##
Welcome to the QMASK data page

###
Shop here for maps and covariance matrices

**Figure 1***: The QMAP and Saskatoon experiments zoomed in on a
bird-shaped region near the North Celestial Pole with a higher
angular resolution than COBE DMR. If you want a ps file
with this image, click on it.*

For how the data were analyzed,
please refer to the the QMASK paper (nickname for QMAP and SASKATOON) that was
submitted to PRD:
Xu et al.
You can also get the details about QMAP experiment from three
QMAP papers.
You'll find additional information on
the QMAP sites at
Penn and
Princeton and on
Angelica's mapmaking page. The details about SASK experiment can be gotten
by
Netterfield et
al .
The analyzed science data sets are now
public and you can download them below.
## Raw maps

As described on the above-mentioned paper, the multi-gigabyte
data set from both QMAP and SASK was reduced to a combined sky map. Here it is:
- qmask_cmbmap.dat (QMAP+SASK)
- index.dat shows where the pixels are in the
square sky map. It is not needed for sceintific analysis of
the maps, but is useful for plotting. We devided the sky region, including the
observed part, into 256x256 pixels. The 1st column tells the pixel sequence
number. The 2nd one tells if those pixels are observed or not. If it isn't, then
give a value 0; If it is, then give the sequence number in the file
qmask_cmbmap.dat.

In the file qmask_cmbmap.dat, each row corresponds to a pixel on the sky.
The five columns are x, y, z, T and sigma,
where (x,y,z) is a unit vector in the direction of the pixel
in equatorial coordinates (see the figure below),
T is the measured temperature fluctuation and sigma is the
rms noise. T and sigma are in units of micro-Kelvin.
In other words, the vector x in the papers corresponds to the 4th column. You
may notice that some values (noise level) in the 5th column are pretty big, or
in other words, those pixels looks pretty "noisy". But they are unignorable
because they are correlated noise from the deconvolution technique as described
in paper so that it's important that they use the noise covariance matrix.
# Noise covariance matrices

As described in the papers, the complete data product needed for a cosmological
analysis is the temperature vector and the noise covariance matrix.
These matrices are also public. If you want them, you can download
qmask_Sigma.raw, **WARNING!!!!**
*Please click your ***RIGHT** mouse button and then choose "Save link
as...", ** NEVER** use the left button!!!. or
email
xuyz@hep.upenn.edu
to get an ftp directory.
Here is a piece of
fortran code that reads these covariance matrices
and explains the format in which they are stored.
The 5th column in a map file above is merely the square root of the
diagonal elements in the corresponding noise covariance matrix.
# Wiener-filtered maps

Since some pixels are much noisier than others and there are
noise correlations between pixels, Wiener filtered maps
are more useful than the raw ones for visual inspection.
The images above and below are both Wiener filtered.
Here they are:
- qmask_wiener.dat (Wiener-filtered
vector Xw)
- W.raw (Xw=WX, X is the raw vector).
**WARNING!!!!**
*Please click your ***RIGHT** mouse button and then choose "Save link
as...", ** NEVER** use the left button!!!.
- Sw.raw and Nw.raw
(Sw=WSW^t;Nw=WNW^t).
**WARNING!!!!**
*Please click your ***RIGHT** mouse button and then choose "Save link
as...", ** NEVER** use the left button!!!.

The pixels are in the same order as for the raw maps above.
qmask_wiener.dat contains
only one column, which shows the temperature (in unit micro-Kelvin) after
Wiener-filtering.

**Figure 2***: Wiener-filtered map of the QMASK combined data.
The CMB temperature is shown in coordinates where the NCP is near
the center, with RA being zero at the top and increasing clockwise.
If you want a clearer ps file
with this image, click on it.*
# QMASK Power spectrum

Angular power spectrum from QMASK (l~30-200) is available now. Some points have
got the smallest errorbars until April 30th, 2001 (BOOMERanG and DASI will
release their new results). The figure is shown in Fig. 3, and the data are
available in
Xu et
al

. The corresponding window
functions is shown in Fig. 4.

**Figure 3***: Power spectrum from the QMASK combined data.
If you want a clearer ps file
with this image, click on it.*

**Figure 4***: Window functions for QMASK power spectrum. The first
four curves is corresponding to the four points in Fig. 3.
If you want a clearer ps file
with this image, click on it.*

Return to my home page

This page was last modified Apr. 27th, 2001.
xuyz@hep.upenn.edu