data from which wmap sky map is produced

[charles kaufman]

Is the data from which the sky maps are produced (for example figure 12 in
G.Hinshaw et al ApJS 180 225) publicly available and if yes, could you direct
me to it? That is the actual final numbers of delta T as a function of direction
that are color-coded and plotted, not the data at any earlier stage in the
analysis.
Thanks.
CK

[Ben Gold]

Fig 12 is the ILC map, downloadable from:
http://lambda.gsfc.nasa.gov/product/map/dr3/ilc_map_get.cfm

In general, there's a wide variety of data products available:
http://lambda.gsfc.nasa.gov/product/map/dr3/m_products.cfm

[charles kaufman]

Thank you.

Is that apparent from the paper?

I did look through http://lambda.gsfc.nasa.gov/product/map/dr3/

but couldn't tell that 'internal linear combination' was the one I wanted.

Thanks for the pointer.

CK

[Anze Slosar]

In general, you should read the papers carefully enough to know which one you want... ILC is unlikely to be wanted as its statistical properties are complicated and not very well understood. It makes nice pictures, though.

[charles kaufman]

Thank you for that warning.

My naive reading of the papers has led me to believe that
an appropriate transform of the data used for that picture would produce the
power vs l-value (spectrum) that shows three broad peaks. So I thought it
would be interesting to try it and see.

Thanks
CK

[Ben Gold]

Basically, the power spectrum of Fig 12 is the power spectrum of the maps after carefully taking into account (and "undoing") the beam and noise properties of the experiment. Unfortunately this isn't a trivial thing to do.

The ILC is a combination of smoothed maps, which means you'll really see only one peak as the higher resolution info just isn't there. Further, the peak probably won't be where you expected it to be unless you've undone the effects of the smoothing.

If you really do want to play with transforming maps into power spectra, what you could look at instead is a transform of the (unsmoothed) V or W-band data, with an appropriate sky cut to mask out foregrounds. The first peak should just jump right out because it's at large enough scales to not be affected so much by beam size and noise issues. Reconstructing the second and especially third peaks, however, requires a bit of effort to take into account the properties of the instrument.

[charles kaufman]

Basically, the power spectrum of Fig 12 is the power spectrum of the maps after carefully taking into account (and "undoing") the beam and noise properties of the experiment. Unfortunately this isn't a trivial thing to do.

That's what I was trying to get at in my original question. I knew there'd be a lot
of processing between the instrument and the pictures but thought there might
be a simple and direct connection between the picture and the spectrum.
Or that the highly processed data from which the spectrum comes would
be available.

Yes I do want to try to do a transform. So Ill try to learn about foregrounds
and masking.