PLANCK REFERENCE SKY VS WMAP DATA TEMPERATURE

INTRO:
The aim of this work is to investigate the limitation and eventually possible improvement of Planck's reference sky.
The Planck reference sky consist in  a simple model of the galactic foreground collected into a set of easy to use IDL routines, which can predict the sky brightness and polarization at any sky position and frequency. The diffuse galactic components model currently includes 4 components which are calculated by distinct IDL functions: Free-Free, Synchrotron, Dust,spinning dust.
For more information on that:

http://www.cesr.fr/~bernard/PSM/README.more


A better knowledge of foregrounds contamination will be the key factor to  improve our knowledge about cmb; in fact the foreground contamination is  estimated to be the main non-systematic sources of contamination for a cmb map.

We can't suppress this contamination because we are embedded in these galactic and extragalactic emissions then we need to characterize foregrounds contamination to be able to discriminate between such mix of foreground incoming radiations and the CMB background.
This estimation is very critical in polarization for the almost complete lack of data.

The Planck reference sky is the best that we can do with the available data but is still far to be enough. The main problem are:
  1. lack of data close to the cosmological window
  2. lack of data in polarization
The current philosophy for what concern the first item is:
This is exactly the case of synchrotron radiation.In this case, for example, we can't forget that the physical mechanism are very different in the radio band and in the microwave. The amplitude and the morphology will be both completely different.
About the second item the situation is even worse. We used to "invent" everything: a polarization angle, a fraction of polarization sometimes having some clue sometimes unfortunately not...

The WMAP data and the incoming Image WMAP polarization data will offer a good opportunity to test the reliability of Planck's reference sky.

The bands, frequency and beams for WMAP total intensity map are:
band   =    'k'         'ka'       'q'        'v'       'w'
v        =    23         33        41        61       94
beam  =  0.88      0.66     0.51     0.35     0.22             in deg

The first step will be test the power spectra of WMAP data versus the power spectra of Planck reference sky.
Here there is a detailed file about the UNITS used (and in which files) and also the conversion applied in order to obtain the same units for the simulated map and the WMAP survey map.

PROCEDURE #1 diffuse components all sky:
  1. create cmb+synchrotron dust and free-free map  according to Planck reference sky at WMAP frequency.
    [cmb+3 components * 5channel =20 map] sara's components maps
  2. co-adding the component maps [5 map ]
  3. calculate the power spectra of the five maps
  4. convolving with a gaussian beam to mimic the smearing due to the beam cl/(g_beam)^2 where g_beam is the window function in l space for a gaussian axisymmetric beam of given FWHM (provided by gaussbeam.pro)
  5. calculate the power spectra of the five WMAP maps
results in:
Feb 24 sara's plots
Feb 27 sara's plots

in the second set of plots I just added a line with cl*l(l+1)*gaussbeam using a flat power spectrum in cl*l(l+1)(roughly rescaled with the cl of wmap) just to take a look at the beam effect.

The Planck reference sky seems at very low respect the amplitude measured by WMAP survey.


PROCEDURE #2 diffuse components theta cut degree=5.0:

the item 1) 2) 3) are identical
  1. ''
  2. ''
  3. ''
  4. calculate the power spectra of the five maps cutting the galaxy +-5 deg
  5. calculate the power spectra of the five WMAP maps cutting the galaxy +-5 deg
results in:
Feb 28 sara's plots

How we can expect the results are even worse. In fact in general the foreground contamination is better understood into the galactic plane because in that area there are much more available surveys.



...What nice thing  I have to do tomorrow? Image

Include spinning dust, where it is?Image
Include sources Image

1,MARCH NOTE
PROCEDURE #3 diffuse components and sources:

I made the same procedure including also:
  1. galaxy clusters (Sunyaev Zel'dovich emission)
  2. infrared sources
  3. radio sources
results in:
March 2 sara's plots

It's clear the sources contamination have some effect at scale not so well depicted from an exeperiment with such resolution.
For example comparing all sky power spectrum k band with just diffuse foreground k log plot #1 and all sky power spectrum k band with diffuse foreground and sources k log plot #2.


ABOUT BEAM gaussian versus non gaussian beam:
The WMAP beams are hightly nongaussian Page et all. The WMAP team provide the window functions appropriate to each of the ten WMAP differencing assemblies have been derived in-flight from the calibrated time-ordered observations of Jupiter.
Making the average in frequency I obtained this plot .
I re-run the procedure #3 comparing the gaussian beam and the non-gaussian beam. Seems that this correction don't have an huge effect at interesting scale.
Btw the plots are collected here:
March 3 sara's plots


ANOTHER TIME ABOUT UNITS in previous plots there was an error:
In previous graphs I plotted WMAP data in mK thermodinamic themperature (instead the ref sky was in antenna themperature). This correction help a little bit in fact the correction will be ~1 at 22 GHz, and 0.66 at 90 GHz.
I'll show again some plots with power spectrum of diffuse foreground (just dust +free free + synchrotron) and point sources coming from ref sky convolved with gaussian and non gaussian beams VERSUS wmap power specrum, this time (I hope) with the right units.
I don't want rerun all the graphs at this point of my work but from know I'll plot everithing on the right units. Legenda: black dashed is wmap black solid is the ref sky and gray lines are the ref sky convolved with gaussian (up) and non-gaussian beam (down) the plots are collected here:
March 6 sara's plots


