|
|
| (147 intermediate revisions by 4 users not shown) |
| Line 1: |
Line 1: |
| == Overview ==
| | #REDIRECT [[EVLA 6cmWideband Tutorial SN2010FZ-CASA4.4]] |
| This article describes the calibration and imaging of a single-pointing 6cm EVLA wideband continuum dataset on the galaxy NGC2967 (UGC5180) which was the location of the supernova candidate SN2010FZ. No supernova was detected in this observation, but the galactic continuum emission from this face-on spiral is adequately imaged. The data were taken in RSRO mode, with 1024 MHz of bandwidth in each of two widely spaced basebands (comprised each of 8 128 MHz spectral windows), spanning 4.5 to 7.5 GHz. We will use wideband imaging techniques in this tutorial.
| |
| | |
| This is a more advanced tutorial, and if you are a relative novice (and <em>particularly</em> for EVLA continuum calibration and imaging), it is <em>strongly</em> recommended that you start with the [[EVLA Continuum Tutorial 3C391]] before tackling this dataset. We will not include basic information on CASA processing in this tutorial.
| |
| | |
| == CASA Versions ==
| |
| | |
| This tutorial was written for the CASA Version 3.2.1 (release r15198 26 May 2011).
| |
| | |
| == Obtaining the Data ==
| |
| | |
| The scheduling block (SB) processed appears in the EVLA archive under program AS1015 as
| |
| <tt>AS1015_sb1658169_1.55388.89474846065</tt>
| |
| and was run on 2010-07-11 from 21:28 to 22:28 UT (size 37.74GB).
| |
| | |
| For the purposes of this tutorial, we have provided the raw SDM data (as would be extracted from the archive) as well as measurement sets created by filling the data (with the {{importevla}} task) and upon time-averaging to 10s (after application of the online flags).
| |
| | |
| To start your tutorial, depending on which dataset you start with, proceed to:
| |
| * <em>To start with the raw SDM data:</em> Start with the section below titled "[[#Importing your EVLA data from SDM]]". This is where you would start if you were reducing data from the archive.
| |
| * <em>To start with the raw filled MS:</em> Start with the section below titled "[[#Application of Online Flags and Averaging your MS]]".
| |
| * <em>To start with the flagged and averaged MS:</em> Start with the section below titled "[[#Examining and Flagging your Averaged MS]]".
| |
| | |
| == Importing your EVLA data from SDM ==
| |
| | |
| For the purposes of this tutorial, we assume that the SDM is resident on disk, in this case at the location:
| |
| <pre>/lustre/smyers/AS1015/AS1015_sb1658169_1.55388.89474846065</pre>
| |
| Use the actual location of your data when you carry out the commands.
| |
| | |
| The {{listsdm}} task will print out a summary of the scans, fields, spectral windows, and antennas present in your SDM.
| |
| | |
| <source lang="python">
| |
| # In CASA
| |
| listsdm('/lustre/smyers/AS1015/AS1015_sb1658169_1.55388.89474846065')
| |
| </source>
| |
| | |
| In the logger you should see:
| |
| <pre>
| |
| ================================================================================
| |
| SDM File: /lustre/smyers/AS1015/AS1015_sb1658169_1.55388.89474846065
| |
| ================================================================================
| |
| Observer: Dr. Alicia M. Soderberg
| |
| Facility: EVLA, D-configuration
| |
| Observed from 2010/07/11/21:28:28.41 to 2010/07/11/22:28:17.73 (UTC)
| |
| Total integration time = 3589.32 seconds (1.00 hours)
| |
|
| |
| Scan listing:
| |
| Timerange (UTC) Scan FldID FieldName SpwIDs Intent(s)
| |
| 21:28:28.41 - 21:29:27.40 1 0 J0925+0019 [0, 1] CALIBRATE_PHASE
| |
| 21:29:27.40 - 21:30:57.16 2 0 J0925+0019 [0, 1] CALIBRATE_PHASE
| |
| 21:30:57.16 - 21:32:26.91 3 0 J0925+0019 [0, 1] CALIBRATE_PHASE
| |
| 21:32:26.91 - 21:33:56.67 4 0 J0925+0019 [0, 1] CALIBRATE_PHASE
| |
| 21:33:56.67 - 21:34:56.50 5 0 J0925+0019 [0, 1] CALIBRATE_PHASE
| |
| 21:34:56.50 - 21:35:56.34 6 0 J0925+0019 [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] CALIBRATE_PHASE
| |
| 21:35:56.34 - 21:37:26.09 7 0 J0925+0019 [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] CALIBRATE_PHASE
| |
| 21:37:26.09 - 21:38:25.93 8 0 J0925+0019 [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] CALIBRATE_PHASE
| |
| 21:38:25.93 - 21:39:55.68 9 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 21:39:55.68 - 21:41:25.44 10 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 21:41:25.44 - 21:42:55.19 11 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 21:42:55.19 - 21:44:24.94 12 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 21:44:24.94 - 21:45:54.70 13 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 21:45:54.70 - 21:47:24.45 14 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 21:47:24.45 - 21:47:54.37 15 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 21:47:54.37 - 21:49:24.12 16 0 J0925+0019 [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] CALIBRATE_PHASE
| |
| 21:49:24.12 - 21:50:53.88 17 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 21:50:53.88 - 21:52:23.63 18 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 21:52:23.63 - 21:53:53.39 19 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 21:53:53.39 - 21:55:23.14 20 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 21:55:23.14 - 21:56:52.89 21 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 21:56:52.89 - 21:58:22.65 22 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 21:58:22.65 - 21:58:52.57 23 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 21:58:52.57 - 22:00:22.32 24 0 J0925+0019 [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] CALIBRATE_PHASE
| |
| 22:00:22.32 - 22:01:52.07 25 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 22:01:52.07 - 22:03:21.83 26 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 22:03:21.83 - 22:04:51.58 27 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 22:04:51.58 - 22:06:21.34 28 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 22:06:21.34 - 22:07:51.09 29 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 22:07:51.09 - 22:09:20.85 30 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 22:09:20.85 - 22:09:50.76 31 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 22:09:50.76 - 22:11:20.52 32 0 J0925+0019 [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] CALIBRATE_PHASE
| |
| 22:11:20.52 - 22:12:50.27 33 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 22:12:50.27 - 22:14:20.02 34 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 22:14:20.02 - 22:15:49.78 35 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 22:15:49.78 - 22:17:19.53 36 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 22:17:19.53 - 22:18:49.29 37 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 22:18:49.29 - 22:20:19.04 38 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 22:20:19.04 - 22:20:48.96 39 1 SN2010FZ [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] OBSERVE_TARGET
| |
| 22:20:48.96 - 22:22:18.71 40 0 J0925+0019 [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] CALIBRATE_PHASE
| |
| 22:22:18.71 - 22:23:48.47 41 2 3C286 [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] CALIBRATE_BANDPASS CALIBRATE_AMPLI
| |
| 22:23:48.47 - 22:25:18.22 42 2 3C286 [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] CALIBRATE_BANDPASS CALIBRATE_AMPLI
| |
| 22:25:18.22 - 22:26:47.98 43 2 3C286 [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] CALIBRATE_BANDPASS CALIBRATE_AMPLI
| |
| 22:26:47.98 - 22:28:17.73 44 2 3C286 [2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17] CALIBRATE_BANDPASS CALIBRATE_AMPLI
| |
|
| |
| Spectral window information:
| |
| SpwID #Chans Ch0(MHz) ChWidth(kHz) TotBW(MHz) Baseband
| |
| 0 64 7686.0 2000.0 128.0 BB_4
| |
| 1 64 7836.0 2000.0 128.0 BB_8
| |
| 2 64 4488.0 2000.0 128.0 BB_4
| |
| 3 64 4616.0 2000.0 128.0 BB_4
| |
| 4 64 4744.0 2000.0 128.0 BB_4
| |
| 5 64 4872.0 2000.0 128.0 BB_4
| |
| 6 64 5000.0 2000.0 128.0 BB_4
| |
| 7 64 5128.0 2000.0 128.0 BB_4
| |
| 8 64 5256.0 2000.0 128.0 BB_4
| |
| 9 64 5384.0 2000.0 128.0 BB_4
| |
| 10 64 6488.0 2000.0 128.0 BB_8
| |
| 11 64 6616.0 2000.0 128.0 BB_8
| |
| 12 64 6744.0 2000.0 128.0 BB_8
| |
| 13 64 6872.0 2000.0 128.0 BB_8
| |
| 14 64 7000.0 2000.0 128.0 BB_8
| |
| 15 64 7128.0 2000.0 128.0 BB_8
| |
| 16 64 7256.0 2000.0 128.0 BB_8
| |
| 17 64 7384.0 2000.0 128.0 BB_8
| |
|
| |
| Field information:
| |
| FldID Code Name RA Dec SrcID
| |
| 0 D J0925+0019 09:25:07.82 +000.19.13.933 0
| |
| 1 NONE SN2010FZ 09:42:04.77 +000.19.51.000 1
| |
| 2 K 3C286 13:31:08.29 +030.30.32.959 2
| |
|
| |
| Antennas (27):
| |
| ID Name Station Diam.(m) Lat. Long.
