Spectral Line RFI Characterization Notes
Introduction
This wiki page discribes preliminary work done to characterize RFI found in spectral line data. The initial focus was on creating features found in spectral line data known to contain RFI.
Code
Preprocessing
We started by looking through the data in GBTIDL to find RFI signals using getfs to calibrate and plots scans. The analysis included overlaying different polarizations to see if there are any major differences, variations in Tsys across integrations, and baseline changes across integration. We then used the GBT pipeline is to produce the calibrated data from an SDFITS file for a give scan with RFI to compute values for the analysis we did using GBTIDL.
Features
Polarization differences: Unlike astronomical signals, RFI is polarized. One feature we generated was the difference between two polarizations for a given spectrum integration. The difference was computed by subtracting pol2 from pol1. Differences over the spectrum are not reported for each channel, instead I give the average and standard deviation. This is probably not ideal and should be rethought.
python features/PolarizationDifferences.py generates a table with the following information.
- Date
- Frequency
- Band
- Pol Diff Avg
- Pol Diff Std
- Where or not there is a significant difference in the polarization.
- Tsys1
- Tsys2
--
MikeMcCarty - 2012-09-12