[daip] Re: AIPS questions

[Magdalena Kunert-Bajraszewska]

Hi Amy,

>  > after the first run of
>  > FRING (for phase calibrator) there is nothing on the map
> On which map, the map of the phase calibrator or the map of the target?

on the map of the target

>   Did you have many failed solutions from FRING?  FRING (and CALIB) will

not many, about 6%, but for aparm(7)=3, and there was no search window

> attempt to make the data look like the the model so can create a source
> if the source is sufficiently weak.  How strong was the resulting
> source?  You could try FRING fitting on the target with a point source

the resulting source has about 10 mJy

> or a different model and see what you get.  You could also try using

I tried this (FRING fitting on the target with a point source) and I got
single point in the center of the target map.

> CALIB, I find that CALIB works better for weak sources, if you have all
> VLBA antennas and are pretty confident that the delays and rates are
> small, since CALIB will not solve for them.  Also, by using a model and
> fringe fitting on the target you have pretty much lost all positional
> accuracy.

I have only 9 VLBA antennas, but I tried to use CALIB, and I got the same
result as FRING on a target with a point
source, which means a single point in the center of the map.
Does it mean that the source structure I got from the FRING run on the
target source (2cm) with a phase-referenced model from the 6cm is an
artefact created by FRING?

Actually I have made some test :) I run fring on the same source but at
the 6cm with a point source as a model. Maybe I should say first that the
phase-referencing on this source at 6cm worked very well and I got a good
map of a double source.  But after fringe fitting with a point source I
also got this double structure but much weaker and in the center of the
map there was another third single component. So as I understand corectly
this central component has been also created by FRING? But why does it
happens like that? The source was too weak to be fringe fitting like that?

> I am interested in why the phase referencing didn't work.  What did the
> CL table look like?  How far away was the phase calibrator and how
> strong was it?  I find that editing the last CL table and getting rid of
> spurious phase winds as one of the best way to improve your phase
> referenced image.

Many delays are in range -0.5 - 0.5 microseconds and rates are in range
-50 to 100 milliHZ, but there are no many single points away from the mean
value, there are rather trends in the points in such ranges.
My source name is 1159+395 and the phase calibrator is J1153+4036
(about 380 mJy at 6cm) and it
is 2 degrees from the target. Also for this target source I have noticed
a difference in coordinates between 18 and 6 cm in order of 0.004
arcseconds.

>
>  > 2. Should I always run FRING on my target source with a
>  > phase-referenced map as a model?
>
> Well it depends on how strong your sources are and your scientific
> goals.  I would avoid using models from other frequencies unless you
> absolutely had to.  In the case where you had a pretty weak source as
> the model, I would not use it at all, but use a point source model if
> the phase referencing did not work, and then fix things up with some
> self-calibration.  If a source is strong enough to run FRING on it is
> strong enough for self-calibration.

OK, but what about the situation that I run FRING on the target source
with a phase-referenced map as a model at the SAME frequency?
Is this step necessary? or maybe When is this necessary?

>  > 3. I have noticed that for some of my sources there is a difference in
>  > coordinates between 18 and 6 cm in order of 0.004 arcseconds.
>
> This can easily be caused by the ionosphere which is important at 18 cm
> but not at 6 and 4 cm.  This can also be a result of slightly different
> structures in the calibrator between 18 and 6 cm.  It is REALLY hard to
> do astrometry at 18 cm without a in-beam phase calibrator because of the
> ionosphere.

Ok, so there is nothing I can do now with this problem during the data
reduction?

Uff, I think that is all for now :)

Best regards,

Magdalena