| |
| 0 ea01 W09 25.0 +000.00.00.0 +000.00.00.0
| |
| 1 ea02 E02 25.0 +033.53.51.0 -107.37.25.2
| |
| 2 ea03 E09 25.0 +033.54.01.1 -107.37.04.4
| |
| 3 ea04 W01 25.0 +033.53.53.6 -107.36.45.1
| |
| 4 ea05 W08 25.0 +033.54.00.5 -107.37.05.9
| |
| 5 ea06 N06 25.0 +033.53.53.0 -107.37.21.6
| |
| 6 ea08 N01 25.0 +033.54.10.3 -107.37.06.9
| |
| 7 ea09 E06 25.0 +033.54.01.8 -107.37.06.0
| |
| 8 ea10 N03 25.0 +033.53.57.7 -107.36.55.6
| |
| 9 ea11 E04 25.0 +033.54.04.8 -107.37.06.3
| |
| 10 ea12 E08 25.0 +033.53.59.7 -107.37.00.8
| |
| 11 ea13 N07 25.0 +033.53.55.1 -107.36.48.9
| |
| 12 ea14 E05 25.0 +033.54.12.9 -107.37.07.2
| |
| 13 ea15 W06 25.0 +033.53.58.8 -107.36.58.4
| |
| 14 ea16 W02 25.0 +033.53.56.4 -107.37.15.6
| |
| 15 ea17 W07 25.0 +033.54.00.9 -107.37.07.5
| |
| 16 ea18 N09 25.0 +033.53.54.8 -107.37.18.4
| |
| 17 ea19 W04 25.0 +033.54.19.0 -107.37.07.8
| |
| 18 ea20 N05 25.0 +033.53.59.1 -107.37.10.8
| |
| 19 ea21 E01 25.0 +033.54.08.0 -107.37.06.7
| |
| 20 ea22 N04 25.0 +033.53.59.2 -107.37.05.7
| |
| 21 ea23 E07 25.0 +033.54.06.1 -107.37.06.5
| |
| 22 ea24 W05 25.0 +033.53.56.5 -107.36.52.4
| |
| 23 ea25 N02 25.0 +033.53.57.8 -107.37.13.0
| |
| 24 ea26 W03 25.0 +033.54.03.5 -107.37.06.2
| |
| 25 ea27 E03 25.0 +033.54.00.1 -107.37.08.9
| |
| 26 ea28 N08 25.0 +033.54.00.5 -107.37.02.8
| |
| </pre>
| |
| The C-band data of interest is contained in scans 6-44 and spans spectral windows 2 to 17.
| |
| | |
| We use the {{importevla}} task to convert the SDM dataset from the archive to a CASA Measurement Set (MS).
| |
| | |
| <source lang="python">
| |
| # In CASA
| |
| importevla(asdm='/lustre/smyers/AS1015/AS1015_sb1658169_1.55388.89474846065', \
| |
| vis='SN2010FZ_filled.ms',online=True,flagzero=True, \
| |
| shadow=True,applyflags=False,tbuff=1.5,flagbackup=False)
| |
| </source>
| |
| | |
| Here we had the task create (but not apply) the online flagging commands, plus flags for zero-clipping and shadowing. The timeranges for the online flags were extended by 1.5sec (the integration time was 1sec) to account for some timing mismatches present in the EVLA data at this time. These online flags indicated times where the antennas were not on source (e.g. slewing) or had other detectable faults. The created flagging commands will be stored in the <tt>FLAG_CMD</tt> MS table and can be applied later. Note that if you set <tt>applyflags=True</tt> here then after filling the task will go ahead and apply the flags for you.
| |
| | |
| For the purposes of this exercise, in order to save time and disk space, we have turned off the automatic creation of flag column backups by setting <tt>flagbackup=False</tt>. If we make a mistake and need to recover flags then we will have to rerun all previous commands. We recommend that for real data processing that you leave the default value <tt>flagbackup=True</tt> in this and subsequent tasks.
| |
| | |
| You now have a MS called <tt>SN2010FZ_filled.ms</tt> in your working area. This should be 37GB like the SDM.
| |
| | |
| == Application of Online Flags and Averaging your MS ==
| |
| | |
| If you are starting from the filled MS, you can find this at the AOC at:
| |
| <pre>/lustre/smyers/AS1015/SN2010FZ_filled.ms</pre>
| |
| Again, use the actual location of this file for your system.
| |
| | |
| NOTE: the following step will not work in Version 3.2.1 (you will get a blank plot) but should in later versions).
| |
| You can examine the commands stored in the <tt>FLAG_CMD</tt> table using {{flagcmd}}.
| |
| <source lang="python">
| |
| # In CASA
| |
| flagcmd(vis='SN2010FZ_filled.ms',flagmode='table',optype='plot')
| |
| </source>
| |
| This will bring up a <tt>matplotlib</tt> plotter. You can have it plot to a PNG file instead:
| |
| <source lang="python">
| |
| # In CASA
| |
| flagcmd(vis='SN2010FZ_filled.ms',flagmode='table',optype='plot',outfile='plotSN2010FZ_flagcmd.png')
| |
| </source>
| |
| | |
| To apply the flags also use {{flagcmd}}:
| |
| <source lang="python">
| |
| # In CASA
| |
| flagcmd(vis='SN2010FZ_filled.ms',flagmode='table',optype='apply',flagbackup=False)
| |
| </source>
| |
| This can take a while for our 37GB dataset. It took 20min on my workstation.
| |
| | |
| With the known bad data flagged, we can now split out the data we want and also average down in time to make a smaller MS.
| |
| For D-configuration (max baselines 1km) we can safely average to 3s or even 10s to reduce dataset size:
| |
| <source lang="python">
| |
| # In CASA
| |
| split(vis='SN2010FZ_filled.ms',outputvis='SN2010FZ_filled10s.ms',datacolumn='data',timebin='10s')
| |
| </source>
| |
| This can also take a while for our 37GB dataset. It took 20min on my workstation.
| |
| | |
| You now have a MS called <tt>SN2010FZ_filled10s.ms</tt> in your working area. This should be 3.2GB in size.
| |
| | |
| == Examining and Flagging your Averaged MS ==
| |
| | |
| If you are starting from the pre-flagged averaged split MS, you can find this at the AOC at:
| |
| <pre>/lustre/smyers/AS1015/SN2010FZ_filled10s.ms</pre>
| |
| | |
| We use {{listobs}} to summarize our new MS:
| |
| <source lang="python">
| |
| # In CASA
| |
| listobs('SN2010FZ_filled10s.ms')
| |
| </source>
| |
| Scan 6 is a dummy scan so we will use scans 7 to 44 when we process our data.
| |
| | |
| [[Image:plotSN2010FZ_plotants.png|200px|thumb|right|plotants plotter]]
| |
| To plot up the antenna positions in the array:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotants('SN2010FZ_filled10s.ms')
| |
| </source>
| |
| NOTE: if after this point or any other you get table locks, use {{clearstat}} to clear them:
| |
| <source lang="python">
| |
| # In CASA
| |
| clearstat
| |
| </source>
| |
| | |
| Now we examine the MS looking for bad data to flag. We will use {{plotms}} to bring up an interactive GUI that will display 2-D Y vs.X style line plots. NOTE: We do not recommend using the editing/flagging features of plotms. It is very easy to mess up your data that way. Also, we will be restricting the scope of plotting so most box/flag operations would not get rid of all the bad data. We will instead use plotms to identify bad data and then use flagcmd to flag it. This will also allow full scripting of the flagging.
| |
| | |
| WARNING: The Flag button on the plotms GUI is close to other buttons you will be using, in particular the one that gets rid of boxes you have drawn. Be careful and don't hit the Flag button by mistake!
| |
| | |
| The useful spw are 2~17. To get an idea of the data layout, plot a single baseline/channel versus time:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_filled10s.ms',field='',spw='2~17:31~31', \
| |
| antenna='ea01&ea02',correlation='RR,LL',xaxis='time',yaxis='amp')
| |
| </source>
| |
| | |
| [[Image:plotSN2010FZ_plotms_ants.png|200px|thumb|right|plotms ant2 vs ea01]]
| |
| Look for bad antennas by picking the last field and plotting baselines versus antenna <tt>ea01</tt>:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_filled10s.ms',field='2',spw='2~17:31~31', \
| |
| antenna='ea01',correlation='RR,LL',xaxis='antenna2',yaxis='amp')
| |
| </source>
| |
| You should be able to see that antenna 11 (= ea13) is bad (very low amplitude, it has no C-band receiver!) and that some of the spectral windows on 15 and 23 (ea17,ea25) are also on the low side. Boxing and using Locate will show that spw 10~17 are suspect for these antennas.
| |
| | |
| [[Image:plotSN2010FZ_plotms_ea02.png|200px|thumb|right|plotms ea02 vs frequency]]
| |
| Now look at the bandpass for ea02 - it is in the inner core and a prospective reference antenna. Exclude ea13 using negation in the selection:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_filled10s.ms',field='2',spw='2~17', \
| |
| antenna='ea02;!ea13',correlation='RR,LL',xaxis='frequency',yaxis='amp')
| |
| </source>
| |
| There is clearly less data for spw 11, and use of Locate shows spw 11 data only for ea02,ea03,04,08,09,11,12. We will later delete this incomplete spw. Note also the very strong RFI spike at 6614MHz (spw 10 ch 63) with clear ringing contaminating both spw 10 and 11. There is also a tremendous roll-off in spw 10. We will drop these spectral window when we process the data.
| |
| | |
| [[Image:plotSN2010FZ_plotms_ea02ea20.png|200px|thumb|right|plotms ea02&ea20 iteration phase vs frequency]]
| |
| We can also step through the baselines to our antenna using iteraxis - use the ">" button to step through:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_filled10s.ms',field='2',spw='2~17',antenna='ea02;!ea13', \
| |
| correlation='RR,LL',xaxis='frequency',yaxis='amp',iteraxis='baseline')
| |
| </source>
| |
| This will make it easier to isolate the bad antennas. Now plot the phases, iterating through baselines to ea02:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_filled10s.ms',field='2',spw='2~17',antenna='ea02;!ea13', \
| |
| correlation='RR,LL',xaxis='frequency',yaxis='phase',iteraxis='baseline')
| |
| </source>
| |
| You see the slopes due to residual delays. Mostly a turn or less over a 128MHz subband, but there are some outliers.
| |
| Step through to ea20. You see that there is a very large delay in RR (via locate) for
| |
| the first baseband (spw 0~7). We will delete this (will also delete LL so there are no orphan polarization products).
| |
| Note ea17 and ea25 baselines drop close to zero in the middle of upper baseband (e.g. plot 'ea17&ea25') so we will delete these.
| |
| | |
| To carry out flagging, we again use {{flagcmd}} in the mode where it takes a list of command strings:
| |
| <source lang="python">
| |
| # In CASA
| |
| flaglist = ['antenna="ea13"',
| |
| 'antenna="ea17" spw="10~17"',
| |
| 'antenna="ea25" spw="10~17"',
| |
| 'antenna="ea20" spw="2~9"']
| |
| flagcmd(vis='SN2010FZ_filled10s.ms',flagmode='cmd',command=flaglist,optype='apply',flagbackup=False)
| |
| </source>
| |
| These commands will be carried out as well as being added to the FLAG_CMD table (marked as applied).
