How can we see the CMB?

[TooMany]

When mapping the CMB, the "foreground" is carefully removed. The local Galactic radiation should exceed the CMB in intensity (in the CMB part of the spectrum) because the Galactic sources are warmer and are also roughly black bodies. I think galaxies in the Local Group are also removed, but I'm not sure how deep this removal process goes.

More generally, if you examine the HUDF images, it is clear that the density of distant galaxies at an angular scale of say 15' is quite high. There may also be many more galaxies than those visible in these images at great distances. Since the angular resolution of WMAP is about the same (15'), any WMAP pixel contains contains radiation from many galaxies that is not explicitly removed from the CMB map (as far as I know). From z=0 to z=1, surface brightness of galaxies with the same luminosity is roughly constant. Thus, since galaxies are a source of substantial EM radiation within the CMB spectrum, why isn't the entire CMB contaminated by galaxies (not to mention gas clouds)?

Further, since galaxies are distributed densely over the entire sky, there is galactic radiation across the entire spectrum in any 15' sample. How do astronomers precisely isolate the CMB component from the galactic background in order to conclude that the CMB has a perfect black body spectrum?

It almost seems like you would have to find a 30' X 13.3 GLy "hole" in space to measure the CMB without contamination using WMAP.

Imagine you are in a snowstorm. Close by objects are not significantly obscured by the snowflakes because the individual flakes are spaced several inches apart, but distance objects become invisible as in a fog, all you see is flakes. How can we see the CMB through the fog of galaxies?

Sorry to include multiple questions, but they are focused on the same general issue.

[korjik]

Galaxies dont take up as much angular space as snowflakes do.

Also, the CMB isnt a perfect blackbody. If it were, it would not be the slightest bit interesting. There are certain things that can be done with the CMB tho. Once you take out the influence of the Milky way any multi-degree distortion can only be something that is in the CMB not a foreground effect, as it would require an invisible object the size of a supercluster to make a foreground distortion of that kind of scale. Before you wonder, yes they would check to make absolutely sure, to the limit of observation, that there isnt an invisible object out there.

That leads to the second point. Everything in cosmology today should be taken with the qualifier 'To the limits of observation'. We dont have really good views of most of the universe, so there may be some large error in cosmology somewhere. That is one of the reasons to actually study the CMB. If certain observations come up that dont make any sense under current theory, then new theory will have to be made to take the observation into account. Dark energy is a good example of that, it only being a bit more than a decade since it was discovered.

[TooMany]

Galaxies dont take up as much angular space as snowflakes do.

Perhaps they don't, but as you look deeper into the universe (assuming a uniform distribution) galaxies will occupy more and the more of the sky area until there is not much space left where a galaxy is not present. I believe the same mathematics applies as in the mean free path of a molecule in a gas. The mean free path (or in this case mean free view) depends directly on the product of the number density of galaxies per unit volume and the size of galaxies (at least in the case of a uniform random distribution).

Also, the CMB isnt a perfect blackbody. If it were, it would not be the slightest bit interesting.

My understanding is that it is claimed to be the most perfect black body ever observed (but the temperature varies very slightly in different directions).

There are certain things that can be done with the CMB tho. Once you take out the influence of the Milky way any multi-degree distortion can only be something that is in the CMB not a foreground effect, as it would require an invisible object the size of a supercluster to make a foreground distortion of that kind of scale.

But at least hundreds (and perhaps thousands) of galaxies populate every WMAP pixel-size area of sky. That doesn't matter? Why?

Before you wonder, yes they would check to make absolutely sure, to the limit of observation, that there isnt an invisible object out there.

This sounds like a claim of "holes in space" through which we may view the uncontaminated CMB. But there very likely are invisible and certainly visible objects in every WMAP pixel if the rest of the sky looks like the HUDF. I should add that the HUDF objects are only those that the camera is sensitive enough to detect. We know the ones we detect are actually there. Are they removed when measuring the CMB? I don't see how that would be possible without a HUDF of the entire sky.

I think I need a more quantitative analysis to understand how we can sort out the CMB from the 13.3 GLy deep galactic foregound to a degree that we can assert that the CMB is "the most perfect black body" every measured. I'm looking for this in the literature, but I just started and haven't found an answer yet. It seems as though after the Galaxy and perhaps the Local Group are removed, there is an assumption of insignificant contamination.

[parejkoj]

But at least hundreds (and perhaps thousands) of galaxies populate every WMAP pixel-size area of sky. That doesn't matter? Why?

