Positional error in restored image
Moderator: Mark.Wieringa
-
- Posts: 14
- Joined: Thu Jan 24, 2013 4:19 pm
Positional error in restored image
Reducing year 2000 data- dual C/X data from C711. The reduction is nothing out of the ordinary, the usual invert, clean and restor tasks are run but when I load a kvis annotation file with known sources onto the restored image, the image is inaccurate by ~ 15 arcsec. Is there something I've forgotten?
-
- Site Admin
- Posts: 220
- Joined: Thu Feb 04, 2010 3:27 pm
- Location: Paul Wild Observatory Narrabri NSW
Re: Positional error in restored image
Hi Stacy,
Have you checked that your phase calibrator was observed at the correct location? If your phase calibrator is away from where you observed it, then this will manifest as a positional error in your target field.
Have you checked that your phase calibrator was observed at the correct location? If your phase calibrator is away from where you observed it, then this will manifest as a positional error in your target field.
cheers
Jamie Stevens
ATCA Senior System Scientist
Jamie Stevens
ATCA Senior System Scientist
-
- Posts: 14
- Joined: Thu Jan 24, 2013 4:19 pm
Re: Positional error in restored image
Hi Jamie,
There is only 7.5 degrees between source and phase calibrator 1933-400... In processing all the L, C and X data, the same effect is seen. Odd.
There is only 7.5 degrees between source and phase calibrator 1933-400... In processing all the L, C and X data, the same effect is seen. Odd.
-
- Site Admin
- Posts: 220
- Joined: Thu Feb 04, 2010 3:27 pm
- Location: Paul Wild Observatory Narrabri NSW
Re: Positional error in restored image
Hi Stacy,
That is very indicative of an offset phase calibrator. If you think the phase calibrator is at position RA = A, Dec = B, but it is actually somewhere else at RA = A + dA, Dec = B + dB, then when you gpcal the phase calibrator it will incorporate this shift into the phase solution. Then when you apply this to the target field, the whole field will be shifted by the dA, dB.
We've been talking about this in the ATCA Forum group, as Baerbel realised there were a couple of calibrators with wrong positions in the database. I'd suggest imaging the phase calibrator field before doing the gpcal step (the phase solution from the flux calibrator should be sufficient for a first pass), and looking for an offset from the field centre.
That is very indicative of an offset phase calibrator. If you think the phase calibrator is at position RA = A, Dec = B, but it is actually somewhere else at RA = A + dA, Dec = B + dB, then when you gpcal the phase calibrator it will incorporate this shift into the phase solution. Then when you apply this to the target field, the whole field will be shifted by the dA, dB.
We've been talking about this in the ATCA Forum group, as Baerbel realised there were a couple of calibrators with wrong positions in the database. I'd suggest imaging the phase calibrator field before doing the gpcal step (the phase solution from the flux calibrator should be sufficient for a first pass), and looking for an offset from the field centre.
cheers
Jamie Stevens
ATCA Senior System Scientist
Jamie Stevens
ATCA Senior System Scientist
-
- Posts: 14
- Joined: Thu Jan 24, 2013 4:19 pm
Re: Positional error in restored image
Just understood your meaning. I'll look at the position.
-
- Posts: 14
- Joined: Thu Jan 24, 2013 4:19 pm
Re: Positional error in restored image
imfit: Revision 1.11, 2013/08/30 01:49:21 UTC
-------------------------------------------------
Object 1933-400_3cm
RMS residual is 1.22E-02 (theoretical image noise is 1.24E-04)
Using the following beam parameters when
deconvolving and converting to integrated flux
Beam Major, minor axes (arcsec): 2.45 2.27
Beam Position angle (degrees): 18.2
Scaling error estimates by 6.1 to account for
noise correlation between pixels
Source 1, Object type: point
Peak value: 0.6516 +/- 1.7269E-02
Offset Position (arcsec): -0.280 -0.072
Positional errors (arcsec): 0.036 0.039
Pos error ellipse (arcsec): 0.039 0.036 18.10
Right Ascension: 19:37:16.194
Declination: -39:58:01.622
-------------------------------------------------
Difference between the imfit position and calibrator database is 0.3 arcsec. So seems the problem is downstream...
-------------------------------------------------
Object 1933-400_3cm
RMS residual is 1.22E-02 (theoretical image noise is 1.24E-04)
Using the following beam parameters when
deconvolving and converting to integrated flux
Beam Major, minor axes (arcsec): 2.45 2.27
Beam Position angle (degrees): 18.2
Scaling error estimates by 6.1 to account for
noise correlation between pixels
Source 1, Object type: point
Peak value: 0.6516 +/- 1.7269E-02
Offset Position (arcsec): -0.280 -0.072
Positional errors (arcsec): 0.036 0.039
Pos error ellipse (arcsec): 0.039 0.036 18.10
Right Ascension: 19:37:16.194
Declination: -39:58:01.622
-------------------------------------------------
Difference between the imfit position and calibrator database is 0.3 arcsec. So seems the problem is downstream...