| |
| | |
| Plot the data again now having flagged:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_filled10s.ms',field='2',spw='2~17',antenna='ea02', \
| |
| correlation='RR,LL',xaxis='frequency',yaxis='amp',scan='7~44')
| |
| </source>
| |
| | |
| [[Image:plotSN2010FZ_plotms_ea02fld0.png|200px|thumb|right|plotms field 0 ea02 amp vs frequency]]
| |
| Now our phase calibrator - it is weaker, and we now start to see the RFI:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_filled10s.ms',field='0',spw='2~17',antenna='ea02', \
| |
| correlation='RR,LL',xaxis='frequency',yaxis='amp',scan='7~44')
| |
| </source>
| |
| Use the Zoom feature, Mark rectangles and use Locate to identify the frequency/channel of RFI. In particular, we note in our analysis:
| |
| * 6614MHz (spw 10 ch 63) super strong
| |
| * 6772-6778MHz (spw 12 ch 14-17)
| |
| * 7260-7264MHz (spw 16 ch 2-4)
| |
| * 7314-7340MHz (spw 16 ch 29-42)
| |
| * 7402-7418MHz (spw 17 ch 9-17)
| |
| * 7458-7466MHz (spw 17 ch 37-41)
| |
| * 7488MHz (spw 17 ch 52)
| |
| | |
| If you plot all antennas and avoid the band edges you see spw 16,17 are pretty wiped out:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_filled10s.ms',field='0',spw='2~17:4~59',antenna='', \
| |
| correlation='RR,LL',xaxis='frequency',yaxis='amp',scan='7~44')
| |
| </source>
| |
| For now we will not flag these channels but note them to mask out when creating continuum calibration tables and images.
| |
| | |
| Finally, split off the good scans and spw, this will allow us to work on the data without having to start completely over if we mess something up badly as well as letting us do simpler selection:
| |
| <source lang="python">
| |
| # In CASA
| |
| split(vis='SN2010FZ_filled10s.ms',outputvis='SN2010FZ_flagged10s.ms',datacolumn='data',spw='2~9,12~17',scan='7~44')
| |
| </source>
| |
| | |
| You now have a MS called <tt>SN2010FZ_flagged10s.ms</tt> in your working area. This should be 2.8GB in size.
| |
| | |
| == Calibration ==
| |
| | |
| Summarize the split flagged MS:
| |
| <source lang="python">
| |
| # In CASA
| |
| listobs('SN2010FZ_flagged10s.ms')
| |
| </source>
| |
| In the logger we see:
| |
| <pre>
| |
| ================================================================================
| |
| MeasurementSet Name: /home/sandrock2/smyers/casa/tutorials/SN2010FZ/SN2010FZ_flagged10s.ms MS Version 2
| |
| ================================================================================
| |
| Observer: Dr. Alicia M. Soderberg Project: T.B.D.
| |
| Observation: EVLA
| |
| Data records: 1374548 Total integration time = 3042 seconds
| |
| Observed from 11-Jul-2010/21:36:01.0 to 11-Jul-2010/22:26:43.0 (UTC)
| |
| | |
| ObservationID = 0 ArrayID = 0
| |
| Date Timerange (UTC) Scan FldId FieldName nVis Int(s) SpwIds ScanIntent
| |
| 11-Jul-2010/21:36:01.0 - 21:38:20.5 7 0 J0925+0019 73710 9.93 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]CALIBRATE_PHASE#UNSPECIFIED
| |
| 21:38:44.0 - 21:39:51.0 9 1 SN2010FZ 39312 9.11 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 21:40:01.0 - 21:41:20.5 10 1 SN2010FZ 44226 9.89 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 21:41:30.0 - 21:42:50.0 11 1 SN2010FZ 44226 10 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 21:43:00.0 - 21:44:20.0 12 1 SN2010FZ 44226 10 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 21:44:30.0 - 21:45:50.0 13 1 SN2010FZ 44226 10 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 21:46:00.0 - 21:47:19.5 14 1 SN2010FZ 44226 9.89 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 21:47:29.0 - 21:47:48.5 15 1 SN2010FZ 14742 9.67 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 21:48:12.0 - 21:49:18.5 16 0 J0925+0019 39312 8.92 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]CALIBRATE_PHASE#UNSPECIFIED
| |
| 21:49:42.0 - 21:50:49.0 17 1 SN2010FZ 39312 9.11 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 21:50:59.0 - 21:52:19.0 18 1 SN2010FZ 44226 10 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 21:52:29.0 - 21:53:48.5 19 1 SN2010FZ 44226 9.89 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 21:53:58.0 - 21:55:18.0 20 1 SN2010FZ 44226 10 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 21:55:28.0 - 21:56:48.0 21 1 SN2010FZ 44226 10 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 21:56:58.0 - 21:58:18.0 22 1 SN2010FZ 44226 10 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 21:58:28.0 - 21:58:47.0 23 1 SN2010FZ 14742 9.33 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 21:59:11.0 - 22:00:17.0 24 0 J0925+0019 39312 8.93 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]CALIBRATE_PHASE#UNSPECIFIED
| |
| 22:00:40.0 - 22:01:47.0 25 1 SN2010FZ 39312 9.12 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 22:01:57.0 - 22:03:17.0 26 1 SN2010FZ 44226 10 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 22:03:27.0 - 22:04:47.0 27 1 SN2010FZ 44226 10 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 22:04:57.0 - 22:06:16.5 28 1 SN2010FZ 44226 9.89 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 22:06:26.0 - 22:07:46.0 29 1 SN2010FZ 44226 10 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 22:07:56.0 - 22:09:16.0 30 1 SN2010FZ 44226 10 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 22:09:26.0 - 22:09:45.0 31 1 SN2010FZ 14742 9.33 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 22:10:08.5 - 22:11:15.0 32 0 J0925+0019 39312 8.94 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]CALIBRATE_PHASE#UNSPECIFIED
| |
| 22:11:38.5 - 22:12:45.5 33 1 SN2010FZ 39312 9.11 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 22:12:55.0 - 22:14:15.0 34 1 SN2010FZ 44226 10 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 22:14:25.0 - 22:15:45.0 35 1 SN2010FZ 44226 10 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 22:15:55.0 - 22:17:15.0 36 1 SN2010FZ 44226 10 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 22:17:25.0 - 22:18:44.5 37 1 SN2010FZ 44226 9.89 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 22:18:54.0 - 22:20:14.0 38 1 SN2010FZ 44226 10 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 22:20:24.0 - 22:20:43.5 39 1 SN2010FZ 14742 9.6 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]OBSERVE_TARGET#UNSPECIFIED
| |
| 22:21:06.5 - 22:22:13.5 40 0 J0925+0019 38584 8.96 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]CALIBRATE_PHASE#UNSPECIFIED
| |
| 22:25:13.0 - 22:25:13.0 42 2 3C286 770 2.65 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]CALIBRATE_BANDPASS#UNSPECIFIED,CALIBRATE_AMPLI#UNSPECIFIED
| |
| 22:25:23.0 - 22:26:43.0 43 2 3C286 42812 9.62 [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]CALIBRATE_BANDPASS#UNSPECIFIED,CALIBRATE_AMPLI#UNSPECIFIED
| |
| (nVis = Total number of time/baseline visibilities per scan)
| |
| Fields: 3
| |
| ID Code Name RA Decl Epoch SrcId nVis
| |
| 0 D J0925+0019 09:25:07.8150 +00.19.13.9334 J2000 0 230230
| |
| 1 NONE SN2010FZ 09:42:04.7700 +00.19.51.0000 J2000 1 1100736
| |
| 2 K 3C286 13:31:08.2880 +30.30.32.9589 J2000 2 43582
| |
| (nVis = Total number of time/baseline visibilities per field)
| |
| Spectral Windows: (14 unique spectral windows and 1 unique polarization setups)
| |
| SpwID #Chans Frame Ch1(MHz) ChanWid(kHz)TotBW(kHz) Ref(MHz) Corrs
| |
| 0 64 TOPO 4488 2000 128000 4488 RR RL LR LL
| |
| 1 64 TOPO 4616 2000 128000 4616 RR RL LR LL
| |
| 2 64 TOPO 4744 2000 128000 4744 RR RL LR LL
| |
| 3 64 TOPO 4872 2000 128000 4872 RR RL LR LL
| |
| 4 64 TOPO 5000 2000 128000 5000 RR RL LR LL
| |
| 5 64 TOPO 5128 2000 128000 5128 RR RL LR LL
| |
| 6 64 TOPO 5256 2000 128000 5256 RR RL LR LL
| |
| 7 64 TOPO 5384 2000 128000 5384 RR RL LR LL
| |
| 8 64 TOPO 6744 2000 128000 6744 RR RL LR LL
| |
| 9 64 TOPO 6872 2000 128000 6872 RR RL LR LL
| |
| 10 64 TOPO 7000 2000 128000 7000 RR RL LR LL
| |
| 11 64 TOPO 7128 2000 128000 7128 RR RL LR LL
| |
| 12 64 TOPO 7256 2000 128000 7256 RR RL LR LL
| |
| 13 64 TOPO 7384 2000 128000 7384 RR RL LR LL
| |
| Sources: 42
| |
| ID Name SpwId RestFreq(MHz) SysVel(km/s)
| |
| 0 J0925+0019 0 - -
| |
| 0 J0925+0019 1 - -
| |
| 0 J0925+0019 2 - -
| |
| 0 J0925+0019 3 - -
| |
| 0 J0925+0019 4 - -
| |
| 0 J0925+0019 5 - -
| |
| 0 J0925+0019 6 - -
| |
| 0 J0925+0019 7 - -
| |
| 0 J0925+0019 8 - -
| |
| 0 J0925+0019 9 - -
| |
| 0 J0925+0019 10 - -
| |
| 0 J0925+0019 11 - -
| |
| 0 J0925+0019 12 - -
| |
| 0 J0925+0019 13 - -
| |
| 1 SN2010FZ 0 - -
| |
| 1 SN2010FZ 1 - -
| |
| 1 SN2010FZ 2 - -
| |
| 1 SN2010FZ 3 - -
| |
| 1 SN2010FZ 4 - -
| |
| 1 SN2010FZ 5 - -
| |
| 1 SN2010FZ 6 - -
| |
| 1 SN2010FZ 7 - -
| |
| 1 SN2010FZ 8 - -
| |
| 1 SN2010FZ 9 - -
| |
| 1 SN2010FZ 10 - -
| |
| 1 SN2010FZ 11 - -
| |
| 1 SN2010FZ 12 - -
| |
| 1 SN2010FZ 13 - -
| |
| 2 3C286 0 - -
| |
| 2 3C286 1 - -
| |
| 2 3C286 2 - -
| |
| 2 3C286 3 - -
| |
| 2 3C286 4 - -
| |
| 2 3C286 5 - -
| |
| 2 3C286 6 - -
| |
| 2 3C286 7 - -
| |
| 2 3C286 8 - -
| |
| 2 3C286 9 - -
| |
| 2 3C286 10 - -
| |
| 2 3C286 11 - -
| |
| 2 3C286 12 - -
| |
| 2 3C286 13 - -
| |
| Antennas: 27:
| |
| ID Name Station Diam. Long. Lat.
