How to use your single-dish observations?

Our algorithm to produce the short-spacing information is coded in the UV_SHORT task. This task have about 25 input parameters which are not always intuitive to fill in (in particular the relative weight between single dish and interferometric data). We thus wrote a SIC procedure which wraps up most of the task input parameters according to the context. This procedure is driven by the widget called short-spacings processing in the main MAPPING menu. Although you have the possibility to access all the control parameters of the UV_SHORT task through this widget, you should have only the file names and the single-dish data unit to fill in, all other parameters being guessed by the procedure.

This task starts from the data format produced by the CLASS TABLE command. Basically, this is a GDF table containing one line per spectrum, the columns representing the lambda offset, beta offset, weight, and the spectrum intensities. This format is subject to change: Please, refer to the TABLE documentation for up-to-date information. At least two characteristics of this table must be tuned on the characteristics of the interferometric observations:

The velocity axis

because all the computations are done plane by plane without velocity resampling.

The center of projection

because all the computations are done using the offset to this center.

To do this, the user needs to work both in MAPPING and CLASS.

Inside MAPPING,

both informations can be found with the following commands

       1 read uv field-1
       2 uv_map
       3 write dirty field-1
       4 header field-1.lmv

Comments:

Steps 1-3

Images the first field of a mosaic because the information about the velocity axis will be much easier to dig up in the dirty cube field-1.lmv than in the $uv$ table field-1.uvt.

Step 4

Displays the header as follow:

       1  File : field-1.lmv                                        REAL*4
       2  Size      Reference Pixel        Value                  Increment
       3     256   129.0000000000000   0.000000000000000   -1.5514037841057871E-06
       4     256   129.0000000000000   0.000000000000000    1.5514037841057871E-06
       5      49   25.50000000000000   10.50000000000000    0.2000000029802322
       6       1   0.000000000000000   0.000000000000000    0.000000000000000
       7 ...
       8  Axis 1     A0       05:40:54.270        Axis 2     D0      -02:28:00.00
       9 ...

Line 8

Indicates that the axes 1 and 2 are space coordinates around the phase center $\alpha_{2000} = 05:40:54.270$, $\beta_{2000} =-02:28:00.00$.

Lines 3-6

Describes the axes and in particular the velocity axis as line 5, i.e. 49 channels of 0.2 km/s width, the velocity of the 25.5 channel being 10.5 km/s.

Inside, CLASS

you need to type the following commands to align the characteristics of your single-dish data to the above values

       1 file in  single-dish-reduced.30m
       2 file out single-dish-reprojected.30m single
       3 set level 5
       4 set system equatorial 2000.00
       5 find
       6 for ientry 1 to found
       7    get next
       8    modify position 05:40:54.270 -02:28:00.00
       9    write
      10 next ientry 
      11 file in single-dish-reprojected.30m
      12 find
      13 table short-spacings new /resample 49 25.50 10.50 0.2000000029802322 V
      14 xy_map short-spacings
      15 header short-spacings.lmv

Steps 1-10

Will ensure that the used data is in Equatorial 2000 coordinate system (The default for PdBI data) with the right projection center. Steps 1 and 2 opens the input and output files. Step 5 decreases the verbosity of CLASS: This is useful when dealing with OTF maps with a large number of spectra. Step 4 enforces the use of the 2000 equatorial coordinate system. Steps 5 to 10 modifies the coordinates of all spectra inside the input file and write them in the output files.

Steps 11-13

Create the table with an on-the-fly resampling of the velocity axis. In the future, the CLASS TABLE will also be able to modify the projection center on-the-fly.

Steps 14-15

Enable to verify by eye that the consistency of the single-dish data with the interferometric data, i.e. the displayed header information for velocity axis and projection center should be identical to the above ones. Those steps are optional as the input really used by the short-spacing processing is produced at step 13.

It is clear that your single-dish data must be reduced before applying the above steps. CLASS has many facilities to visualize and reduce your data.