PROCEDURE #3 diffuse components and sources cutting the galaxy [+-5 degrees]:
the plots are collected here:
March 6 sara's plots (#2)


SPATIAL CORRELATION SCATTER PLOTS:
In order to study spatial correlation between the Planck reference sky and WMAP survey it will be useful make some scatter plots with Planck ref sky versus wmap maps. We hope to see some correlation.
First and brute steps is just to make a scatter plot between all sky maps with cmb+diffuse components [synch+ff+dust]+sources versus wmap the plots are collected here:
March 6 sara's plots (#3)
attention: files slow to open [.ps with 12*1024^2 points vs 12*1024^2 points]
Second step is to make a scatter plot highlighting the galaxy structure. I used a galactic patch of the same templates (theta cut +-5) the plots are collected here:
March 6 sara's plots (#4)
Another option will be study the spatial correlation between a reference sky component and a map of wmap. For example I'll expect some correlation between synchrotron reference sky map and wmap k band map . .
Here the scatter plot on the usual galaxy cut.
Idem for the dust reference sky and the wmap w band .
Here the scatter plot on the usual galaxy cut.

Today was not so proficuous: btw I run correlation and covariance of the scatter plot on the galactic cut.I tryed to calculate a fit but the usual IDL routines seem don't work very good. I already tryed linfit.pro and ladfit.pro. March 7 sara's plots


I rerun correlation and covariance of the scatter plot on the galactic cut. I made a linear fit (ladfit.pro) and I rescaled the reference sky with the inverse of the angular coefficient of the fit. I recalculated the parameter of the fit (just to be sure). In these plots you can see:
  1. scatter plot wmap versus rescaled ref_sky
  2. correlation
parameter of the linear fit (a,b -> wmap versus ref_sky) and (aa,bb -> wmap versus rescaled ref_sky; aa have to be 1.)
March 8 sara's plots


I run the usual procedure usin diffuse components +sources + cmb on the galactic cut. In these plots you can see:
  1. cl wmap (into the galactic plane +-5)
  2. cl ref_sky (into the galactic plane +-5)
  3. cl ref_sky convolved with wmap nongaussian beam (into the galactic plane +-5)


March 9 sara's plots
I summarized all this results in this much more clear web page: here

PLANCK REFERENCE SKY VS WMAP DATA POLARIZATION

INTRO:
In one day, hopefully, we will have the WMAP data release, and we will expect T and Q maps. Will be very interesting run similar comparison in polarization in order to understand how much the Planck Reference Sky will be relaiable in polarization.
I'm repeating the usual procedure including all the available templates in polarization.
This is the list of the components in polarization:

  1. CMB
  2. synchrotron
  3. dust
  4. infrared sources
  5. radio sources


I put the maps March 15 sara's plots

I put the power spectra [old window function]March 20 sara's plots
I put the new window function March 20 #2 sara's plots
I introduced the 3 year window function and I put more clear plots of power spectra March 20 #3 sara's plots
Attention!!! Each .ps contains 12 pages. TT EE BB TE for 3 different cut [all sky, in and out galactic plain (+_5)]
Attention!!! I updated these plots (bacause I was wrong with wmap_beam)
In the previous plot I show the ref sky convolved with the various WMAP beams; in this plots I show the opposite (WMAP cls decorrelated) March 22 sara's plots
*********************************************************************

Questi sono gli spettri nuovi, la convoluzione deconvoluzione per il beam dovrebbe essere giusta ho fatto i due casi: cl_ref_sky*(bb)^2 [K.ps Ka.ps....] e cl_wmap/(bb)^2 [de_K.ps de_Ka.ps ...]
NEW Power spectra
Ho messo anche i grafici zoommati l=[1,500]NEW Power spectra zoom
SCATTER PLOT IN POLscatter in pol
Per federico:per essere sicura delle unita' ho preso le "mie" mappe di wmap (convertite in antenna temperature) le ho tagliate con la maschera loro e ho confrontato il pws (moltiplico per 10^6 perche' io sono sempre in mK^2 loro in uK^2) nelle varie bande con quello dell'articolo di Page fig 17 e la mia
Mi sembra che ci siamo con le unita' no? cerco di pensare se posso aver fatto quanche casino con il reference sky...ma non credo [anche perche' in temperatura viene come ci aspettiamo]

Power spectra BB and BB modes Maik Wolleben survey [DRAO 1.4 GHz]...non era difficile qui!
*********************************************************************

Allora, qui c'e' un confronto interessante credo ma va raffinato. Nei grafici ci sono gli spettri di potenza di WMAP K band, WMAP K band deconvoluta per il beam, reference sky riscalato a 23.GHz e la Leiden survey riscalata con uno spectral index costante(=3) a 23 GHz. Non ho tenuto conto che le mappe sono circa 1/2 cielo. EE and BB

Ho tenuto conto che le mappe sono 41.7% del cielo, ho smoothato per visualizzare meglio. EE and BB

Non vorrei dire ma sembrano in buon accordo ancora una volta con WMAP mentre non c'entrano nulla col template di Giardino (assumendo che dust+cmb contino poco a 23 GHz). Questo ovviamente non rules out la spinning dust ...


sara