| |
| 0 ea01 W09 25.0 m -107.37.25.2 +33.53.51.0
| |
| 1 ea02 E02 25.0 m -107.37.04.4 +33.54.01.1
| |
| 2 ea03 E09 25.0 m -107.36.45.1 +33.53.53.6
| |
| 3 ea04 W01 25.0 m -107.37.05.9 +33.54.00.5
| |
| 4 ea05 W08 25.0 m -107.37.21.6 +33.53.53.0
| |
| 5 ea06 N06 25.0 m -107.37.06.9 +33.54.10.3
| |
| 6 ea08 N01 25.0 m -107.37.06.0 +33.54.01.8
| |
| 7 ea09 E06 25.0 m -107.36.55.6 +33.53.57.7
| |
| 8 ea10 N03 25.0 m -107.37.06.3 +33.54.04.8
| |
| 9 ea11 E04 25.0 m -107.37.00.8 +33.53.59.7
| |
| 10 ea12 E08 25.0 m -107.36.48.9 +33.53.55.1
| |
| 11 ea13 N07 25.0 m -107.37.07.2 +33.54.12.9
| |
| 12 ea14 E05 25.0 m -107.36.58.4 +33.53.58.8
| |
| 13 ea15 W06 25.0 m -107.37.15.6 +33.53.56.4
| |
| 14 ea16 W02 25.0 m -107.37.07.5 +33.54.00.9
| |
| 15 ea17 W07 25.0 m -107.37.18.4 +33.53.54.8
| |
| 16 ea18 N09 25.0 m -107.37.07.8 +33.54.19.0
| |
| 17 ea19 W04 25.0 m -107.37.10.8 +33.53.59.1
| |
| 18 ea20 N05 25.0 m -107.37.06.7 +33.54.08.0
| |
| 19 ea21 E01 25.0 m -107.37.05.7 +33.53.59.2
| |
| 20 ea22 N04 25.0 m -107.37.06.5 +33.54.06.1
| |
| 21 ea23 E07 25.0 m -107.36.52.4 +33.53.56.5
| |
| 22 ea24 W05 25.0 m -107.37.13.0 +33.53.57.8
| |
| 23 ea25 N02 25.0 m -107.37.06.2 +33.54.03.5
| |
| 24 ea26 W03 25.0 m -107.37.08.9 +33.54.00.1
| |
| 25 ea27 E03 25.0 m -107.37.02.8 +33.54.00.5
| |
| 26 ea28 N08 25.0 m -107.37.07.5 +33.54.15.8
| |
| </pre>
| |
| Note that the spws are re-numbered to 0-13.
| |
| | |
| Prepare the MS for calibration by adding scratch columns. This will take a few minutes.
| |
| <source lang="python">
| |
| # In CASA
| |
| clearcal('SN2010FZ_flagged10s.ms')
| |
| </source>
| |
| | |
| === Setting the flux density scale ===
| |
| | |
| Before calibrating, we put a model for flux calibration source 3C286 into the MS (in the MODEL_DATA column we just created).
| |
| Have it set the flux on a per-channel basis.
| |
| NOTE: This uses the AOC Unix location for casapy models (yours may be different):
| |
| <source lang="python">
| |
| # In CASA
| |
| setjy(vis='SN2010FZ_flagged10s.ms',field='2',spw='',scalebychan=True, \
| |
| modimage='/usr/lib64/casapy/data/nrao/VLA/CalModels/3C286_C.im')
| |
| </source>
| |
| It reports to logger that its about 7.7Jy at lower end to 5.7Jy at upper frequency limit.
| |
| | |
| === Calibrating delays and bandpass ===
| |
| | |
| First, we do a phase-only calibration solution on a narrow range of channels in each spw on the bandpass/flux calibrator 3c286 to flatten them before solving for the bandpass. Note where we saw RFI in the higher spw and avoid those channels. The range 23~28 should work. Pick a refant near center - ea02 is a reasonable bet:
| |
| <source lang="python">
| |
| # In CASA
| |
| gaincal(vis='SN2010FZ_flagged10s.ms',caltable='calSN2010FZ.G0',field='2',spw='0~13:23~28', \
| |
| gaintype='G',refant='ea02',calmode='p',solint='int',minsnr=3)
| |
| </source>
| |
| | |
| [[Image:plotSN2010FZ_plotcal_G0p1.png|200px|thumb|right|plotcal G0 phase ant 0~15]]
| |
| [[Image:plotSN2010FZ_plotcal_G0p2.png|200px|thumb|right|plotcal G0 phase ant 16~26]]
| |
| Plot the phase solutions (using full phase range instead of autorange)
| |
| <source lang="python">
| |
| # In CASA
| |
| plotcal(caltable='calSN2010FZ.G0',xaxis='time',yaxis='phase',iteration='antenna', \
| |
| plotrange=[-1,-1,-180,180])
| |
| </source>
| |
| Step through the antenna-based solutions. They look good (and fairly flat over the scans).
| |
| NOTE: If you want to make single-page multipanel plots (like those shown to the right), particularly for
| |
| hardcopy (where it only shows the first page), you can do:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotcal(caltable='calSN2010FZ.G0',xaxis='time',yaxis='phase', \
| |
| antenna='0~10,12~15',subplot=531,iteration='antenna', \
| |
| plotrange=[-1,-1,-180,180],showgui=False,fontsize=6.0, \
| |
| figfile='plotSN2010FZ_plotcal_G0p1.png')
| |
| plotcal(caltable='calSN2010FZ.G0',xaxis='time',yaxis='phase', \
| |
| antenna='16~26',subplot=531,iteration='antenna', \
| |
| plotrange=[-1,-1,-180,180],showgui=False,fontsize=6.0, \
| |
| figfile='plotSN2010FZ_plotcal_G0p2.png')
| |
| </source>
| |
| | |
| We can now solve for the residual antenna-based delays that we saw in phase vs. frequency.
| |
| This uses the gaintype='K' option in gaincal (this is not documented but is available in
| |
| Version 3.2.1). Note that this currently does not do a "global fringe-fitting" solution for delays,
| |
| but instead does a baseline-based delay solution to all baselines to the refant, treating these
| |
| as antenna-based delays. In most cases with high-enough S/N to get baseline-based delay solutions
| |
| this will suffice.
| |
| <source lang="python">
| |
| # In CASA
| |
| gaincal(vis='SN2010FZ_flagged10s.ms',caltable='calSN2010FZ.K0',gaintable='calSN2010FZ.G0', \
| |
| field='2',spw='0~13:4~59',gaintype='K', \
| |
| refant='ea02',combine='scan',solint='inf',minsnr=3)
| |
| </source>
| |
| We pre-apply our initial phase table, and produce a new K-type caltable for input to bandpass calibration.
| |
| The delays found are sent to the terminal. They are up to a few nanoseconds.
| |
| | |
| [[Image:plotSN2010FZ_plotcal_B0ea14.png|200px|thumb|right|plotcal B0 amp phase vs. freq for ea01]]
| |
| Now solve for the bandpass using the previous tables:
| |
| <source lang="python">
| |
| # In CASA
| |
| bandpass(vis='SN2010FZ_flagged10s.ms',caltable='calSN2010FZ.B0', \
| |
| gaintable=['calSN2010FZ.G0','calSN2010FZ.K0'], \
| |
| field='2',refant='ea02',solnorm=True, \
| |
| bandtype='B', combine='scan', solint='inf')
| |
| </source>
| |
| Now plot this, in amplitude then phase:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotcal(caltable='calSN2010FZ.B0',xaxis='freq',yaxis='amp',iteration='antenna')
| |
| | |
| plotcal(caltable='calSN2010FZ.B0',xaxis='freq',yaxis='phase',iteration='antenna', \
| |
| plotrange=[-1,-1,-180,180])
| |
| </source>
| |
| In the bandpass phase you no longer see the residual antenna delays but there are some band edge effects.
| |
| Note that some antennas have a little strange bandpasses at upper end of lower baseband in spw 5,6,7
| |
| (e.g. ea14,ea16,ea17,ea25).
