Data reduction for Zone of Avoidance (ZOA) survey with ALFA - LiveData

P. Henning, 13 July 2005 (update 17 March 2006)

The data reduction is very similar to that done for the AGES survey,
which has nice documentation at www.naic.edu/~ages

We use the packages LiveData and Gridzilla, which were originally
written for the Parkes multibeam surveys, and which have been
modified to handle ALFA data.  LiveData does the system temperature
calibration, bandpass calibration, corrects to heliocentric frame,
applies a spectral smoothing, and writes out the data in sdfits
format, ready for gridding.  (Note 3/17/2006: right now, the cubes are 
produced with an indication that the velocity scale isn't quite right, with
error on the scale of the Doppler correction.  We may need to run
LiveData again once this is understood.)

Log in as "zoa" (for password, check with Trish, Arun, Emmanuel, Barbara, or
Chris).  
For record keeping, I have been making an area to collect the sdfits files 
for each observing day's data.  Thus, for the commensal project a2004 mapping
the Taurus region, I write the sdfits files to an area called, for instance:

/share/zoa/zoa_a2004/20050710 

This collects the sdfits files created by LiveData from day 20050710.

Before running LiveData, you need to create such an area (mkdir) 
under, eg. the zoa_a2056 directory for the relevant day of data you are 
about to calibrate.

For instance, observations on 23 April 2006 should go into a directory you
create under the existing zoa_a2056 directory, called:

/share/zoa/zoa_a2056/20060423

Do this for each day of observing.

I've been running LiveData on one of the cluster machines aoxcN,
where N = 1 to 4, depending on which is least busy.  To start, type 
"livedata".

Now, consult the webpage for AGES LiveData.  Use the AUDS configuration.
Set the read directory to find the right data:  /share/wappdata
File wildcard would be set to the *project ID*.fits.  Set the write
directory to the directory you've created for the sdfits files.  Set the 
bandpass calibration to do a hanning smoothing, and use bandpass calibration
parameters "extended", and "median".  The Max cycles and Box size depends 
of the length of time for each scan.  Set them long enough to get a loop
of basketweave.  For instance, each loop of the Taurus map of a2004 
takes 8.5 minutes times 2, plus cals, plus some overhead, equals 1028 sec.
I have set the Max cycles and Box size to 1080 to be sure nothing is
missed.  

For project a2056, a smooth observing session has about 18 basketweave 
loops, each known as a "lambda".  Each part of a loop takes 3.1 minutes, 
thus a full loop is 6.2 minutes plus overhead, including calibration.  We have
been using 450 sec for box size and max cycles in livedata.

Set everything else the same as for AGES.  Do set Spectral baseline
fit to Robust (adaptive) linear fit.  To run livedata, select the files
you want with the mouse, hit "queue selection", then start.  I do like
to monitor the output, using "multibeam view".  You can also monitor various
statistics by clicking on "statistics".  All this is documented via the
AGES data reduction website.  When you've successfully LiveData'd, update
the reduction log, for instance for a2056, update the log:

/share/zoa/zoa_a2056/a2056_reduction_log.txt  

When in doubt about any of this, ask Trish, Chris, Barbara, or Emmanuel!

When all the data are successfully through LiveData, you are ready to
grid into cubes with Gridzilla, which is described in document "Data 
reduction for Zone of Avoidance (ZOA) survey with ALFA - Gridzilla".
If in doubt about what should be gridded, check with Trish.