Answer the following questions, and you'll have your answer:

What is the angular size of a Milky Way-type galaxy at z=0.1, z=1, and z=5? What is the redshifted flux in the 5 WMAP bands from a galaxy with a star formation rate of 0.1,1, 10, and 100 solar masses per year at each such redshift? What is the on-sky density of such galaxies at each such redshift? And finally, what is the total integrated flux in the WMAP beam from those galaxies at each such redshift, in each WMAP band?

Now, compare those numbers to the integrated flux from the CMB in the WMAP beam in each band.

You'll find that there's quite a difference between the pairs of numbers, even if you assume all the galaxies at, say, z=0.1 have a star formation rate >100 M_*/year. Which is a completely unrealistic assumption, but that's the only way to get the numbers to be even close.

[korjik]

Perhaps they don't, but as you look deeper into the universe (assuming a uniform distribution) galaxies will occupy more and the more of the sky area until there is not much space left where a galaxy is not present. I believe the same mathematics applies as in the mean free path of a molecule in a gas. The mean free path (or in this case mean free view) depends directly on the product of the number density of galaxies per unit volume and the size of galaxies (at least in the case of a uniform random distribution).



My understanding is that it is claimed to be the most perfect black body ever observed (but the temperature varies very slightly in different directions).



But at least hundreds (and perhaps thousands) of galaxies populate every WMAP pixel-size area of sky. That doesn't matter? Why?



This sounds like a claim of "holes in space" through which we may view the uncontaminated CMB. But there very likely are invisible and certainly visible objects in every WMAP pixel if the rest of the sky looks like the HUDF. I should add that the HUDF objects are only those that the camera is sensitive enough to detect. We know the ones we detect are actually there. Are they removed when measuring the CMB? I don't see how that would be possible without a HUDF of the entire sky.

I think I need a more quantitative analysis to understand how we can sort out the CMB from the 13.3 GLy deep galactic foregound to a degree that we can assert that the CMB is "the most perfect black body" every measured. I'm looking for this in the literature, but I just started and haven't found an answer yet. It seems as though after the Galaxy and perhaps the Local Group are removed, there is an assumption of insignificant contamination.

'Most perfect' and 'perfect' are two very different things. If the CMB were an actual perfect blackbody, then it would be pretty boring. It is the variations from that blackbody, most of which are very small, that make it interesting, because they are indications of structure in the very young universe.

Your assumption that the mean free path for a CMB photon is smaller than the observable universe is a bad one. Most paths dont actually hit anything. Also, a galaxy is going to have a different temp blackbody warmer than the CMB. It isnt all that hard to subtract off the superimposed signal. You are also missing the point that the variations from perfect blackbody that cant be attrubuted to low-z visible objects are indications of high-z invisible objects. Even then, like Parejkoj said, the amount of distorting signal is quite small compared to the CMB.

[Tensor]

I'm looking for this in the literature, but I just started and haven't found an answer yet.

Here , enjoy.

It seems as though after the Galaxy and perhaps the Local Group are removed, there is an assumption of insignificant contamination.

I think it's a bit highhanded to specifically state that you just started looking for an answer and then basically accuse the WMAP team of willfully ignoring contamination, by assuming any that is left, after removing the galaxy and local group, is insignificant. Shouldn't you wait until you've done more of a search before jumping in with an accusation like that?

[TooMany]

Here , enjoy.


I think it's a bit highhanded to specifically state that you just started looking for an answer and then basically accuse the WMAP team of willfully ignoring contamination, by assuming any that is left, after removing the galaxy and local group, is insignificant. Shouldn't you wait until you've done more of a search before jumping in with an accusation like that?

I have not accused anyone of anything. I resent your accusation. I've asked a question that I'm curious about and I'm trying to find an answer. I have carefully framed it in order to get a specific answer. So far I've gotten a homework assignment that might be beyond my ability to complete and a link that I'll study tomorrow.

I have a quick question for parejkoj. What is the presumed ratio of intensity of the CMB to the averaged intensity of galaxies and all other matter in the same part of the spectrum? (Excluding the Galaxy.) In other words, how far does the CMB outshine galaxies (in the appropriate part of the spectrum)?

[Tensor]

I have not accused anyone of anything.

Your specific statement:It seems as though after the Galaxy and perhaps the Local Group are removed, there is an assumption of insignificant contamination.

From your statement, it appears that you accept that the WMAP team takes out the Milky Way, may or may not take out the Local Group, and that's it. That you think they haven't considered anything else. What, exactly, leads you to that conclusion? Especially since you specifically stated you just started your search.