| |
| To plot amp and phase for a single antenna versus frequency (see plots at right):
| |
| <source lang="python">
| |
| # In CASA
| |
| plotcal(caltable='calSN2010FZ.B0',xaxis='freq',yaxis='amp', \
| |
| antenna='ea14',subplot=211,showgui=False, \
| |
| figfile='')
| |
| plotcal(caltable='calSN2010FZ.B0',xaxis='freq',yaxis='phase', \
| |
| antenna='ea14',subplot=212,showgui=False, \
| |
| plotrange=[-1,-1,-180,180],figfile='plotSN2010FZ_plotcal_B0ea14.png')
| |
| </source>
| |
| | |
| === Final phase and amplitude calibration ===
| |
| | |
| [[Image:plotSN2010FZ_plotcal_G1p1.png|200px|thumb|right|plotcal G1 phase ant 0~15]]
| |
| [[Image:plotSN2010FZ_plotcal_G1p2.png|200px|thumb|right|plotcal G1 phase ant 16~26]]
| |
| Now calibrate phases for real with wider bandwidth. First the flux calibrator again, with a per-integration solution time:
| |
| <source lang="python">
| |
| # In CASA
| |
| gaincal(vis='SN2010FZ_flagged10s.ms',caltable='calSN2010FZ.G1', \
| |
| gaintable=['calSN2010FZ.K0','calSN2010FZ.B0'], \
| |
| field='2',refant='ea02',solnorm=F, spw='0~13:4~59', \
| |
| solint='int',gaintype='G',calmode='p')
| |
| </source>
| |
| Next our phase calibrator, appending these solutions to previous table.
| |
| Exclude RFI channels here, and obtain one solution per scan:
| |
| <source lang="python">
| |
| # In CASA
| |
| gaincal(vis='SN2010FZ_flagged10s.ms',caltable='calSN2010FZ.G1',
| |
| gaintable=['calSN2010FZ.K0','calSN2010FZ.B0'], \
| |
| field='0',refant='ea02',solnorm=F, \
| |
| spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
| |
| solint='inf',gaintype='G',calmode='p',append=True)
| |
| </source>
| |
| The phases look reasonably connected:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotcal(caltable='calSN2010FZ.G1',xaxis='time',yaxis='phase',iteration='antenna', \
| |
| plotrange=[-1,-1,-180,180])
| |
| </source>
| |
| | |
| Now solve for amplitudes on a per scan interval. Do these separately using gainfield so phases don't get
| |
| transferred across fields. Uses linear interpolation of the previously determined phases by default.
| |
| Pre-apply the gaincurve also:
| |
| <source lang="python">
| |
| # In CASA
| |
| gaincal(vis='SN2010FZ_flagged10s.ms', caltable='calSN2010FZ.G2', \
| |
| gaintable=['calSN2010FZ.K0','calSN2010FZ.B0','calSN2010FZ.G1'], \
| |
| gainfield=['2','2','2'], field='2',refant='ea02',solnorm=F,
| |
| spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
| |
| solint='inf',gaintype='G',calmode='a',gaincurve=True)
| |
| | |
| gaincal(vis='SN2010FZ_flagged10s.ms', caltable='calSN2010FZ.G2', \
| |
| gaintable=['calSN2010FZ.K0','calSN2010FZ.B0','calSN2010FZ.G1'],\
| |
| gainfield=['2','2','0'], field='0',refant='ea02',solnorm=F, \
| |
| spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
| |
| solint='inf',gaintype='G',calmode='a',gaincurve=True,append=True)
| |
| </source>
| |
| | |
| [[Image:plotSN2010FZ_plotcal_F2a1.png|200px|thumb|right|plotcal F2 amp ant 0~15]]
| |
| [[Image:plotSN2010FZ_plotcal_F2a2.png|200px|thumb|right|plotcal F2 amp ant 16~26]]
| |
| Since the flux on the gain calibrator is not scaled to its correct flux (but to 1.0Jy)
| |
| use fluxscale to transfer the amplitude gains from 3c286:
| |
| <source lang="python">
| |
| # In CASA
| |
| fluxscale(vis='SN2010FZ_flagged10s.ms',caltable='calSN2010FZ.G2', \
| |
| fluxtable='calSN2010FZ.F2',reference='2',transfer='0')
| |
| </source>
| |
| As it so happens, the derived flux for J0925+0019 is about 1 Jy and flat spectrum (you can plot up the raw amplitudes for fields 0,2 and convince yourself this is indeed true and not a bug). Plot these:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotcal(caltable='calSN2010FZ.F2',xaxis='time',yaxis='amp',iteration='antenna')
| |
| </source>
| |
| The phase calibrator has consistent gains across the short run, though there is a slight offset to 3c286 (which is not nearby in any event).
| |
| | |
| == Applying the Calibration and Final Editing ==
| |
| | |
| Next we actually apply all our accumulated calibration tables. We apply these to the
| |
| calibration fields individually using the appropriate gainfields and interpolation for each:
| |
| <source lang="python">
| |
| # In CASA
| |
| applycal(vis='SN2010FZ_flagged10s.ms',field='2', \
| |
| gaintable=['calSN2010FZ.K0','calSN2010FZ.B0','calSN2010FZ.G1','calSN2010FZ.F2'], \
| |
| gainfield=['','','2','2'],interp=['nearest','nearest','linear','nearest'], \
| |
| parang=False,calwt=F,gaincurve=T)
| |
| </source>
| |
| [[Image:plotSN2010FZ_plotms_applied_fld2.png|200px|thumb|right|plotms cal applied fld2]]
| |
| For the nearby calibrator we did only scan-based phase solutions so we use nearest interpolation:
| |
| <source lang="python">
| |
| # In CASA
| |
| applycal(vis='SN2010FZ_flagged10s.ms',field='0', \
| |
| gaintable=['calSN2010FZ.K0','calSN2010FZ.B0','calSN2010FZ.G1','calSN2010FZ.F2'], \
| |
| gainfield=['','','0','0'], interp=['nearest','nearest','nearest','nearest'], \
| |
| parang=False,calwt=F,gaincurve=T)
| |
| </source>
| |
| Finally we apply calibration from field 0 to the target field 1. This takes a few minutes:
| |
| <source lang="python">
| |
| # In CASA
| |
| applycal(vis='SN2010FZ_flagged10s.ms',field='1', \
| |
| gaintable=['calSN2010FZ.K0','calSN2010FZ.B0','calSN2010FZ.G1','calSN2010FZ.F2'], \
| |
| gainfield=['','','0','0'], interp=['nearest','nearest','linear','linear'], \
| |
| parang=False,calwt=F,gaincurve=T)
| |
| </source>
| |
| | |
| We can examine the corrected data on 3c286 using our RFI mask from above and avoiding band edges
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_flagged10s.ms',field='2', \
| |
| spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
| |
| correlation='RR,LL',xaxis='frequency',yaxis='amp',ydatacolumn='corrected')
| |
| </source>
| |
| We now see problems in spw 5 and 6 for baseline ea17&ea25, which gives a really strange response:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_flagged10s.ms',field='2', \
| |
| spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
| |
| antenna='ea17&ea25', \
| |
| correlation='RR,LL',xaxis='frequency',yaxis='amp',ydatacolumn='corrected')
| |
| </source>
| |
| You can exclude this through antenna negation:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_flagged10s.ms',field='2', \
| |
| spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
| |
| antenna='!ea17&ea25', \
| |
| correlation='RR,LL',xaxis='frequency',yaxis='amp',ydatacolumn='corrected')
| |
| </source>
| |
| Then use Locate for the other bad points, which seem to indicate spw 5,6,7 for ea14,ea16,ea17,ea25.
| |
| Exclude these and replot:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_flagged10s.ms',field='2', \
| |
| spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
| |
| antenna='!ea14;!ea16;!ea17;!ea25', \
| |
| correlation='RR,LL',xaxis='frequency',yaxis='amp',ydatacolumn='corrected')
| |
| </source>
| |
| [[Image:plotSN2010FZ_plotms_appliedflags_fld2.png|200px|thumb|right|plotms cal applied flagged fld2]]
| |
| This now looks clean except for the RFI in the upper subbands.
| |
| | |
| Do flagging based on these:
| |
| <source lang="python">
| |
| # In CASA
| |
| flaglist = ['antenna="ea14,ea16,ea17,ea25" spw="5~7"']
| |
| flagcmd(vis='SN2010FZ_flagged10s.ms',flagmode='cmd',command=flaglist, \
| |
| optype='apply',flagbackup=False)
| |
| </source>
| |
| | |
| Now replot the corrected data (you may have to force reload if you plotted same thing right before this):
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_flagged10s.ms',field='2', \
| |
| spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
| |
| correlation='RR,LL',xaxis='frequency',yaxis='amp',ydatacolumn='corrected')
| |
| </source>
| |
| Looks good. If we were more diligent we would go back and recalibrate, but this looks good enough for now.
| |
| | |
| Plot the phase:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_flagged10s.ms',field='2', \
| |
| spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
| |
| correlation='RR,LL',xaxis='frequency',yaxis='phase',ydatacolumn='corrected')
| |
| </source>
| |
| [[Image:plotSN2010FZ_plotms_appliedflags_fld2_phase.png|200px|thumb|right|plotms cal applied flagged fld2 phase]]
| |
| Note the characteristic "bowtie" pattern of the phases about the sub-band centers.
| |
| Here we can see the effect of the EVLA "delay clunking", where the delay steps through discrete values such that
| |
| the phase goes from -11deg to +11deg across the sub-band as the delay changes due to geometry. This is D-configuration so the delays change slowly, it will change faster in wider configurations. As of Q3 2011 we have not enabled the corrections for this
| |
| in the EVLA system so you will always have this remaining delay error in your data. In principle you could solve for delays on short timescales and take this out, in practice this in not possible for your weaker target source in any event.
| |
| | |
| Now lets plot the corrected data amplitude for the phase calibrator:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_flagged10s.ms',field='0', \
| |
| spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
| |
| correlation='RR,LL',xaxis='frequency',yaxis='amp',ydatacolumn='corrected')
| |
| </source>
| |
| [[Image:plotSN2010FZ_plotms_appliedflags_fld0_amp.png|200px|thumb|right|plotms cal applied flagged fld0 amp]]
| |
| [[Image:plotSN2010FZ_plotms_appliedflags_fld0_phase.png|200px|thumb|right|plotms cal applied flagged fld0 phase]]
| |
| You can see the roll-off pumping up noise at the baseband edges (about 8-16 channels worth). Also, we can see some RFI we missed:
| |
| * <6804MHz spw 8 below ch 30 lots of bad stuff (alot from ea18,ea22 but others too)
| |
| * 7168MHz spw 11 ch 20
| |
| * pretty much all of spw 12,13
| |
| The ch 20 ones are all harmonics of the notorious 128MHz tone.
