-
Notifications
You must be signed in to change notification settings - Fork 9
Tasks
This page summarizes a list of tasks and projects that need to be done.
We need to settle on how we plan to do the polarization calibration.
There are many possible schemes for doing this. Currently we are trying to calibrate off of 3c286.
Since we plan to move to guppi for the back end of our observations, it might be best to use the polarization calibration scheme that the pulsar observers use. The tools for this calibration are in the psarchive library (apparently python wrappers exist).
Right now Tabitha and Victor are working on this.
Testing the spectral stability of the noise cal is essential. Even if we do not have a perfect absolute calibration, errors in the spectral calibration will introduce false frequency structure in the foregrounds making them impossible to subtract. The noise cal needs to be spectrally stable to of order a percent (I made that number up) on 3-6 hour time scales (the amount of time between calibrator observations). This is fairly easy to test by comparing pairs of calibrator scans (on the same calibrator) separated by several hours. If the noise cal is not stable, our spectral calibration will need to come from the foregrounds.
I think Aravind is looking into this.
We are still applying the a single flux calibration to all data. This clearly needs to change, with the pipeline automatically choosing a calibrator scan for each data file. Any takers?
Kiyo and Ue-Li are working on this.
While we are a ways from this point still, power spectrum estimation is something we should begin to think about. This could be quite an independent effort.
Changing the optical wigglez catalogs into maps and bringing them to the same format as the 21cm data.
I think Tzu-Ching is working on this.