Tanelorn said:
In your description, why would the microphones all measure different signals?
In WMAP, the reason is that they get different amounts of the signal from the galactic plane. The reason they get different amounts of the galactic signal is that the galactic signal is
not a black-body spectrum at a temperature of 2.725K.
In the microphone analogy, this is similar to there being another person in the room, Fred, who is also talking, sometimes much louder than George, but Fred is not situated at the same location, so some microphones pick up Fred's voice louder than other microphones. Thus we can use a linear combination of the microphones to cancel out Fred's voice while keeping George's. As long as Fred isn't moving around, this is a relatively simple operation.
Tanelorn said:
Anyway, I think I will have to move on from this question until I can get a more detailed picture of the actual RF signals entering each of the difference receivers. There is plenty of other sky to consider.
Well, understanding the RF signals themselves isn't all that likely to give you an understanding of how the WMAP team went back and subtracted the galaxy from the maps to get the CMB.
But the basic idea is that each detector is a radiometer which detects the amount of radiation impacting the telescope from a particular direction within a small range of frequencies. The central frequencies of the detectors are 23GHz, 33GHz, 41GHz, 61GHz, and 94GHz. The instrument subtracts the signal from a detector looking at one part of the sky from the value of a corresponding detector (which measures the sky at the same frequency) looking at another part of the sky. As WMAP orbits the Sun, WMAP spins around, scanning the sky so that it is always pointed away from the Sun. Every six months, WMAP scans the whole sky once (well, almost...some holes are left, but these are covered in the next six months).
From the time-ordered data, the WMAP team produces maps of the whole sky at each frequency. These are not maps of absolute amount of radiation hitting the telescope at each frequency, but of the deviation from the average amount of radiation hitting the telescope.
Tanelorn said:
It appears that with the center of the frequency peak at 160GHz so we would prefer to measure the signal there, but this is not easy or as accurate with today's technology. eg. The highest frequency I have worked on is 18GHz. What method was used to determine this frequency peak?
This was done earlier with the COBE satellite's FIRAS instrument, which measured the spectrum of the CMB to tremendous accuracy:
http://lambda.gsfc.nasa.gov/product/cobe/cobe_images/firas_spectrum.jpg
On this plot, the error bars are smaller than the trend line at every point. Not just slightly smaller, but mind-bogglingly, absurdly smaller. The measurement error at the peak is around 0.0035%. All but the last three data points have an error less than 1%.
Tanelorn said:
The WMAP is looking for tiny differences in RF amplitude, but shouldn't we be measuring the frequency of the frequency peak to determine the actual CBR temperature?
Well, this was done by FIRAS to tremendous accuracy, and so isn't really that interesting to scientists any longer. We want to know the small deviations in temperature across the sky, and in order to distinguish between the CMB and other sources, we need to look at the sky at multiple different frequencies.
Planck, by the way, should measure the CMB near the peak. It has detectors that look at the sky in nine different frequency bands from 30GHz to 857GHz. The 30GHz-70GHz detectors are radiometers, like WMAP (but actively cooled instead of using the differencing strategy). The 100GHz-857GHz detectors are bolometers, which instead of being antennas are designed so that energy deposited in the bolometer increases the temperature of the bolometer slightly, leading to a change in electrical resistance that can be measured. I believe the frequency of the radiation that strikes each detector is singled out by constructing a horn which only permits the passage of radiation within a specific wave band.
Tanelorn said:
Has there been any resolution of the following: Since CBR is received from the edge of the observable universe, and we are at the center of our observable universe, why is there a large red shift on the CBR? How does the red shift affect the WMAP measurements?
The redshift comes from the expansion of the universe. The universe has expanded by a factor of about 1090 since the emission of the CMB, which has multiplied the wavelength of each photon emitted at that time by a factor of 1090.
Tanelorn said:
Also what is the average attenuation of the CMB from after traveling 13.8 Billion light years? What is the spread on this attenuation and how does this affect the WMAP measurements?
About 8% of the radiation is lost in transit. This can be measured due to the correlation between the polarized signal and the unpolarized signal, as the ionized gas in between us and the CMB is sensitive to the polarization of the light.