| |
| | |
| We will not flag these, but exclude them in imaging, so a good channel selection string for imaging might be:
| |
| <pre>
| |
| spw = '0:16~59,1~6:4~59,7:4~54,8:30~59,9~10:4~59,11:4~19;21~59'
| |
| </pre>
| |
| | |
| We can also plot the corrected phase - looks fairly good
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_flagged10s.ms',field='0', \
| |
| spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
| |
| correlation='RR,LL',xaxis='frequency',yaxis='phase',ydatacolumn='corrected')
| |
| </source>
| |
| | |
| We can average over baseline and each scan:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_flagged10s.ms',field='0', \
| |
| spw='0:10~59,1~7:4~59,8:4~13;18~59,9~11:4~59,12:4~13;18~29;31~33;46~51;53~59,13:4~8;15~36;42~59', \
| |
| correlation='RR,LL',avgbaseline=True,avgtime='600s',
| |
| xaxis='frequency',yaxis='phase',ydatacolumn='corrected')
| |
| </source>
| |
| [[Image:plotSN2010FZ_plotms_appliedflags_fld0_phaseavg.png|200px|thumb|right|plotms cal applied flagged fld0 phase averaged]]
| |
| In this case we can see the residual effect of the EVLA "delay clunking" described above. It is reduced due to the averaging
| |
| that we applied, but it is still there.
| |
| | |
| [[Image:plotSN2010FZ_plotms_appliedflags_fld1.png|200px|thumb|right|plotms cal applied flagged fld1 amp averaged]]
| |
| You can look at the target source field='1' but there are lots of data so you will need to do alot of averaging.
| |
| For example:
| |
| <source lang="python">
| |
| # In CASA
| |
| plotms(vis='SN2010FZ_flagged10s.ms',field='1',avgtime='300s', \
| |
| spw='0:16~59,1~6:4~59,7:4~54,8:30~59,9~10:4~59,11:4~19;21~59', \
| |
| correlation='RR,LL',xaxis='frequency',yaxis='amp',ydatacolumn='corrected')
| |
| </source>
| |
| Alas, the upper baseband still has lots of low level RFI.
| |
| | |
| Now split off the data for calibrators and target, to avoid later issues that can corrupt the MSes.
| |
| These will take a bit of time...field 1 took 18min on my workstation.
| |
| <source lang="python">
| |
| # In CASA
| |
| split(vis='SN2010FZ_flagged10s.ms',outputvis='SN2010FZ_split10s.ms',datacolumn='corrected',field='1')
| |
| | |
| split(vis='SN2010FZ_flagged10s.ms',outputvis='SN2010FZ_3c28610s.ms',datacolumn='corrected',field='2')
| |
| | |
| split(vis='SN2010FZ_flagged10s.ms',outputvis='SN2010FZ_J092510s.ms',datacolumn='corrected',field='0')
| |
| </source>
| |
| | |
| == Imaging ==
| |
| | |
| This is D-config data at C-band, look at the Obs Status Summary:
| |
| [http://evlaguides.nrao.edu/index.php?title=Observational_Status_Summary_-_Current]
| |
| Synthesized beam should be 12" at 6GHz with primary beam FOV 7.5arcmin (450").
| |
| Our data spans 4.5-7.5GHz : beam 9.6 at 7.5GHz and FOV 10' at 4.5GHz
| |
| A cellsize of 3" should work, with an imsize > 200 to cover 2xFWHM FOV
| |
| An imsize of 400+ will put the main beam inside the inner quarter safely
| |
| The Briggs robust (0.5) weighting is somewhere between uniform and natural
| |
| and will give reasonable resolution but still see some larger scale structure.
| |
| | |
| Due to the numerology of FFTW (which clean uses underneath for FFTs) optimal sizes,
| |
| imsize should be composite number with two and only two prime factors chosen from
| |
| 2,3,5. Taking into account the x1.2 padding that clean uses internally to the imsize
| |
| you give it (and 1.2=2*3/5), we choose 640 or 1280 as our imsize (640=2^7*5). Other
| |
| reasonable sets would be 405,1215,etc. (405=3^4*5) or 432,648,1296 (these are 2^n*3^m*5).
| |
| In practice, if you give it non-optimal values for imsize, you may find that the transforms
| |
| take a bit longer, which is noticeable if you are doing interactive clean.
| |
| | |
| Lets interactively clean one of the lower baseband spw (5):
| |
| NOTE: this first time will take a few minutes at start to create scratch columns
| |
| in the MS in case we want to do self-calibration later.
| |
| | |
| === Cleaning a single spectral window ===
| |
| | |
| [[Image:viewSN2010FZ_spw5_clean640.png|200px|thumb|right|clean spw5 640x640]]
| |
| [[Image:viewSN2010FZ_spw5_clean1280.png|200px|thumb|right|clean spw5 1280x1280 before clean]]
| |
| [[Image:viewSN2010FZ_spw5_clean1280final.png|200px|thumb|right|clean spw5 1280x1280 restored image]]
| |
| Start carefully by boxing the bright source and setting iterations to 10 at first
| |
| Gradually add more boxes and increase the number of iterations
| |
| Since this is not much more than a snapshot you see the six-fold sidelobe pattern
| |
| of the extended emission in the center of the map. This decreases as you clean
| |
| out this emission.
| |
| | |
| Stop cleaning when the residuals look like noise (and you cannot clearly see sources).
| |
| To stop click the red X button.
| |
| <source lang="python">
| |
| # In CASA
| |
| clean(vis='SN2010FZ_split10s.ms',spw='5:4~59', \
| |
| imagename='imgSN2010FZ10s_spw5_clean640', \
| |
| mode='mfs',nterms=1,niter=10000,gain=0.1,threshold='0.0mJy', \
| |
| psfmode='clark',imsize=[640,640],cell=['3.0arcsec'],stokes='I', \
| |
| imagermode='csclean', cyclefactor=1.5, \
| |
| weighting='briggs',robust=0.5,calready=True,interactive=True)
| |
| </source>
| |
| The top figure to the right shows a zoom in on the end state of the clean, where
| |
| we have marked a number of boxes and cleaned them out.
| |
| | |
| Note that there are some strange sidelobe patterns in lower left, possibly
| |
| from a source outside the image area. We can make a bigger image starting from
| |
| our current model:
| |
| <source lang="python">
| |
| # In CASA
| |
| clean(vis='SN2010FZ_split10s.ms',spw='5:4~59', \
| |
| imagename='imgSN2010FZ10s_spw5_clean1280', \
| |
| mode='mfs',nterms=1,niter=10000,gain=0.1,threshold='0.0mJy', \
| |
| psfmode='clark',imsize=[1280,1280],cell=['3.0arcsec'],stokes='I', \
| |
| imagermode='csclean', cyclefactor=1.5, \
| |
| modelimage='imgSN2010FZ10s_spw5_clean640.model', \
| |
| weighting='briggs',robust=0.5,calready=True,interactive=True)
| |
| </source>
| |
| Sure enough, there is a bright source near the lower left (see middle panel at right).
| |
| Box it, clean it a bit, and look again. There is a second source in the mid-left (track
| |
| it down by its sidelobes). Box this one, clean it a bit, and when satisfied stop.
| |
| <source lang="python">
| |
| # In CASA
| |
| viewer('imgSN2010FZ10s_spw5_clean1280.image')
| |
| </source>
| |
| The restored image is shown in bottom panel to the right. I have chosen the Grayscale1 instead of default color
| |
| map as I prefer grayscale to false color for assessing image quality.
| |
| Check the rms of the residuals using the {{imstat}} task:
| |
| <source lang="python">
| |
| # In CASA
| |
| mystat = imstat('imgSN2010FZ10s_spw5_clean1280.residual')
| |
| print 'Residual standard deviation = '+str(mystat['sigma'][0])
| |
| </source>
| |
| I got 31.8uJy for mine.
| |
| | |
| === Cleaning the lower baseband ===
| |
| | |
| [[Image:viewSN2010FZ_spw0to7_clean1280final.png|200px|thumb|right|clean spw0-7 restored image center]]
| |
| Now, image the entire lower baseband (spw 0-7).
| |
| Follow same iterative procedure as before, and get the best
| |
| residuals you can. Because of the bandwidth and frequency
| |
| synthesis, the sidelobe pattern is different than before and
| |
| it is much easier to see fainter emission.
| |
| Be careful cleaning sources that lie near or on sidelobe splotches.
| |
| Clean the central emission region way down first to reduce the sidelobe
| |
| level before adding components in the sidelobe areas.
| |
| <source lang="python">
| |
| # In CASA
| |
| clean(vis='SN2010FZ_split10s.ms',spw='0:16~59,1~6:4~59,7:4~54', \
| |
| imagename='imgSN2010FZ10s_spw0to7_clean1280', \
| |
| mode='mfs',nterms=1,niter=10000,gain=0.1,threshold='0.0mJy', \
| |
| psfmode='clark',imsize=[1280,1280],cell=['3.0arcsec'],stokes='I', \
| |
| imagermode='csclean', cyclefactor=1.5, \
| |
| weighting='briggs',robust=0.5,calready=True,interactive=True)
| |
| | |
| mystat = imstat('imgSN2010FZ10s_spw0to7_clean1280.residual')
| |
| print 'Residual standard deviation = '+str(mystat['sigma'][0])
| |
| </source>
| |
| I got 11.3uJy (and there is clearly structure left in the residual). To the right is a zoom-in on the center
| |
| of the restored image.
| |
| | |
| [[Image:viewSN2010FZ_spw0to7_mfs2clean.png|200px|thumb|right|clean spw0-7 mfs nterms=2 in progress]]
| |
| [[Image:viewSN2010FZ_spw0to7_mfs2loadalpha.png|200px|thumb|right|clean spw0-7 mfs nterms=2 load alpha with LEL]]
| |
| [[Image:viewSN2010FZ_spw0to7_mfs2panelalpha.png|200px|thumb|right|clean spw0-7 mfs nterms=2 tt0 and alpha]]
| |
| Lets try adding a spectral slope to the multi-frequency synthesis using nterms=2 on the lower baseband.
