Initial Look at VLA Measurement Set - BHianca/VLA GitHub Wiki

We should first take a look at the operator log for the data. Go to this website and find the project code to download the observing log pdf. http://www.vla.nrao.edu/cgi-bin/oplogs.cgi The project code was 17A-248 and data was taken in 2017 on May 30. The pdf is available here also https://github.com/Cosentino-Troja/AstroTrans/blob/master/Wiki/2017-05-30_1037_17A-248.pdf One can see from this log that the conditions were very good for acquiring data with clear skies and that there was no down time reported. This is a very good log and probably excellent data.

Let's look first at the data. Open CASA by double clicking on the CASA_5.3 app icon in your applications folder.

This opens a CASA terminal utilizing ipython and a CASA message or log window. Important to see both.

First thing we want to do is get familiar with the data in the measurement set or .ms (we will refer to the .ms as the data hereafter). Find the path to where the data was downloaded to. Create a new folder for it if you'd like to keep things clean and start from the beginning (highly recommended). We want to use the CASA "task" called "listobs()". Put the path and .ms together into listobs('pwd'+'.ms') and hit return. Look at your log message window.

It is important to note a few things. First, CASA uses 'functions' called "tasks". They can be called in the terminal and some have GUIs that can be launched when activated. You can also look at the inputs and outputs of a task, that will be explained next. Second, it is good to have a heirachy and naming convention moving forward that you will use and stick to. Try to be explicit but brief also. Third, I have tried to put tasks and commands in black text and a different font within the screencaps. I will use quotes to refer to them on the wiki page.

If we want to call up listobs.

Now we have created a file. Let's look at it. Note how the text file is much easier to navigate and keep open rather than looking back in the log window. Check out the spectral windows for the science target. Also look at the scans and how the observation was conducted, looking at the source, acquiring data, then going to the phase calibrator, and then back to the source. The field ID changes depending where the VLA is pointed and the designations of the objects in the scheduling blocks that created the observation.

To plot the data, we can use CASA's task plotms(), which can be launched as a GUI like shown below. You will need to open the directory of the .ms but following these example steps. Plotms() is sometimes rather clunky. Be patient when plotting and clicking on tabs and checking boxes or entering characters or modifying fields. This is the GUI. Use it to find the .ms and open it from the appropriate path.

We want to look at the data of 3C48 first. This is because it is bright and will let us know about the general quality of the data from the beginning of the observation if everything was executed as expected. Color by spectral window and frequency. Note RFI and the rest looking like it is zero. Note how the RFI appears. It is very tightly centered on a specific frequency and has a very strong peak amplitude. This is probably a communication satellite, but the VLA can also pick up cell phone signals. These sources of RFI look characteristically different from other radio emitting sources.

Try looking at 3C48 by narrowing the spectral windows to not include data below spw=19. Here we are not seeing radio emission from 3C48 between 2.4-4.0 GHz. We didn't remove the data at frequencies between 2.0-2.4 GHz, we just didn't plot those data. Now the amplitudes are much lower than the RFI but also have broad - nearly continuous / constant amplitude with frequency. That is what you want as a calibrator source - very well-behaved radio emission amplitude with frequency in time.

This last plot here is what data could look like if you do not average or specify which data you'd like to plot. Trying to visualize multiple fields of data like this is NOT RECOMMENDED. DO NOT DO THIS unless you have very small data sets. We will revisit manual flagging later and we will specify narrow windows of time or frequency space to investigate. It is rare to ever try to look at large amounts of VLA data - because of all of the reasons mentioned earlier.