Those people worked extremely hard making sure there was as little contamination as possible. A couple of the data releases were delayed, because they were working to eliminate as much contamination as possible. And then you blithely toss off a line saying they remove the Milky Way, may remove the local group, and after that, they don't appear to consider anything else.

I resent your accusation.

You can resent it all you want, it still looks as if you are accusing them of ignoring anything other than the Milky Way, and possibly some nearby galaxies.

I've asked a question that I'm curious about and I'm trying to find an answer.

I might buy that if it was in the form of a question, instead of in the form of what appears to be an attempt to make it appear as if they didn't do their job. Some question like "what contamination was removed from the WMAP data"?

Anyway, this paper lists point sources that should be removed from the 5-year data release. There are 390 of them.

[korjik]

I have not accused anyone of anything. I resent your accusation. I've asked a question that I'm curious about and I'm trying to find an answer. I have carefully framed it in order to get a specific answer. So far I've gotten a homework assignment that might be beyond my ability to complete and a link that I'll study tomorrow.

I have a quick question for parejkoj. What is the presumed ratio of intensity of the CMB to the averaged intensity of galaxies and all other matter in the same part of the spectrum? (Excluding the Galaxy.) In other words, how far does the CMB outshine galaxies (in the appropriate part of the spectrum)?

You do realize that we can see pretty much all the way back to first light, and that there isnt an Olber's Paradox-type situation where all lines of sight end in a galaxy? You do realize that if that situation did occur that it would not be a 2.7K blackbody but a much higher temp one?

[Jeff Root]

'Most perfect' and 'perfect' are two very different things.
If the CMB were an actual perfect blackbody, then it
would be pretty boring. It is the variations from that
blackbody, most of which are very small, that make it
interesting, because they are indications of structure
in the very young universe.

As far as I know, no variation from a perfect blackbody
curve has been observed in the CMB. It is, within the
limits of measurement, perfect. The variations you
refer to are variations in temperature and intensity in
different directions, not deviations from a perfect
blackbody curve.

As far as I know. My knowledge is superficial.

But given all the contamination that must be there,
it does amaze me that it is possible to say that the
curve is so perfect.

-- Jeff, in Minneapolis

[Jeff Root]

My first thought was korjik's last comment. If the
sky were covered by light sources (stars and nebulae
in countless distant galaxies), it would be very bright.
Instead it is very dark. The vast majority of lines of
sight must not end on a star or nebula.

-- Jeff, in Minneapolis

[ngc3314]

Next hint: from a fair-sized nearby sample, we know what spectral components are important for galaxies in the far-IR and microwave regions. These components have spectral energy distributions which are distinct from each other and from the CMB blackbody. Synchrotron radiation s close to a power-law shape (flux scales with frequency to a power typically ~-0.7), reradiating interstellar dust is a (much warmer) blackbody modified by an emissivity function ~(wavelength)^-2 since the grains are much smaller than the relevant wavelength. Removing these components spectrally, pixel by pixel, is the reason that CMB experiments use multiple frequency/wavelength bands. Since the averaged surface brightness of these components is less than the troublesome parts of Milky Way emission, the same algorithms will work for both (as long as they incorporate the possibility of a significant redshift range for the dust contribution).

[Tensor]

Next hint: from a fair-sized nearby sample, we know what spectral components are important for galaxies in the far-IR and microwave regions. These components have spectral energy distributions which are distinct from each other and from the CMB blackbody.

It's just been the last three-four years that I learned how important the SED is and how it is used in various ways.

[parejkoj]

I have a quick question for parejkoj. What is the presumed ratio of intensity of the CMB to the averaged intensity of galaxies and all other matter in the same part of the spectrum? (Excluding the Galaxy.) In other words, how far does the CMB outshine galaxies (in the appropriate part of the spectrum)?

Work through the questions that I asked, and you'll know the answer!

If you have trouble working through those questions, post your questions and we'll help you.

[TooMany]

Your specific statement:It seems as though after the Galaxy and perhaps the Local Group are removed, there is an assumption of insignificant contamination.

From your statement, it appears that you accept that the WMAP team takes out the Milky Way, may or may not take out the Local Group, and that's it. That you think they haven't considered anything else. What, exactly, leads you to that conclusion? Especially since you specifically stated you just started your search.

I don't know exactly what they remove. That's in part why I asked my question. When I read papers that talk about removing foreground they seem to consider the Galaxy and some "point sources". I haven't seen much mention of other galaxies as yet. Now it may be that other galaxies don't matter because according to some calculations their contribution is insignificant. However I haven't seen the specific analysis that deals with this part of the issue.

parejkoj is suggesting that it is insignificant but asks me to do the quantitative evaluation. Well that's not an easy job for me and I was hoping to find some literature that discusses this in detail and explains how the rest of the galactic foreground is found insignificant or removed.

Those people worked extremely hard making sure there was as little contamination as possible. A couple of the data releases were delayed, because they were working to eliminate as much contamination as possible. And then you blithely toss off a line saying they remove the Milky Way, may remove the local group, and after that, they don't appear to consider anything else.

You can resent it all you want, it still looks as if you are accusing them of ignoring anything other than the Milky Way, and possibly some nearby galaxies. I might buy that if it was in the form of a question, instead of in the form of what appears to be an attempt to make it appear as if they didn't do their job. Some question like "what contamination was removed from the WMAP data"?

Please stop accusing me of things and either help answer the questions or stop posting on my thread. All I'm saying is that reading the literature so far, I see discussion everywhere of removing Galactic and "point source" contributions. I have no idea how the rest is handled.

What I'm trying to do here is understand why this radiation is considered to be a background behind everything else. Moreover, since the measurements are precise enough to show a nearly perfect black body spectrum, then at least intuitively any foreground must either be insignificant or removed with a high degree of precision. I'd like to understand how this is done for all galaxies for my own edification.

Anyway, this paper lists point sources that should be removed from the 5-year data release. There are 390 of them.

Yes I've seen such information. I believe that the number of point sources removed is now around 500 or possibly more. I don't even know yet what these point sources are. Are they stars in the Galaxy? Are they nearby galaxies? Are they AGN or quasars? Perhaps they just any bright compact source above some intensity level found in various surveys? The later certainly makes sense.

[TooMany]

Your assumption that the mean free path for a CMB photon is smaller than the observable universe is a bad one. Most paths dont actually hit anything.

That could be, based on additional assumptions/calculations about the density and luminosity of galaxies and the depth at which the CMB is generated.

Also, a galaxy is going to have a different temp blackbody warmer than the CMB. It isnt all that hard to subtract off the superimposed signal. You are also missing the point that the variations from perfect blackbody that cant be attrubuted to low-z visible objects are indications of high-z invisible objects. Even then, like Parejkoj said, the amount of distorting signal is quite small compared to the CMB.

Yes, and I'm asking just how small it is compared to the CMB. If the more distant galaxies are removed, how is that done, how much needs to be removed? You see I don't understand the big picture here. Is the CMB more luminous by far than all galaxies combined (in it's spectral range)? How much brighter is it? I'm also curious about how much total energy is in the CMB versus the rest of the sources in the universe (in all parts of the spectrum). I have don't have any grasp of these quantities.

[TooMany]

You do realize that we can see pretty much all the way back to first light, and that there isnt an Olber's Paradox-type situation where all lines of sight end in a galaxy? You do realize that if that situation did occur that it would not be a 2.7K blackbody but a much higher temp one?

No I didn't "know" that, but I guessed that it must be so and that is what my questions are about. I'd like to have some quantitative idea about these things. How much of the sky is covered by galaxies? How much power do these galaxies radiate in the CMB range? What happens at high red shifts? There, the density of galaxies must be much higher than locally given the higher density of the universe at high red shift, the recent evidence of early creation of galaxies and the merger theory of the origin of today's galaxies. I've read that these early galaxies have star formation rates 100 times current rates, but they are also much smaller. Are they more luminous that today's galaxies? Was there a peak luminosity of galaxies at some red shift? The light from early galaxies (say z>2) is red shifted so I suppose those galaxies potentially have a greater contribution to the overall galactic foreground in the CMB range than closer galaxies (of similar luminosity). This I conjecture because the peak energies are moved downward toward the CMB range by red shift. Does red shift distort a black body emission curve or does it preserve the black body curve but just shift it downward in temperature?

[TooMany]

Next hint: from a fair-sized nearby sample, we know what spectral components are important for galaxies in the far-IR and microwave regions. These components have spectral energy distributions which are distinct from each other and from the CMB blackbody. Synchrotron radiation s close to a power-law shape (flux scales with frequency to a power typically ~-0.7), reradiating interstellar dust is a (much warmer) blackbody modified by an emissivity function ~(wavelength)^-2 since the grains are much smaller than the relevant wavelength. Removing these components spectrally, pixel by pixel, is the reason that CMB experiments use multiple frequency/wavelength bands. Since the averaged surface brightness of these components is less than the troublesome parts of Milky Way emission, the same algorithms will work for both (as long as they incorporate the possibility of a significant redshift range for the dust contribution).

This is a reasonable qualitative explanation. My question is partly about how much non-CMB radiation must be removed. The less that needs to be removed the more accurately it can be done. How dependent is the perfection of the black body curve on the perfection of foreground removal? At the moment, I have no idea. The removal methods you mention seem to find the radiation patterns of galaxies independently of the CMB and then subtract it to find the CMB with very high accuracy. In other words there is an assumption that galaxies themselves are not CMB emitters and that we can estimate the spectral contributions from a galaxy (synchrotron, dust, molecular clouds and stars) accurately and subtract this from the observed sky radiation to isolate the CMB to a high level of purity. Does the CMB also shine through galaxies or is a substantial amount absorbed and re-emitted?

Recently some significant contamination has been discovered that seems to be emitted by spinning dust particles. I recently read a paper arguing that these dust particles must be magnetic in order to have such a large effect. They conjecture that this dust contains either elemental Fe or magnetic oxides of Fe. I believe these sources have been found in the Magellanic Clouds and perhaps in the galaxy as well. I don't know whether this new discovery has any serious significance in foreground removal.

This stuff is not simple, at least to me. The red shift issue also complicates the calculation, especially if there is a lot of dust that moves the spectral output towards the red in the first place.

[TooMany]

My first thought was korjik's last comment. If the
sky were covered by light sources (stars and nebulae
in countless distant galaxies), it would be very bright.
Instead it is very dark. The vast majority of lines of
sight must not end on a star or nebula.

-- Jeff, in Minneapolis

You certainly have a good qualitative point there. At least visually the sky is dark. You can see the Milky Way easily, Andromeda easily, the Magellanic Clouds and the Orion nebula for examples (under dark skies of course) . I don't know whether any more distance galaxies can be seen by the eye, but otherwise it is pretty black. However our vision drops out at about 600-700 mu which are far smaller the CMB wavelengths. What does the sky look like in the deep infrared? In recent papers I've seen much reference to the CIB (cosmic infrared background). It would be interesting to know whether the CIB is correlated with the CMB. It seems that it should have virtually no correlation since the CMB is an independent background of very uniform temperature. How does the energy in the CIB compare with the CMB?

[TooMany]

Work through the questions that I asked, and you'll know the answer!

If you have trouble working through those questions, post your questions and we'll help you.

It seems to me that this is a very complication calculation. We need to know average galactic densities, sizes, luminosities, spectral envelopes and corresponding red shifts and how all of these vary along the path to the origin of the CMB.

Going back to the snowflake analogy, I think I can conclude this much. The snow storm is like sparse flurries, one flake is right next to our eyeball, the next big flake that we can see in the entire sky (with our eyes) is about 3 million light years away. So indeed it seems that the flakes are very sparse, however to get a real quantitative idea one must consider many other things. How big, how bright and how dense are galaxies at each distance, how much does clustering affect the overall sky density, how much does increasing red shift and therewith increasing density of the universe affect the energy dumped into the CMB range and how does all of that integrate when looking through 13.3 GLy (about 4,000 times the distance to Andromeda). And what about low surface brightness galaxies that are just now being discovered. Do they have any significant contribution? Etc, etc.

No I cannot do this calculation without a lot of information (and also writing a program to do numeric integration of data points at various red shifts).

Now if you have some back of the envelope calculations that you care to present to argue for a very insignificant effect of the galaxies (so that all the mentioned details are of little consequence), I'd like to hear it. E.g. what percentage of the area of the sky is covered by galaxies (including the HI regions). Perhaps also some estimate of extra-galactic sources is needed as well such as gas clouds. Better still, if you know an internet source that analyzes this issue in some depth, I'd love to read it.

This reminds me of the last time you offered me help with understanding something about the SN 1a luminosity/distance relationship. You challenged others to calculate it in another thread. Did anyone succeed in that calculation?

I am not a graduate astrophysics student that you can assign this sort of the problem to and expect a reasonable answer. I'm asking you to be more direct in your answers.

[korjik]

It seems to me that this is a very complication calculation. We need to know average galactic densities, sizes, luminosities, spectral envelopes and corresponding red shifts and how all of these vary along the path to the origin of the CMB.

Going back to the snowflake analogy, I think I can conclude this much. The snow storm is like sparse flurries, one flake is right next to our eyeball, the next big flake that we can see in the entire sky (with our eyes) is about 3 million light years away. So indeed it seems that the flakes are very sparse, however to get a real quantitative idea one must consider many other things. How big, how bright and how dense are galaxies at each distance, how much does clustering affect the overall sky density, how much does increasing red shift and therewith increasing density of the universe affect the energy dumped into the CMB range and how does all of that integrate when looking through 13.3 GLy (about 4,000 times the distance to Andromeda). And what about low surface brightness galaxies that are just now being discovered. Do they have any significant contribution? Etc, etc.

No I cannot do this calculation without a lot of information (and also writing a program to do numeric integration of data points at various red shifts).

Now if you have some back of the envelope calculations that you care to present to argue for a very insignificant effect of the galaxies (so that all the mentioned details are of little consequence), I'd like to hear it. E.g. what percentage of the area of the sky is covered by galaxies (including the HI regions). Perhaps also some estimate of extra-galactic sources is needed as well such as gas clouds. Better still, if you know an internet source that analyzes this issue in some depth, I'd love to read it.

This reminds me of the last time you offered me help with understanding something about the SN 1a luminosity/distance relationship. You challenged others to calculate it in another thread. Did anyone succeed in that calculation?

I am not a graduate astrophysics student that you can assign this sort of the problem to and expect a reasonable answer. I'm asking you to be more direct in your answers.

so, summing up a bit, you dont know anything about the actual values, but you know that the people who do, the ones who do it for a living, must be doing it wrong?

[parejkoj]

It seems to me that this is a very complication calculation.

The calculation suggestions I gave you are pretty straightforward. Unless you want someone to just hand you an answer, you're going to have to try to work it out if you want to understand backgrounds in the CMB. I've been trying to help you acquire the tools to solve these problems yourself. Plenty of questions in astronomy can be solved (or, well-enough approximated) with some basic algebra, geometry, and a little estimation. Except for the redshift-distance relation (see below), this is one of them.

You can use Ned Wright's cosmology calculator to get the distances at each redshift (you'll need both the luminosity distance and the angular diameter distance). You can find the expected luminosity of a Milky Way type galaxy in the WMAP bands from an ADS search (that's the one value I don't know exactly off-hand though assuming it's about 0.1% of the the visual-band luminosity would be a very rough start). The on-sky densities of galaxies evolves with redshift, but you could start by assuming a space density of 0.01 per cubic megaparsec (from, e.g., Blanton et al. (2002)) at each redshift, and see where that gets you. Putting these bits together is then a little algebra and geometry, to get a back-of-the-envelope estimate.

Now if you have some back of the envelope calculations that you care to present to argue for a very insignificant effect of the galaxies (so that all the mentioned details are of little consequence), I'd like to hear it. E.g. what percentage of the area of the sky is covered by galaxies (including the HI regions).

Can you do some very basic geometry and algebra? I'd be happy to help if you get stuck with my above suggestions. In this and other threads you've implied that astronomers are either ignoring very obvious things or are not incorporating things that you think are important. At the same time, you've shown no inclination to do the work yourself, to find out whether these things are important or not. I've tried to point you toward the path to the solutions to your problems, which would help you learn a heck of a lot more than me just giving you some numbers. Which you might not believe anyway.

This reminds me of the last time you offered me help with understanding something about the SN 1a luminosity/distance relationship. You challenged others to calculate it in another thread. Did anyone succeed in that calculation?

So far, a biologist managed it with Mathematica, but that's sadly the only one. There are at least a few more people who have gotten part of the way there.

In the last post in the thread, I suggested just using Ned Wright's cosmology calculator to compute the redshift/distance relation. The disadvantage for that, in your case, is that it wouldn't calculate said relationship for the d_L=cz/H₀ cosmology that you were suggesting was also a good fit to the supernova data. That's a pretty simple cosmology, requiring no integral at all, so I was somewhat surprised that you didn't at least compare it to the data I provided.

I am not a graduate astrophysics student that you can assign this sort of the problem to and expect a reasonable answer. I'm asking you to be more direct in your answers.

Why do you care what my answers are? I don't know more about the properties of the CMB than the WMAP team, nor do I know more about the on-sky distribution of high-flux point sources at ~10mm than they do. If you want the best answers, read their papers. I'm just trying to give you some tools you can use.

[TooMany]

so, summing up a bit, you dont know anything about the actual values, but you know that the people who do, the ones who do it for a living, must be doing it wrong?

That's exactly right. I don't know anything about the actual values. Hence the questions. This is getting old between you and Tensor. I never said anybody is doing something wrong.

What I want to know is how significant that galactic foreground is, how it affects CMB measurements and how it is removed (if it even needs to be removed)! In my question I pointed out the issue of not just removal of Galactic sources and "point sources" but also of deeper sources. I tried to explain specifically the issue I was thinking about by describing it in some detail at the start of the thread. In particular I noted that there are many galaxies in any WMAP pixel. Either those sources are negligible (which seems unlikely) or they are somehow removed. I'm quite certain that this has been considered by astronomers, but I want to know more about it. I'm not asking whether they considered this. I'm asking how they quantitatively evaluate possible contamination from these sources, including some numbers.

Have you got that now? Please, if you cannot contribute scientific information relevant to the question, don't post to this thread.

[TooMany]

The calculation suggestions I gave you are pretty straightforward. Unless you want someone to just hand you an answer, you're going to have to try to work it out if you want to understand backgrounds in the CMB. I've been trying to help you acquire the tools to solve these problems yourself. Plenty of questions in astronomy can be solved (or, well-enough approximated) with some basic algebra, geometry, and a little estimation. Except for the redshift-distance relation (see below), this is one of them.

I can try. (I certainly would not mind if you wanted to offer some numbers directly.)

You can use Ned Wright's cosmology calculator to get the distances at each redshift (you'll need both the luminosity distance and the angular diameter distance).

I'm not sure what these two distances mean, but I can look it up.

You can find the expected luminosity of a Milky Way type galaxy in the WMAP bands from an ADS search (that's the one value I don't know exactly off-hand though assuming it's about 0.1% of the the visual-band luminosity would be a very rough start).

OK, that's an important value so I better find the correct one.

The on-sky densities of galaxies evolves with redshift, but you could start by assuming a space density of 0.01 per cubic megaparsec (from, e.g., Blanton et al. (2002)) at each redshift, and see where that gets you. Putting these bits together is then a little algebra and geometry, to get a back-of-the-envelope estimate.

Density of what is 0.01 per cubic megaparsec? There are galaxies of various sizes and luminosity. I'm not sure what the average size and luminosity are but it might be written somewhere in the world.

So far, a biologist managed it with Mathematica, but that's sadly the only one. There are at least a few more people who have gotten part of the way there.

Good for him! I don't happen have Mathematica or any knowledge of how to use it, but I wouldn't be surprised if it makes life much easier than graph paper, pencil and a calculator.

Why do you care what my answers are? I don't know more about the properties of the CMB than the WMAP team, nor do I know more about the on-sky distribution of high-flux point sources at ~10mm than they do. If you want the best answers, read their papers. I'm just trying to give you some tools you can use.

As I have already stated, so far the only analysis I've seen concerns Galactic sources and point sources. I haven't yet seen the analysis of the rest, perhaps because I'm spending too much time trying to get some answers here. I think I should continue in the literature.

I will try some very rough calculations, ignoring a lot of the details however.

[Reality Check]

That's exactly right. I don't know anything about the actual values. Hence the questions.

You seem to be asking us to do the research that you do not want to do, i.e. a list of all of the sources that are removed in the WMAP foreground removal process.
Why can you not find this list?
What are you going to do with list?

The need for processing WMAP raw data is obvious - to get the actual CMB rather than the CMB + contamination.
Max Tegmark's CMB data analysis center written in 1997 is still relevant.

[TooMany]

You seem to be asking us to do the research that you do not want to do, i.e. a list of all of the sources that are removed in the WMAP foreground removal process.
Why can you not find this list?
What are you going to do with list?

The need for processing WMAP raw data is obvious - to get the actual CMB rather than the CMB + contamination.
Max Tegmark's CMB data analysis center written in 1997 is still relevant.

If you don't want to answer a question don't post to this thread. If you just want to tell me to go find the answer in the literature then there is certainly no point in posting a question here, right?

If you bothered to read the original questions posed you would not be asking me if I found a list of point sources removed. I'm asking about the entire galactic foreground that might contaminate the CMB.

I'll check your link to see if this issue is discussed. Thanks for that.

[Jeff Root]

Reality Check,
korjik,
parejkoj,

Based just on what I read in this thread, TooMany is
trying to understand how we can distinguish the CMB
from foreground sources. He isn't interested in doing
cosmology himself. He is just trying to understand it.
There is a difference.

In particular, I don't see that TooMany asked for "a list
of all of the sources that are removed in the WMAP
foreground removal process." What he wants is an
understanding of what that list contains and how it
is generated. The methodolgy, not the dataset.

The paper that Tensor linked lists the locations of the
point sources, but does not say what those sources
actually are. He wants a characterization of what the
listed sources consist of, so he can get an idea of what
is being filtered out of the observations.

I'm not as interested in the question, but I'm interested
enough to follow the thread, and I'd like to see answers
to TooMany's questions.

His motivation is obviously the fact that disentangling
foreground sources from the CMBR seems an impossible
task. It seems impossible to me. So how is it done?

-- Jeff, in Minneapolis

[Reality Check]

I'm asking about the entire galactic foreground that might contaminate the CMB.

The "entire galactic foreground" is that of the Milky Way. It actually does contaminate the CMB since it contains many sources of relatively strong microwave radiation. This is easy to see in images without forground removal. This and other foreground sources (e.g. the point sources that you are aware of) are thus removed to expose the CMB.

ETA
A source for you: CMB Foregrounds

One of the major concerns in any Cosmic Microwave Background (CMB) anisotropy analysis is to determine which fraction of the observed signal is due to some foreground contaminant. Two sources of foreground contamination have been firmly identified: the diffuse Galactic emission and unresolved point sources. The diffuse Galactic emission includes three components: synchrotron and free-free radiation, which are important mainly at frequencies below 60 GHz, and thermal emission from dust particles, which is important mainly at frequencies above 60 GHz. From the theoretical point of view, it is possible to distinguish these three components by observing their different frequency dependence and spatial morphology. In practice, however, there is no emission component for which both the frequency dependence and spatial template are currently well known. Besides the above cited mechanisms, recent evidence suggests that there are additional physical components that can contaminate the CMB signal.

[Hetman]

Imagine you are in a snowstorm. Close by objects are not significantly obscured by the snowflakes because the individual flakes are spaced several inches apart, but distance objects become invisible as in a fog, all you see is flakes. How can we see the CMB through the fog of galaxies?

It is obvious that we do not see anything except that which was generated by the source and at the time: distance / c before, exactly.

Long ago, over a hundred years ago, various people have calculated directly from the Stefan-Boltzmann law: background of the Galaxy around 5K, and approximately the 3K from the outside.

However, the continuous reception of the signal emitted by the source only in one brief moment, it's a good idea for an invention in the style of perpetual motion.

[TooMany]

The "entire galactic foreground" is that of the Milky Way. It actually does contaminate the CMB since it contains many sources of relatively strong microwave radiation. This is easy to see in images without forground removal. This and other foreground sources (e.g. the point sources that you are aware of) are thus removed to expose the CMB.

ETA
A source for you: CMB Foregrounds

One of the major concerns in any Cosmic Microwave Background (CMB) anisotropy analysis is to determine which fraction of the observed signal is due to some foreground contaminant. Two sources of foreground contamination have been firmly identified: the diffuse Galactic emission and unresolved point sources. The diffuse Galactic emission includes three components: synchrotron and free-free radiation, which are important mainly at frequencies below 60 GHz, and thermal emission from dust particles, which is important mainly at frequencies above 60 GHz. From the theoretical point of view, it is possible to distinguish these three components by observing their different frequency dependence and spatial morphology. In practice, however, there is no emission component for which both the frequency dependence and spatial template are currently well known. Besides the above cited mechanisms, recent evidence suggests that there are additional physical components that can contaminate the CMB signal.

Our Galaxy's various sources are removed. I'm not sure what the unresolved point sources are. Does unresolved mean optically unresolved or unresolved at CMB wavelengths? I guess maybe I better find out specifically what these point sources are.

The sources that I am asking about are not the 500 or so bright point sources, but rather the millions of galaxies scattered over the sky.

This image is from the HUDF. The image is three arcminutes across or about 1/25 the area of a single WMAP pixel. Look at it carefully, it is said to capture about 10,000 galaxies, most of which cannot be seen in this low res reproduction. Thus HST detects 250,000 galaxies in each WMAP pixel. This image allows me to skip some calculations suggested by parejkoj and instead use direct evidence of the sky coverage by galaxies.

The image demonstrates that some significant percentage of the sky is covered with distant galaxies. Moreover the angular extent of radiation at longer wavelengths is at least double the optical size for each galaxy (this is due to the extended H I clouds that are found on the outskirts of galaxies). Because of this additional radiation, the actual percentage of the sky covered by galaxies is about 4 times as much as seen in this optical/near infrared image.

These galaxies are glowing with temperatures higher (on average) than the CMB. Within the area occupied by each galaxy (assuming that they are approximately black bodies), they should affect the CMB spectral range.

I used Planck's law of black-body radiation to determine that a (cool) star at 4000K outshines the CMB at 200GHz by a factor of 2000. Of course stars are actually quite sparse in a galaxy. Nevertheless a galaxy must have an overall temperature that is much higher than 2.7K over it's full area.

With this reasoning I conclude that there is substantial contamination in the spectral range of the CMB due to galaxies. In order to conclude that the CMB is a near perfect black body, this contamination must be very accurately removed. My question is how is that done? Can anyone suggest a reference that explains this?

Please explain if my reasoning about the need to remove this contamination is wrong.