| |
| This will solve for images of the average and spectral slope simultaneously.
| |
| The dirty beam will have lower sidelobes so we turn up cyclefactor for csclean a bit:
| |
| <source lang="python">
| |
| # In CASA
| |
| clean(vis='SN2010FZ_split10s.ms',spw='0:16~59,1~6:4~59,7:4~54', \
| |
| imagename='imgSN2010FZ10s_spw0to7_mfs2_clean1280', \
| |
| mode='mfs',nterms=2,niter=10000,gain=0.1,threshold='0.0mJy', \
| |
| psfmode='clark',imsize=[1280,1280],cell=['3.0arcsec'],stokes='I', \
| |
| imagermode='csclean', cyclefactor=4.5, \
| |
| weighting='briggs',robust=0.5,calready=True,interactive=True)
| |
| | |
| mystat = imstat('imgSN2010FZ10s_spw0to7_mfs2_clean1280.residual.tt0')
| |
| print 'Residual standard deviation = '+str(mystat['sigma'][0])
| |
| </source>
| |
| I got 10.5uJy (somewhat better looking than the nterms=1).
| |
| The top screenshot to the right shows an intermediate but early stage of cleaning where we are looking at
| |
| the central emission and cleaning it out slowly.
| |
| | |
| ==== Cleaning the lower baseband using two MFS Taylor terms ====
| |
| | |
| The mfs nterms=2 option creates two "Taylor Term" images - an average intensity image (with suffix <tt>.image.tt0</tt>)
| |
| and a spectral slope image (with suffix <tt>.image.tt1</tt>) which is intensity x alpha (where alpha is spectral index).
| |
| For convenience there is a spectral index image (with suffix <tt>.image.alpha</tt>). These Taylor expansions are with respect to the "Reference Frequency" of the image (by default the center frequency of the spw selected, but can be specified using the
| |
| reffreq parameter in clean). The convention for spectral index alpha is that
| |
| | |
| <math>
| |
| S \propto \nu^\alpha
| |
| </math>
| |
| | |
| so negative spectral indexes indicate a "steep" spectrum (falling with frequency).
| |
| | |
| You can use the {{viewer}} to load the average intensity
| |
| <source lang="python">
| |
| # In CASA
| |
| viewer('imgSN2010FZ10s_spw0to7_mfs2_clean1280.image.tt0')
| |
| </source>
| |
| and then use the Open Data panel to load <tt>imgSN2010FZ10s_spw0to7_mfs2_clean1280.image</tt>
| |
| which can then be blinked (optionally plotted side-by-side using the Panel Display Options
| |
| panel to set 2 panels in the x direction). Note there is lots of noise in the low-intensity
| |
| regions, and thus filtering the alpha image based on the values in the tt0 image is desirable.
| |
| | |
| You can use the {{immath}} task to make this filtered alpha image explicitly, using a
| |
| Lattice Expression Language (LEL) expression:
| |
| <source lang="python">
| |
| # In CASA
| |
| immath(imagename=['imgSN2010FZ10s_spw0to7_mfs2_clean1280.image.alpha',
| |
| 'imgSN2010FZ10s_spw0to7_mfs2_clean1280.image.tt0'],
| |
| mode='evalexpr',
| |
| expr='IM0[IM1>5.0E-5]',
| |
| outfile='imgSN2010FZ10s_spw0to7_mfs2_clean1280.image.alpha.filtered')
| |
| </source>
| |
| This will use 50uJy (or 5 x the sigma we found) as the cutoff.
| |
| You can then view or manipulate the filtered alpha image as normal.
| |
| | |
| We can also use LEL to filter the alpha image on the intensity on-the-fly when we load this raster in the Open Data panel
| |
| by specifying a LEL string in the LEL box instead of selecting the image from the directory
| |
| listing. The LEL string:
| |
| <pre>
| |
| 'imgSN2010FZ10s_spw0to7_mfs2_clean1280.image.alpha'['imgSN2010FZ10s_spw0to7_mfs2_clean1280.image.tt0'>5.0E-05]
| |
| </pre>
| |
| will replicate what we did above. The middle figure to the right shows the Open Data panel
| |
| with our LEL string in it. Just click the Raster button to load this.
| |
| | |
| The lower panel to the right shows the intensity and LEL-filtered alpha images side-by-side in the viewer, zoomed
| |
| in on the galaxy emission. Mousing over the alpha shows spectral indexes ranging from -1 to +1 in the center, with
| |
| the brightest emission with alpha -0.7 in the knots in the disk.
| |
| | |
| === Cleaning using both basebands combined ===
| |
| | |
| For the ultimate image, use the "clean" part of the upper baseband in addition
| |
| to the lower (use spw 0-11). We will use mfs with nterms=2 (if you try nterms=1
| |
| on this wide bandwidth you will get much poorer residuals). Because of the added
| |
| work and extra data involved, this will take much longer than our other runs of
| |
| clean. Therefore, we will get a head start by doing a non-interactive clean using
| |
| the mask left from the previous clean (spw 0-7). We will insert a clean threshold
| |
| to limit runaway cleaning too far beneath the noise level. This took 40min on my
| |
| workstation.
| |
| <source lang="python">
| |
| # In CASA
| |
| clean(vis='SN2010FZ_split10s.ms', \
| |
| spw='0:16~59,1~6:4~59,7:4~54,8:30~59,9~10:4~59,11:4~19;21~59', \
| |
| imagename='imgSN2010FZ10s_spw0to11_mfs2_clean1280', \
| |
| mode='mfs',nterms=2,niter=3000,gain=0.1,threshold='0.002mJy', \
| |
| psfmode='clark',imsize=[1280,1280],cell=['3.0arcsec'],stokes='I', \
| |
| imagermode='csclean', cyclefactor=4.5, \
| |
| mask=['imgSN2010FZ10s_spw0to7_mfs2_clean1280.mask'], \
| |
| weighting='briggs',robust=0.5,calready=True,interactive=False)
| |
| | |
| mystat = imstat('imgSN2010FZ10s_spw0to11_mfs2_clean1280.residual.tt0')
| |
| print 'Residual standard deviation = '+str(mystat['sigma'][0])
| |
| </source>
| |
| I got 8.6uJy for this effort (noticeably better than the lower baseband only results).
| |
| | |
| [[Image:viewSN2010FZ spw0to11_mfs2resid.png|200px|thumb|right|final residual and mask]]
| |
| Let us see if there is more to clean. Bring this up in interactive mode:
| |
| <source lang="python">
| |
| # In CASA
| |
| clean(vis='SN2010FZ_split10s.ms', \
| |
| spw='0:16~59,1~6:4~59,7:4~54,8:30~59,9~10:4~59,11:4~19;21~59', \
| |
| imagename='imgSN2010FZ10s_spw0to11_mfs2_clean1280', \
| |
| mode='mfs',nterms=2,niter=3000,gain=0.1,threshold='0.001mJy', \
| |
| psfmode='clark',imsize=[1280,1280],cell=['3.0arcsec'],stokes='I', \
| |
| imagermode='csclean', cyclefactor=4.5, \
| |
| weighting='briggs',robust=0.5,calready=True,interactive=True)
| |
| </source>
| |
| I found a few more sources revealed in the outer parts of the image, and also more
| |
| emission around the galaxy disk in the center. So I drew new boxes, extended the box
| |
| in the center, and did about 1000 new iterations. At the end, what was left was dominated
| |
| by the error patterns from mis-calibration. Only self-calibration will get rid of
| |
| these. Stop cleaning for now. See the figure to the right for the interactive display panel
| |
| showing final residuals and mask (changing the colormap to <tt>Greyscale 1</tt>).
| |
| | |
| Check the residual levels:
| |
| <source lang="python">
| |
| # In CASA
| |
| mystat = imstat('imgSN2010FZ10s_spw0to11_mfs2_clean1280.residual.tt0')
| |
| sigma = mystat['sigma'][0]
| |
| print 'Residual standard deviation = '+str(mystat['sigma'][0])
| |
| </source>
| |
| Now I get 8.3uJy.
| |
| | |
| == Analyzing the image and comparing with the Optical/Infrared ==
| |
| | |
| Lets see how close we got to expected noise and dynamic range.
