CMB Power Spectrum and Angular Power Spectrum Plot from WMAP

[TrickyDicky]

Anyone knows if the CMB map of anisotropies from WMAP is used to implement the angular power spectrum plot(acoustic peaks)? I'm not sure, but I tend to think it is not.

 

[Chalnoth]

Anyone knows if the CMB map of anisotropies from WMAP is used to implement the angular power spectrum plot(acoustic peaks)? I'm not sure, but I tend to think it is not.

No. The way it's done is they typically just use an aggressive mask to get rid of the galaxy and point sources, and estimate the power spectrum based upon that. So far as I am aware, foreground removal is not yet part of the power spectrum estimation strategy from WMAP, largely because there is currently no good way of estimating the uncertainties in foreground removal. There's also the point to be made that foreground removal isn't really necessary for WMAP, except in the sense of removing the galaxy and point sources, because WMAP isn't sensitive enough for the foregrounds to become an issue.

With Planck, however, due to its increased sensitivity, foreground removal will be absolutely necessary for obtaining optimal science data from the results.

 

[TrickyDicky]

No. The way it's done is they typically just use an aggressive mask to get rid of the galaxy and point sources, and estimate the power spectrum based upon that. So far as I am aware, foreground removal is not yet part of the power spectrum estimation strategy from WMAP, largely because there is currently no good way of estimating the uncertainties in foreground removal.

There's also the point to be made that foreground removal isn't really necessary for WMAP, except in the sense of removing the galaxy and point sources, because WMAP isn't sensitive enough for the foregrounds to become an issue.

With Planck, however, due to its increased sensitivity, foreground removal will be absolutely necessary for obtaining optimal science data from the results.

Hmmm I thought using a mask to get rid of galaxy emisions and point sources , and foreground removal were basically the same thing, is this not so?
What other foreground is there to be removed?

Thanks for answering

 

[Chalnoth]

Hmmm I thought using a mask to get rid of galaxy emisions and point sources , and foreground removal were basically the same thing, is this not so?

What I was referring to there is more often called component separation, which involves using the multi-frequency data in order to estimate the CMB itself. This map that the WMAP team released, for example, is produced using one particular algorithm for doing this, dubbed Internal Linear Combination (ILC):
http://lambda.gsfc.nasa.gov/product/map/current/map_images/ilc_7yr_gal_moll_1024.png

The basic issue here is that there are things other than the CMB no matter where you look in the sky, but those things tend to have different frequency scaling, so you can pick out which part of the signal is CMB, and which part isn't. It's not a trivial process, and there are multiple techniques of doing it, all with their own strengths and weaknesses. The primary difficulty right now is that proper estimation of errors isn't quite there yet.

 

[TrickyDicky]

I see, it all looks quite complex, I guess they have to be really careful with this foreground cleaning, I would suppose there has to be a fine line between being too agresive masking what is not CMB, with the risk of leaving just some kind of random noise, and not masking enough leaving fluctuations in the map that are not really CMB.