Sunday, June 14, 2009

Combining Spectral Line UV Data (with slightly different channel setups)

Update: See this lovely post by Adrienne

This is a very good question from reader SillyWolf. I know several of us have been wondering how to concatenate spectral line data sets that have approx. the same wavelength coverage, but different channel setups (different widths, different centers). So any advice would be greatly appreciated by many people!

There are two VLA archive data which were observed in spectral line mode.
These two archive data cover very similar frequency range.
The point is that:
not only they have different pointing center (about 1 arcmin away from each other), but also their reference frequencies do not match each other, so their frequencies don't match channel by channel !!
I want to combine these two data together. can I use DBCON to combine them in UV domain, or I have to resample one of these datacubes in image domain and combine them together?



11 comments:

Adrienne said...

You can use CVEL to align the velocity axis, and you can probably shift the phase reference center (so the pointing centers are similar) with UVFIX (I believe). Then you should be able to DBCON them together.

Adrienne said...

I believe CVEL also allows you to change the channel width!

Korinne said...
This comment has been removed by a blog administrator.
amanda said...

Hi Adrienne,

It just so happens that I needed to do something like this. I just double-checked the CVEL and DBCON documentation and it looks like CVEL can't change the channel size. DBCON will shift the velocity/frequency axis, so I don't think you need to use CVEL. I'll experiment and check back.

sillywolf said...

Hi Adrienne and Amanda,
I checked the explain for DBCON too. There is an option named with "DOPOS", which looks like working for the job of spatial and frequency shifting.

I will check back too.

amanda said...
This comment has been removed by the author.
amanda said...

After playing around with stuff last night, it looks like DBCON will shift the center position (RA, DEC), but not the velocity. If you have the same channel size/bandwidth, then you can just use cvel to shift data set 2 to the frequency center of data set 1. However, if your data sets have two different channel size/bandwidth setups, this becomes much harder. There's a couple of tasks in AIPS that can deal with rechopping the channels (AVSPC, SPECR), but I couldn't get the channel size between the two data sets to match. There may be some magic hidden in AIPS to do this, but I couldn't find it. What ended up working was to export the data to Miriad, use uvlist options=spectral to figure out the velocity ranges of each data set, and then just invert including both data sets in the vis parameter and inserting the desired velocity ranges in the line parameter.

Adrienne said...

Darn, I just ran into this myself, Amanda. Unfortunately our data sets are too big (!) for miriad to read in properly. If I wanted to go the miriad route I'd have to split off the data with 4-5 separate time ranges total. not fun. It inspires me to look for some AIPS magic, but I suppose I could say that it's not a hopeful inspiration.

amanda said...

Adrienne: Maybe CASA could handle this? I know someone who's imaged SMA data (large data files) in CASA.

Anonymous said...

you may also be able to re-sort you channel separations using the aips task 'specr' and 'bloat' etc. This may allow you to resort the channel positions and widths of your two data sets so that they match and hence can be dbcon'ed together.

Anonymous said...

Hallo.

I've also troubles with HI VLA Data, wich was observed with 4IF mode. The trouble is the small overlap between two spectral windows, then the HI Line of a galaxy lies exactly on its sides. After assembly this spectral windows I have an attenuation of the lineintensity in the middle of my sweet galaxy. Have someone experience with this trouble or have some suggestions to the solution?