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

Estimating the number of galaxies

  1. Jan 24, 2013 #1
    The Hubble deep sky image is 3 x3 arc minutes and shows an estimated 10,000 galaxies. What percent is 3 x 3 arc minutes of the total area? If this ratio is known, one could estimate the number of total galaxies. Thanks
  2. jcsd
  3. Jan 24, 2013 #2


    User Avatar
    Science Advisor
    Homework Helper

    hi jim! :wink:
    the total area is 4π,

    and 3 arc minutes is π divided by … ? :smile:
  4. Jan 24, 2013 #3
    I know the area of a sphere is 4 x π x r2 . But what percent is the area of 3 x 3 arc minutes? When you look up at a dome, a square is not a square except at the horizon. There may be a simple answer but it is not obvious to me. Thanks
  5. Jan 24, 2013 #4


    User Avatar
    Science Advisor

    For small angles, you can approximate the sky sphere as flat. Just find the area of the square. Should be close enough.
  6. Jan 24, 2013 #5


    User Avatar
    Science Advisor
    Gold Member

    Compared to the celestial sphere, 3x3 arc minutes only comprises about one ten millionth of the surface area of the entire sky. All you need to do is calculate the surface area of the celestial sphere and compare it to 9 square arc minutes.
  7. Jan 25, 2013 #6
    Thanks, I was having a mental block. The result is 5.2 x E10 galaxies which is similar to other estimates assuming a finite universe.
  8. Jan 25, 2013 #7


    User Avatar
    Science Advisor

    Not assuming a finite universe. These are just measures of the visible galaxies. The total number, including the parts of the universe we can't observe, is almost certainly vastly larger. How much larger? We don't know.
  9. Jan 25, 2013 #8
    Good point, I should have said the "observable universe".
    Also, in the original calculations I erroneously inserted the diameter of the Moon rather than the radius. Thus, the correct number of galaxies extrapolated from the Hubble ultra deep field image is 1.56 x E11.
    My calculations are probably unorthodox ( a simpler way may exist). But, I used the Moon's radius to compute the Moon's area. The Moon's distance from the Earth was used to compute the surface area of a sphere. Dividing the larger area by the smaller area defined the number of Moons required to fill the entire sphere, 2 x E5.
    Addressing the ultra image, 3 x 3 arc minutes (9 square arc minutes) is 78 times less than the area of the Moon with a 15 arc minute radius (706 square arc minutes).The factor used to extrapolate from the 10,000 galaxies in the image was 1.56 x E7 (2 x E5 x 78) which gives 1.56 x E11 galaxies.
    I did the same calculation for the Hubble eXtreme deep field which is 2.3 x 2 arc minutes and has an estimated 5,500 galaxies. The answer was 1.67 x E11 galaxies - consistent results.
    The Hubble images support an estimate of 155-165 billion galaxies in the observable universe.
  10. Jan 26, 2013 #9


    User Avatar
    Science Advisor

    I'd just mention that this is a significant underestimate, as there are a great many galaxies that are too dim for Hubble to see.
  11. Jan 26, 2013 #10
    This wiki page seems useful:


    Nice plaque - shame it is completely wrong!
    I also just noticed that this wiki page says 80 Billion galaxies in the obervable Universe, so they must have a different definition for the size of the observable universe..

    To calculate the total number of galaxies in the observable universe, it should be possible to estimate the numbers of galaxies per unit volume
    and calculate the total volume of the Observable Universe based on the CMBR radius.

    On second thoughts we also need to consider that the first galaxies didnt form until around 400Million years after the CMBR so the light from the most distant galaxies cannot reach us despite the matter originally being within the CMBR radius when it was first emitted. Perhaps we need to define the what is meant by the observable universe, because there seems to be more than one possibility.
    Last edited: Jan 26, 2013
  12. Jan 26, 2013 #11
    The Wikipedia article references the 1996 Hubble image and in the original NASA writeup it says there are probably many more.
    My references are based on the 2003/4 images.
    The definition of observable universe is how far we can see, over 13 billion light years to the farthest galaxies.
    Do you still think the estimate is wrong?
  13. Jan 26, 2013 #12
    Thanks Jim, I had been interpreting the observable universe to be the CMBR radius because those atoms are, well, a part of the observable universe.
    In principle we should one day be able to observe galaxies a little older and more distant than the ones we see today, but I am not sure what the limit is.
    It is more complex than the statement in my previous message.
  14. Jan 26, 2013 #13


    User Avatar
    Science Advisor

    One way to think of it is to state the the observable universe is a slice in both space and time. That way the number of galaxies in the observable universe is very well-defined: it's the total number of galaxies that we can observe, neglecting the fact that we might not be able to see a few of them because they are hidden behind other galaxies (or hidden by our own).
  15. Jan 28, 2013 #14
    My unorthodox way of calculating the area is not necessary because the NASA bulletin describing the ultra deep field image stated that the oldest galaxy existed when the universe was 800 million years old. Thus, a radius can be calculated directly: 13.7 billion years - 800 million tears = 12.9 billion light years. This would be the distance to the faintest galaxies. The surface area of this sphere is 2.09 x E21 light years. Since the image was 3 x 3 arc minutes, the area of the image is 1.38 x E14 light years. Dividing the surface area by the image area gives a ratio of 1.52 x E7. This times 10,000 = 1.52 x E11 galaxies. As noted in previous posts, this assumes an observed universe defined at about 13 billion light years.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook