our own galaxy has something like 12 associated galaxies around it (LMC, SMC, Canis Major Dwarf, Sagittarius Dwarf Elliptical, Ursa Minor Dwarf, Sculptor Dwarf, Sextans Dwarf, Fornax Dwarf, Leo I Dwarf, Carina Dwarf, Draco Dwarf, and Leo II Dwarf - possibly more)

  4. #4
    Join Date
    Jul 2006
    Posts
    221

    Olber's Paradox

    Quote Originally Posted by ngc3314 View Post
    ...for galaxies too faint for us to observe now, which becomes a bigger and bigger issue in the distant Universe as larger fractions of galaxies fall below our detection threshold. ...
    Then isn't this an excellent argument that Olber's Paradox could be wrong and that the universe is infinite and eternally old - it is just that the light from these way distant stars/galaxies is too faint to light up the night sky? According to my understanding light gets red-shifted the longer it travels and thus the light fades into nothing beyond a certain distance so no star beyond that distance will ever be observable.

  5. #5
    Join Date
    Dec 2004
    Posts
    12,164
    The cosmic expansion causes light from distant galaxies to be
    redshifted. If there were no expansion, and overall curvature
    of the Universe were not an issue, and the Universe were
    infinite in extent, and infinitely old ... then although any one
    distant galaxy might be too dim to detect, the light from many
    such galaxies, all nearly in a line, would be visible as a small
    bright patch of the sky, and every part of the sky would be
    like that: Bright everywhere because the very distant stars
    in galaxies would be crowded together by near-alignment.

    Distant galaxies are only faint because their constituent stars
    have such a tiny angular area at that distance. Just increase
    the number of galaxies and stars to increase the area until
    it completely fills the sky solid with stars.

    Edit:

    Actually the Universe wouldn't need to be infinite in either
    size or age. Just very, very, very big and very, very, very old.

    -- Jeff, in Minneapolis
    http://www.FreeMars.org/jeff/

    "I find astronomy very interesting, but I wouldn't if I thought we
    were just going to sit here and look." -- "Van Rijn"

    "The other planets? Well, they just happen to be there, but the
    point of rockets is to explore them!" -- Kai Yeves

  6. #6
    Join Date
    Dec 2011
    Posts
    2,355
    The scale of this is truly awesome. However, purely in terms of number size, there are things closer to home, that dwarf even this !

    So, just for fun, consider this ...

    One of the smallest double strand DNA viruses is Hepatitis B, which has ~3,200 base pairs. Its DNA is composed of a sequence of 4 nitrogenous bases. Assuming these all occur in equal amounts (ie: 800 of each), then the number of possible ordered sequences in this DNA is: 3200!/(800!)4. (Although, only one of these results in what is known as Hepatitis B).

    So, 3200! = ~ 7.75x109828 and 800! = ~ 7.71x101976.

    Then there are: 3200!/(800!)4 ~ 1.075x101924, possible ordered sequences for this virus.
    ... Which pretty much dwarfs the ~9 x 1021 stars in the observable universe !!

    Just for fun, lets say the numbers of planets in the obs universe is say, ten orders of magnitude bigger than the number of stars, so say we have ~1032 planets.

    Whilst there is no physical significance in making this comparison, it could be said that:
    - on one planet out of ~1032 planets, we know the sequence for 'Hepatitis B' presently exists;
    - of all the 1.075x101924 possible base pair sequences for this chunk of DNA, one results in what we call 'Heptatitis B', a marginal life-form.

    So, as big as the observable universe is, there is an even bigger space of possible DNA sequences right under our noses !
    (Err... well, on second thoughts, in the case of Hep B .. let's hope its not under our noses ! ).

    Regards
    PS: Gee, I hope I got the calculations right !

  7. #7
    Join Date
    Mar 2004
    Posts
    13,441
    Quote Originally Posted by xylophobe View Post
    Then isn't this an excellent argument that Olber's Paradox could be wrong and that the universe is infinite and eternally old - it is just that the light from these way distant stars/galaxies is too faint to light up the night sky? According to my understanding light gets red-shifted the longer it travels and thus the light fades into nothing beyond a certain distance so no star beyond that distance will ever be observable.
    On its face, no.

    In a flat, Euclidean universe the surface brightness of galaxies is the same, no matter the distance; however, the apparent size falls with distance. So there will be one 'detection threshhold' at which galaxies 'disappear' ... because they become indistinguishable from stars. In our universe apparent surface brightness does depend upon distance.