| |
| <source lang="python">
| |
| # In CASA
| |
| mystat = imstat('imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt0')
| |
| peak = mystat['max'][0]
| |
| print 'Image max flux = '+str(mystat['max'][0])
| |
| | |
| mystat = imstat('imgSN2010FZ10s_spw0to11_mfs2_clean1280.model.tt0')
| |
| total = mystat['sum'][0]
| |
| print 'Model total flux = '+str(mystat['sum'][0])
| |
| | |
| snr = peak/sigma
| |
| print 'SN2010FZ peak S/N = '+str(snr)
| |
| | |
| snr = total/sigma
| |
| print 'SN2010FZ total S/N = '+str(snr)
| |
| </source>
| |
| The output gives:
| |
| <pre>
| |
| Residual standard deviation = 8.33040657806e-06
| |
| Image max flux = 0.00995613634586
| |
| Model total flux = 0.0329372803525
| |
| SN2010FZ peak S/N = 1195.15611304
| |
| SN2010FZ total S/N = 3953.86228077
| |
| </pre>
| |
| | |
| What do we expect? If we do {{listobs}} on this MS we see the scans:
| |
| <pre>
| |
| Date Timerange (UTC) Scan FldId FieldName nVis Int(s)
| |
| 11-Jul-2010/21:38:44.0 - 21:39:51.0 9 0 SN2010FZ 39312 9.11
| |
| 21:40:01.0 - 21:41:20.5 10 0 SN2010FZ 44226 9.89
| |
| 21:41:30.0 - 21:42:50.0 11 0 SN2010FZ 44226 10
| |
| 21:43:00.0 - 21:44:20.0 12 0 SN2010FZ 44226 10
| |
| 21:44:30.0 - 21:45:50.0 13 0 SN2010FZ 44226 10
| |
| 21:46:00.0 - 21:47:19.5 14 0 SN2010FZ 44226 9.89
| |
| 21:47:29.0 - 21:47:48.5 15 0 SN2010FZ 14742 9.67
| |
| 21:49:42.0 - 21:50:49.0 17 0 SN2010FZ 39312 9.11
| |
| 21:50:59.0 - 21:52:19.0 18 0 SN2010FZ 44226 10
| |
| 21:52:29.0 - 21:53:48.5 19 0 SN2010FZ 44226 9.89
| |
| 21:53:58.0 - 21:55:18.0 20 0 SN2010FZ 44226 10
| |
| 21:55:28.0 - 21:56:48.0 21 0 SN2010FZ 44226 10
| |
| 21:56:58.0 - 21:58:18.0 22 0 SN2010FZ 44226 10
| |
| 21:58:28.0 - 21:58:47.0 23 0 SN2010FZ 14742 9.33
| |
| 22:00:40.0 - 22:01:47.0 25 0 SN2010FZ 39312 9.12
| |
| 22:01:57.0 - 22:03:17.0 26 0 SN2010FZ 44226 10
| |
| 22:03:27.0 - 22:04:47.0 27 0 SN2010FZ 44226 10
| |
| 22:04:57.0 - 22:06:16.5 28 0 SN2010FZ 44226 9.89
| |
| 22:06:26.0 - 22:07:46.0 29 0 SN2010FZ 44226 10
| |
| 22:07:56.0 - 22:09:16.0 30 0 SN2010FZ 44226 10
| |
| 22:09:26.0 - 22:09:45.0 31 0 SN2010FZ 14742 9.33
| |
| 22:11:38.5 - 22:12:45.5 33 0 SN2010FZ 39312 9.11
| |
| 22:12:55.0 - 22:14:15.0 34 0 SN2010FZ 44226 10
| |
| 22:14:25.0 - 22:15:45.0 35 0 SN2010FZ 44226 10
| |
| 22:15:55.0 - 22:17:15.0 36 0 SN2010FZ 44226 10
| |
| 22:17:25.0 - 22:18:44.5 37 0 SN2010FZ 44226 9.89
| |
| 22:18:54.0 - 22:20:14.0 38 0 SN2010FZ 44226 10
| |
| 22:20:24.0 - 22:20:43.5 39 0 SN2010FZ 14742 9.6
| |
| (nVis = Total number of time/baseline visibilities per scan)
| |
| </pre>
| |
| (listing columns truncated) and we estimate about 37min on target. We had
| |
| about 25 antennas on average, and our spw selection picked out
| |
| 610 channels (2MHz each) for a total of 1220MHz bandwidth. If we plug this
| |
| into the
| |
| [ https://science.nrao.edu/facilities/evla/calibration-and-tools/exposure EVLA exposure calculator]
| |
| then we find that we expect a rms thermal noise level of 7.6uJy so we are close!
| |
| | |
| [[Image:plotSN2010FZ viewerfinal.png|200px|thumb|right|final image]]
| |
| Look at this in the viewer:
| |
| <source lang="python">
| |
| # In CASA
| |
| viewer('imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt0')
| |
| </source>
| |
| Zoom in on the center (see figure to the right).
| |
| | |
| [[Image:viewSN2010FZ mfs2tt1.png|200px|thumb|right|final tt1 image with box]]
| |
| In the previous section we demonstrated how to process and display the spectral index image. You can do
| |
| the same for this final image. Here, we will do some rough analysis on the spectral index to determine
| |
| an intensity-weighted mean spectral index over the core region.
| |
| The <tt>.image.tt1</tt> from our mfs is an intensity times alpha image. See the figure to the right.
| |
| Lets gate the Taylor-term images on intensity as before:
| |
| <source lang="python">
| |
| # In CASA
| |
| immath(imagename=['imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt1',
| |
| 'imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt0'],
| |
| mode='evalexpr',
| |
| expr='IM0[IM1>5.0E-5]',
| |
| outfile='imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt1.filtered')
| |
| | |
| immath(imagename=['imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt0'],
| |
| mode='evalexpr',
| |
| expr='IM0[IM0>5.0E-5]',
| |
| outfile='imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt0.filtered')
| |
| </source>
| |
| | |
| We can identify a box containing the central emission (see figure of tt1 in viewer) and note the corners.
| |
| (We could also use the region tools from the viewer, but that is for another exercise.)
| |
| Let us compute the intensity-weighted spectral index over this box by averaging
| |
| these masked images using {{imstat}} and computing the ratio:
| |
| <source lang="python">
| |
| # In CASA
| |
| mystat = imstat('imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt1.filtered',
| |
| box='503,533,756,762')
| |
| avgtt0alpha = mystat['mean'][0]
| |
| | |
| mystat = imstat('imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt0.filtered',
| |
| box='503,533,756,762')
| |
| avgtt0 = mystat['mean'][0]
| |
| avgalpha = avgtt0alpha/avgtt0
| |
| print 'SN2010FZ I-weighted Alpha = '+str(avgalpha)
| |
| </source>
| |
| We get
| |
| <pre>
| |
| SN2010FZ I-weighted Alpha = -1.54687681754
| |
| </pre>
| |
| The emission in this source is on the steep side. At this point we do not know how reliable this is or
| |
| what we expect (though our calibrators come out with correct spectral indexes if we image them the
| |
| same way). But this illustrates a way to extract spectral information from our wideband mfs images.
| |
| | |
| As a final comparison, here is our image side by side with a i-band image from the Sloan Digital Sky Survey (SDSS) registered
| |
| to our image:
| |
| [[Image:plotSN2010FZ viewerfinalandSDSS.png|600px|thumb|center|final and sdss image]]
| |
| You can also find this image and load it into your viewer, and blink against our 6cm image.
| |
| | |
| We can also plot one as a raster and the other overlaid as contours. You can load the SDSS image
| |
| from the viewer Load Data panel and fiddle with contours. Once you know contour levels, you can
| |
| also use the imview task to load a raster and contour image:
| |
| <source lang="python">
| |
| # In CASA
| |
| imview(raster={ 'file' : 'imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt0'},
| |
| contour = { 'file' : 'NGC_2967_UGC_5180_IRAS_09394+0033-i.fits',
| |
| 'levels' : [0.2, 0.5, 1, 1.5, 3],
| |
| 'base' : 0.0,
| |
| 'unit' : 1.0 } )
| |
| </source>
| |
| The figure below shows the SDSS contours overlaid on our 6cm image (after fiddling with the
| |
| colormap shift/slope for the EVLA raster image).
| |
| [[Image:viewSN2010FZ_spw0to11_mfs2tt0plusSDSS.png|400px|thumb|center|6cm EVLA raster plus SDSS i-band contours]]
| |
| | |
| Likewise, we can plot the SDSS image as a raster and overlay EVLA 6cm contours:
| |
| <source lang="python">
| |
| # In CASA
| |
| imview(raster={ 'file' : 'NGC_2967_UGC_5180_IRAS_09394+0033-i.fits',
| |
| 'scaling' : -2.0,
| |
| 'range' : [0,10] },
| |
| contour = { 'file' : 'imgSN2010FZ10s_spw0to11_mfs2_clean1280.image.tt0',
| |
| 'levels' : [0.04, 0.08, 0.16, 0.32, 0.64, 1.28, 2.56],
| |
| 'base' : 0.0,
| |
| 'unit' : 0.001 },
| |
| zoom = { 'blc' : [397,300],
| |
| 'trc' : [1567,1231] } )
| |
| </source>
| |
| This is shown in the figure below. Is the compact 6cm emission in upper left associated with a
| |
| spiral arm?
| |
| [[Image:viewSN2010FZ_spw0to11_SDSSiplusEVLA6cm.png|400px|thumb|center|SDSS i-band raster plus EVLA 6cm contours]]
| |
| | |
| == What to do next: some exercises for the user ==
| |
| | |
| Here are a number of things you can try after completing this tutorial:
| |
| | |
| # Use self-calibration to improve the data and re-clean to make a better image.
| |
| # Use multi-scale clean by adding non-zero scales to the <tt>multiscale</tt> parameter.
| |
| # Image the calibrators. What sort of dynamic range can you get on them? Is self-calibration needed (and if so what dynamic range do you get when you use it)?
| |
| # Try the <tt>testautoflag</tt> task (in 3.3.0 and later) to automatically flag RFI from the upper sideband.
| |
| | |
| == Credits ==
| |
| | |
| EVLA data taken by A. Soderberg et al. See
| |
| [https://science.nrao.edu/enews/3.8/index.shtml#evlanoise NRAO eNews 3.8] (1-Sep-2010) for more on this result.
| |
| | |
| SDSS image courtesy David Hogg & Michael Blanton, private communication. Data comes from
| |
| SDSS DR7, see [http://adsabs.harvard.edu/abs/2009ApJS..182..543A Abazajian et. al 2009].
| |
| | |
| <blockquote><i>
| |
| Funding for the SDSS and SDSS-II has been provided by the Alfred P. Sloan Foundation, the Participating Institutions, the National Science Foundation, the U.S. Department of Energy, the National Aeronautics and Space Administration, the Japanese Monbukagakusho, the Max Planck Society, and the Higher Education Funding Council for England. The SDSS Web Site is [http://www.sdss.org/].
| |
| | |
| The SDSS is managed by the Astrophysical Research Consortium for the Participating Institutions. The Participating Institutions are the American Museum of Natural History, Astrophysical Institute Potsdam, University of Basel, University of Cambridge, Case Western Reserve University, University of Chicago, Drexel University, Fermilab, the Institute for Advanced Study, the Japan Participation Group, Johns Hopkins University, the Joint Institute for Nuclear Astrophysics, the Kavli Institute for Particle Astrophysics and Cosmology, the Korean Scientist Group, the Chinese Academy of Sciences (LAMOST), Los Alamos National Laboratory, the Max-Planck-Institute for Astronomy (MPIA), the Max-Planck-Institute for Astrophysics (MPA), New Mexico State University, Ohio State University, University of Pittsburgh, University of Portsmouth, Princeton University, the United States Naval Observatory, and the University of Washington.
| |
| </i></blockquote>
| |