    Here's a BAUT thread which looks at some of these things in more detail: Bias effects in galaxy detection

  8. #8
    Join Date
    Jan 2010
    Posts
    4,362
    Quote Originally Posted by Selfsim View Post
    One of the smallest double strand DNA viruses is Hepatitis B, which has ~3,200 base pairs. Its DNA is composed of a sequence of 4 nitrogenous bases. Assuming these all occur in equal amounts (ie: 800 of each), then the number of possible ordered sequences in this DNA is: 3200!/(800!)4. (Although, only one of these results in what is known as Hepatitis B).
    Is there a specific reason to assume they occur in equal amounts? If that assumption isn't made the calculation becomes a lot easier, simply possible ordered sequences of length 3200 with 4 base pairs.

  9. #9
    Join Date
    Jan 2010
    Posts
    509
    Quote Originally Posted by xylophobe View Post
    Then isn't this an excellent argument that Olber's Paradox could be wrong
    For all i know Olbers' paradox was based on the assumption that the universe is static.

    So Olber's Paradox has been wrong (rather than being a paradox) ever since it was found that the universe is not static.

  10. #10
    Join Date
    Dec 2011
    Posts
    2,355
    Quote Originally Posted by caveman1917 View Post
    Is there a specific reason to assume they occur in equal amounts?
    (Chuckle, chuckle) … well, if there's a hard way to do things, I seem to always find myself taking that path, eh ?


    Thanks for the query, Caveman. I wish I could say that I originally had a good reason for assuming equal amounts, but I'll have to admit it was purely a guess!

    Following your query however, I looked it up. The answer is not as easy as I had hoped for. It seems the measure of determining the ratios of nitrogenous bases in the DNA molecule, is the ‘GC Content’. This is defined as being the percentage of nitrogenous bases on a DNA molecule, that are either Guanine or Cytosine, (hence the ‘GC’ term). From what I can see, (and I’m happy to be corrected on this), several sources say the GC ratio of Hepatitis B is 43%, which suggests it is within 7% of being a 50/50 mix of G-C/A-T bases. (Which would seem to partially answer the question).

    As far as other organisms are concerned, it seems that the GC ratio can vary widely, and there has been a lot of effort to classify various organisms into groupings based on their GC percentages. There has also been a lot of statistical/distribution analysis done on G-C content, within the coding and non-coding parts of particular DNA.

    I’ll attach some references below, (just for the record)1..

    I’d have to say, at the end of the day, at the highest level, your result is at least, calculable (and displayable) on web-based calculators. My result on the other hand, is not calculable (most display result in infinity). Either way, both results are way bigger than my estimate of numbers of planets !

    Regards

    1. References:

    VIRsiRNAdb;
    Identification of Hep B in the Human Genome;
    Structural differences between Coding and Noncoding DNA;
    Organism classification according to two codon patterns.



  11. #11
    Join Date
    Jan 2010
    Posts
    4,362
    Quote Originally Posted by Selfsim View Post
    (Chuckle, chuckle) … well, if there's a hard way to do things, I seem to always find myself taking that path, eh ?
    Which in the end also means an advantage in getting practice with it

    Would perhaps a better calculation be to take the length of the coding dna sequence rather than the total length? Because of all the noncoding dna, the total length seems to be a bit immaterial to this. Then there's the issue that not all random coding dna sequences will "make sense" (so a fraction of them should be disregarded), but this seems to be much harder to factor in.

  12. #12
    Join Date
    Dec 2011
    Posts
    2,355
    Quote Originally Posted by caveman1917 View Post
    Which in the end also means an advantage in getting practice with it
    Yes … I've learnt this week that unfortunately, some might perceive such things as 'woo'. But this is definitely not the intention. I found the exercise kind of interesting. I find it gives us a feeling for our relationship with a vast universe. The calculation is what causes that feeling … and for me, that feeling is real .. even if the relationship is tenuous and, perhaps, unphysical.

    Quote Originally Posted by caveman1917
    Would perhaps a better calculation be to take the length of the coding dna sequence rather than the total length? Because of all the noncoding dna, the total length seems to be a bit immaterial to this. Then there's the issue that not all random coding dna sequences will "make sense" (so a fraction of them should be disregarded), but this seems to be much harder to factor in.
    Now that would be complicating it lots more . The 'Structural Differences …' reference in my previous post, contains the base information to perform such a calculation. However, somewhere, somehow, perhaps, on one of the 1033 exo-planets, that DNA might code, resulting in something
    Somewhere in our future, some microbiologist might also find that these sequences do code on Earth, under certain circumstances …

    Perhaps we should call these thoughts 'Extra Dimensions' ?

    Regards

  13. #13
    Join Date
    Jun 2004
    Posts
    6,011
    ... and you's think it is I that is mad.. I rest my case.. and am enjoying the points made.. good fun..;

    and am a little concerned as I seem to find agreement.. this can not be real.. I must need some sleep.

    I have been spending hours on the comparison of images.. searching for little dots..

    Using the 'Go-to' to a galaxy and it follows I am working in two dimensions and the focus being the third...

    and at any mention of 'Extra' Despite the millions of possible micro possible string structures of 'Hep c..'

    It's blown into not the last word in comparison to cosmology.. the minute does not match the infinite..

    Oh dear me, oh my.. must find sleep...

  14. #14
    Join Date
    Dec 2011
    Posts
    2,355
    Quote Originally Posted by astromark View Post
    ... and you's think it is I that is mad.. I rest my case.. and am enjoying the points made.. good fun..;

    and am a little concerned as I seem to find agreement.. this can not be real.. I must need some sleep.
    Oh, I understand the irrationality of it all.
    Feelings are real .. but only to the individual … not so for the community.
    And all mine were expressed without the need to rubbish anything 'communal'.

    And the feeling was made real via the mathematics .. reality was precipitated by doing the maths, created by the community .. I didn't just pull it out of thin air !

    Quote Originally Posted by astromark
    It's blown into not the last word in comparison to cosmology.. the minute does not match the infinite..
    No .. the minute infinity exceeds the macro infinite universe in size .. but how do I know this ? (Check out Georg Cantor and Hilbert's paradox of the Grand Hotel … and they were both great mathematicians !)

    Regards

  15. #15
    Join Date
    Jun 2005
    Posts
    3,221
    I'd like to thank ngc 3314and Shaula for their prompt and precise answers to my questions.

    Shaula - The point you bring up is probably very nontrivial. If the galaxies in the HUDF compute too about 80 bn galaxies in the entire observable universe, then the little and dim satellites you mention are probably not even counted, as even in the HUDF, a small satellite of, say a galaxie 2 gpc away is probably far beyond the detection limit (isn't it?).

  16. #16
    Join Date
    Mar 2010
    Location
    United Kingdom
    Posts
    4,154
    Quote Originally Posted by Arneb View Post
    Shaula - The point you bring up is probably very nontrivial. If the galaxies in the HUDF compute too about 80 bn galaxies in the entire observable universe, then the little and dim satellites you mention are probably not even counted, as even in the HUDF, a small satellite of, say a galaxie 2 gpc away is probably far beyond the detection limit (isn't it?).
    It all depends on how they have estimated that number of 100 billion. If could be worked out from all-sky surveys or it could be done by looking at the local group and using the characteristics of it to try to get an idea how many galaxies there are (or more accurately estimate their size/frequency distribution so we can estimate how many galaxies we are missing for each bright one we see). Trouble is local statistics won't be characteristic of distant (far in the past) galactic clusters, surveys will always miss dim ones.

    But trying to answer your other question - we have issues seeing these things from inside our own galaxy thanks to geometric effects. Combine geometry and distance and you have a recipe for missing more than you spot! Especially when the distance creeps into gigparsecs.

  17. #17
    Join Date
    Jun 2005
    Posts
    3,221
    OTOH, shouldn't we assume that if we look at galaxies at cosmological distances, the Universe should be isotropic at those distances, and hence, it shouldn't matter that 30 % of the sky are blocked by our own galaxy?

  18. #18
    Join Date
    Mar 2010
    Location
    United Kingdom
    Posts
    4,154
    By geometric effects I just meant that the satellites could be behind or too close to the galaxies to be spotted, even if they were bright enough to be seen. Unlike binary star systems we don't really have the luxury of watching them rotate around each other!

  19. #19
    Join Date
    Jun 2005
    Posts
    3,221
    Very true!

  20. #20
    Join Date
    Dec 2004
    Posts
    12,164
    If you are trying to estimate the number of galaxies in the
    visible Universe, then it seems to me you need to:

    - Estimate the number of galaxies per unit volume from
    whatever regions nearby that we happen to be able to see,

    - Notice that every large region in the visible Universe
    appears to be pretty similar to those nearby -- or would be,
    if we could see what they look like "now" (which of course
    we can't), and

    - Figure out how big the visible Universe is.

    So it doesn't matter what we can or can't see in
    distant parts of the visible Universe, right?

    -- Jeff, in Minneapolis
    http://www.FreeMars.org/jeff/

    "I find astronomy very interesting, but I wouldn't if I thought we
    were just going to sit here and look." -- "Van Rijn"

    "The other planets? Well, they just happen to be there, but the
    point of rockets is to explore them!" -- Kai Yeves

  21. #21
    Join Date
    Mar 2010
    Location
    United Kingdom
    Posts
    4,154
    What matters is that the further we go back the worse our local galaxy distribution will fit. Galaxies evolve, merge and so on. So our estimates based on everything being similar to our locale will get less and less accurate as we increase the distance. And even within our local cluster our sample used to estimate this distribution is heavily biased by our probability of detection. Several of our own galaxy's satellites were only found by careful mapping of stellar velocities and types - something we simply cannot do once we get much further out. We could have 26 satellites for our own galaxy, we may have found 21 for Andromeda, Triangulum may have one satellite or may be a satellite of Andromeda, NGC3109 may have one, the other close galaxies are mostly isolated or in pairs - there is a strong link between distance and number of satellites which seems likely to be related to detectability rather than anything else.

    Plus the distribution (spatial this time) of galaxies is highly non-uniform - they are broken into all sort of filaments and walls and so on. So we need to have some guideline as to how much of the universe is empty and how much is full of galaxies - which makes what we can see very relevant.

    In essence one of the ways we can do this is to estimate a distribution of brightness/frequency from local conditions, then estimate what we should be able to see at various distances. Then where we can see galaxies we use these two facts to estimate a number. The two critical approximations we are making are
    1) Local distributions approximate distant ones (we know this is false to some extent because of the aforementioned evolution of galactic clusters)
    2) We have an adequate map of intervening material that lets us estimate what we should see very well (a stronger assumption - we have spectroscopic techniques that help out here)

    As I sort of said before - how you handle the approximations (i.e. do you use cosmological models to try to 'fix' the galaxy distributions for each time interval, do you assume a certain volume is representative of everything out there) and how you make the overall estimate is really critical to what number you get out. I probably would not trust it to be more than within an order of magnitude or so at best.

  22. #22
    Join Date
    Sep 2003
    Posts
    10,347
    Quote Originally Posted by Arneb View Post
    Is my math crap? Is my cosmology crap? Is the region of the sky that was imaged especially empty of galaxies? Or do we really only see 5 % of the galaxies (never mind dark matter or some such) that are there even if we take such beautiful, ultradeep images? Help appreciated...
    I'm glad you asked, and your math is not fishy (carp-like). [I've been claiming 131 billion was the best answer, and I think it might be the 125 billion. I suppose someone could actually count the no. of stars in the image assumed to have 10,000 galaxies. Maybe they have by now.]

    Here's a thread from 2009 when the HUDF info came upon us....

    I think the HUDF (Hubble Ultra Deep Field) is the famous million second exposure taken over the course of several months.

    As I recall, the HDF came first and with great success. This was followed by a southern version to see if it made sense to make extrapolations. The south gave similar galaxy counts.

    I didn't check my old link to the camera sensor specs., but it looks like 125 billion may be a reasonable estimate of galaxies, though this will increase since this is not the earliest period for galaxy formation for the visible universe, exluding all regions beyond.
    We know time flies, we just can't see its wings.

  23. #23
    Join Date
    Dec 2004
    Posts
    12,164
    Shaula,

    It seems to me that what we can't see isn't important,
    because there is so much that we *can* see. If we can't
    see small galaxies hidden behind our own, it doesn't
    matter because there should be about equal numbers
    on the side we can see and the side we can't see.
    If we can't distinguish small galaxies from neighboring
    big galaxies, it doesn't matter because the small ones
    are included in the big ones.

    Maybe I just don't get the significance of the number
    of galaxies. It seems completely unimportant, like trying
    to determine the number of rocks on Earth or the number
    of clouds in Earth's atmosphere at a given moment.

    -- Jeff, in Minneapolis
    http://www.FreeMars.org/jeff/

    "I find astronomy very interesting, but I wouldn't if I thought we
    were just going to sit here and look." -- "Van Rijn"

    "The other planets? Well, they just happen to be there, but the
    point of rockets is to explore them!" -- Kai Yeves

  24. #24
    Join Date
    Mar 2010
    Location
    United Kingdom
    Posts
    4,154
    Assuming symmetry is another assumption to add to the mix - it seems to be broken for Andromeda. So we end up generalising the properties of elliptical and a range of spiral sizes/types from incomplete data taken from two fairly similar spirals that have partially characterised and not hugely similar distributions of satellites!

    I'd agree with you wholeheartedly that the actual number of galaxies is not very important and subject to huge uncertainties for all the reasons laid out above and more. Your cloud analogy is exactly right - it is like trying to estimate the number of clouds worldwide over a year from observations taken in one city at in Summer!

Similar Threads

  1. VISTA View Is Chock Full Of Galaxies
    By Fraser in forum Universe Today
    Replies: 0
    Last Post: 2012-Mar-22, 05:20 PM
  2. Replies: 3
    Last Post: 2011-Jan-13, 01:09 PM
  3. First Views of ISS at Full Length, Full Power
    By Fraser in forum Universe Today
    Replies: 1
    Last Post: 2009-Mar-26, 11:00 AM
  4. Full full tanks and empty bellies
    By banquo's_bumble_puppy in forum Science and Technology
    Replies: 121
    Last Post: 2006-Sep-05, 07:16 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
here
The forum is sponsored in-part by: