Template:Checked 6.5.4: Difference between revisions

From CASA Guides
Jump to navigationJump to search
No edit summary
(Undo revision 36494 by Tcandela (talk))
Tags: Replaced Undo
 
Line 2: Line 2:
Last checked on CASA Version 6.5.4
Last checked on CASA Version 6.5.4
</div>
</div>
<!--By Bjorn Emonts-->
<b>This CASA Guide is for Version 6.5.2 of CASA.</b>
{| class="wikitable"
|-
| style="background: red; color: white;" | '''DISCLAIMER'''
|-
| The following guide is for users who are experts in data reduction using CASA. If you are a beginner or novice user, please review other CASA guides found at [https://casaguides.nrao.edu/index.php/Karl_G._Jansky_VLA_Tutorials VLA Tutorials]. <b> Note: This guide may take up to 2-days to complete </b>
|}
== Overview ==
This tutorial describes how to use CASA 6.5.2 to reduce spectral-line data in the low-frequency P-band of the VLA (230&ndash;470 MHz). The goal is to make an image cube containing HI 21cm absorption against the strong radio continuum of gravitationally lensed radio galaxy MG0414+0534. As a result of the high redshift of z=2.6365, the HI absorption signal in MG0414+0534 is redshifted to an observed frequency of 390.597 MHz.  The HI absorption in MG0414+0534 was previously imaged with the VLA by Moore, Carilli & Menten 1999 ([http://iopscience.iop.org/article/10.1086/311818/meta ApJ, 510, 87]), before the upgrade to the WIDAR system.
We would like to note that the CASAviewer (supporting tasks viewer, imview and msview) has not been maintained for a few years now and will be removed from future versions of CASA. The NRAO replacement visualization tool for images and cubes is CARTA, the “Cube Analysis and Rendering Tool for Astronomy”. It is available from the [https://cartavis.org/ CARTA website]. We strongly recommend to use CARTA, as it provides a much more efficient, stable, and feature rich user experience. A comparison of the CASAviewer and CARTA, as well as instructions on how to use CARTA at NRAO is provided in the respective section of the [https://casadocs.readthedocs.io/en/stable/notebooks/carta.html CASA docs]. This tutorial shows Figures generated with CARTA for visualization.
=== Observing strategy ===
To perform P-band spectrsocopy, there are three important considerations for planning the observations:
* Use a bandpass calibrator that is strong enough to accurately calibrate the frequency-dependent gain variations. This is particularly important for most HI 21cm absorption projects, which are typically performed against strong radio continuum sources. As a rule-of-thumb, use ''t''<sub>cal</sub> > ''t''<sub>obj</sub> × (''S''<sub>obj</sub> / ''S''<sub>cal</sub>)<sup>2</sup>, with ''t'' the exposure time and ''S'' the source flux density. This rule-of-thumb serves to avoid introducing excessive additional noise in the spectrum.
* Additionally, if very high spectral dynamic range is needed (i.e., when expecting a ratio between the detection limit and the radio continuum of about 1:10,000 or more), consider observing a bandpass calibrator several times during your run to be able to correct for time-varying bandpass changes, which scale with the continuum emission in the target field. See the [https://science.nrao.edu/facilities/vla/docs/manuals/obsguide/calibration Calibration section] of the Guide to Observing with the VLA for more information.
* Use a bandwidth that is wide enough to perform accurate self calibration. Using a wide bandwidth for self-calibration is important for fields with relatively weak continuum sources. For strong continuum sources, a narrower bandwidth can be used to avoid excessive RFI.
The P-band test data on MG0414+0534 that we use in this tutorial were obtained using a large bandwidth. This is to ensure that good bandpass solutions can be procured, and that self-calibration can accurately be performed. However, due to the strong radio continuum of MG0414+0534 (3.3 Jy at 390 MHz), and the large amounts of RFI across the entire band, we only use a small fraction of the total band for data reduction and analysis in this tutorial.
== Obtaining the data ==
<font color=red>'''This 33 GB data set can be downloaded by clicking [https://casa.nrao.edu/Data/EVLA/MG0414/MG0414_d1_data.ms.tgz here].'''</font> The filename should be MG0414_d1_data.ms.tgz (add the ".tgz" if this is missing).  We are providing this starting data set, rather than the original data set for two reasons.  First, many of these initial processing steps can be rather time consuming (> 1 hr). Second, while necessary, many of these steps are not fundamental to the calibration and imaging process, which is the focus of this tutorial.  We will use test data that was taken in a hybrid configuration when the VLA array was being moved from B-config to A-config.
The entire data set can be downloaded from the [https://data.nrao.edu NRAO archive] by searching for the following Project Code:
<div style="text-align:justify;">
<pre style="background-color: #gray;">
TSUB0001
</pre>
</div>
This returns a long list with test-data that are publicly available. Our observations were performed on 14 Sept 2016:
<div style="text-align:justify;">
<pre style="background-color: #gray;">
TSUB0001.sb32720781.eb32763188.57645.263958564814
</pre>
</div>
Note that this observation was duplicated on 15 September 2016. <!-- This second day of observations can be found under filename:
TSUB0001.sb32720781.eb32773507.57646.24443702547 -->
The 15 GB data set from the download has already had online flagging and Hanning smoothing applied, and the smaller data set split out from the larger data set. If you wish, you can download the SDM from the NRAO archive in order to apply the online flags and run Hanning smoothing on the large data set before splitting out the smaller data set to work on. To apply the on line flag use the task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.flagging.flagcmd.html flagcmd]; to apply the Hanning smoothing use the task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.manipulation.hanningsmooth.html hanningsmooth] on scans 7 to 16 and field 3 and 4; and to split the data, use the task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.manipulation.split.html split]. For the purpose of this tutorial, we only reduce and image the first data set. Then, you will need to split only the spw 17, and only channels 20 to 485 to remove the borders. 
<!-- The second data set can be reduced in an identical way (pending the flagging, given that the RFI conditions most likely changed). During the imaging stage you can then combine both reduced data sets. -->
<!-- To download the data, fill in your email, select either the "SDM-BDF (all files)" or “MS” option, and check the box next to the data sets that you want. The practical difference between selecting the MeasurementSet (MS) and Science Data Model (SDM) is that for MS the first step in the data reduction described below, namely reading in the SDM data to save it as a MeasurementSet, is already done by the archival engine. Note that you can also opt to download the data as a tar-file by tickking the appropriate box.
Optionally, you can request to discard data marked as ‘bad’ by ticking the box “Apply Telescope Flags”, which gets rid of data taken during times of instrument calibration, shadowing, slewing, etc. However, it also applies all flags on the science data that were automatically created during the observations. Therefore, a safer option is to not apply any flags before downloading the data, but first inspect and subsequently apply the flags generated during the observations using [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.flagging.flagcmd.html flagcmd], as part of the data reduction plan.
Click "Get My Data" will forward you to the next page, where you should choose the delivery method (either downloading over the internet, or sending home a hard-drive with the data). If you opt to retrieve the data over the internet, wait until you get an email confirming that the data is ready for download. -->
== Examining the Data ==
=== Loading data into CASA ===
We will have to untar the 15 GB data set that you've downloaded.
<pre style="background-color:lightgrey;">
# in a terminal, outside of CASA:
tar -xzf MG0414_d1_data.ms.tgz
</pre>
Please use CASA 6.5.2 for this tutorial (typing ''casa -ls'' in a linux window shows the available versions and the current version; to explicitly change the version type, e.g., ''casa -r 6.5.2-26-py3.6'')
<source lang='bash'>
casa
</source>
on the command line. This should start a CASA interactive python (iPython) session, and open a separate log window. To guarantee that the below mentioned procedure for data reduction and imaging works, make sure you are using CASA version 6.5.2. While older versions may work as well for the purpose of this tutorial, it is good to read instructions on how to [https://casa.nrao.edu/casa_obtaining.shtml download and install] the latest version of CASA.
<!--
We will begin by importing our data from the binary format (SDM-BDF) into the MeasurementSet format, which is the standard for CASA data. For this, we use <i>importevla</i>:
<source lang='python'>
# In CASA
importevla(asdm='TSUB0001.sb32720781.eb32763188.57645.263958564814', vis='MG0414_d1.ms', flagpol=False, applyflags=False, savecmds=True, outfile='flagfile.txt')
</source>
<!-- In this case, we do not apply the flags that were created as part of the observations, but we will write them out to a <i>flagfile.txt</i> file. We will inspect and apply the flags as follows:
<source lang='python'>
# In CASA
flagcmd(vis='MG0414_d1.ms', action='plot')
flagcmd(vis='MG0414_d1.ms', action='apply')
</source>
As an alternative to <i>flagcmd</i>, given that we saved the flags to <i>flagfile.txt</i>, we can also use
<source lang='python'>
# In CASA
flagdata(vis='MG0414_d1.ms', mode='list', inpfile='flagfile.txt', action='apply')
</source>
to apply the flags.
One of the flag commands is to <i>clip</i> all data points that have a 0-value, as these most likely have not recorded any science data. This is normally done with the settings <i>correlation=`ABS_RR’</i> and <i>correlation=`ABS_LL’</i>. Because the VLA P-band and 4-band systems use linear XX and YY polarisation, rather than the circular RR and LL polarisation of the other VLA bands, you will encounter an error message when running <i>flagcmd</i>. We can get around this issue by manually clipping the 0-data for the XX and LL polarization with <i>flagdata</i>:
<source lang='python'>
# In CASA
flagdata(vis='MG0414_d1.ms', mode='clip', correlation='ABS_XX,ABS_YY', clipzeros=True, action='apply')
</source>
-->
It is a good practice to read the Operator Log that was created by the duty operator during the observations. This log can provide additional information on data that should be flagged manually during the data reduction stage. The Operator logs can be found at [http://www.vla.nrao.edu/cgi-bin/oplogs.cgi Ops Logs form entry], and this is the [http://www.vla.nrao.edu/operators/logs/2016/9/2016-09-14_0620_TSUB0001.pdf direct link for the log of the observation] (in PDF format).
=== Inspecting the observation info ===
We will inspect the content of our data set using CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.information.listobs.html listobs].
<source lang='python'>
# In CASA
listobs(vis='MG0414_d1_data.ms')
</source>
The task plots an overview of the data in the CASA logger. Notice that the task returns a dictionary on the terminal screen. Using the parameter <i>listfile=`listobs.txt’</i>, this information is written out to the file <i>listobs.txt</i>:
<source lang='python'>
# In CASA
listobs(vis='MG0414_d1_data.ms', listfile='listobs.txt')
</source>
The content is included below:
<small style="font-size:80%;"><pre>
  Observer: . Frazer Owen    Project: uid://evla/pdb/1695465 
Observation: EVLA
Data records: 1179738      Total elapsed time = 9390 seconds
  Observed from  14-Sep-2016/06:43:48.0  to  14-Sep-2016/09:20:18.0 (UTC)
  ObservationID = 0        ArrayID = 0
  Date        Timerange (UTC)          Scan  FldId FieldName            nRows    SpwIds  Average Interval(s)    ScanIntent
  14-Sep-2016/06:43:48.0 - 06:48:42.0    7      0 3C48                    37044  [0]  [3] [CALIBRATE_AMPLI#UNSPECIFIED,CALIBRATE_BANDPASS#UNSPECIFIED,CALIBRATE_FLUX#UNSPECIFIED,CALIBRATE_PHASE#UNSPECIFIED]
              06:48:45.0 - 07:08:39.0    8      1 MG0414+0534            150444  [0]  [3] [OBSERVE_TARGET#UNSPECIFIED]
              07:08:42.0 - 07:28:36.0    9      1 MG0414+0534            150444  [0]  [3] [OBSERVE_TARGET#UNSPECIFIED]
              07:28:39.0 - 07:48:33.0    10      1 MG0414+0534            150444  [0]  [3] [OBSERVE_TARGET#UNSPECIFIED]
              07:48:36.0 - 08:08:30.0    11      1 MG0414+0534            150444  [0]  [3] [OBSERVE_TARGET#UNSPECIFIED]
              08:08:33.0 - 08:14:30.0    12      0 3C48                    44982  [0]  [3] [CALIBRATE_AMPLI#UNSPECIFIED,CALIBRATE_BANDPASS#UNSPECIFIED,CALIBRATE_FLUX#UNSPECIFIED,CALIBRATE_PHASE#UNSPECIFIED]
              08:14:33.0 - 08:34:27.0    13      1 MG0414+0534            150444  [0]  [3] [OBSERVE_TARGET#UNSPECIFIED]
              08:34:30.0 - 08:54:21.0    14      1 MG0414+0534            150066  [0]  [3] [OBSERVE_TARGET#UNSPECIFIED]
              08:54:24.0 - 09:14:18.0    15      1 MG0414+0534            150444  [0]  [3] [OBSERVE_TARGET#UNSPECIFIED]
              09:14:21.0 - 09:20:18.0    16      0 3C48                    44982  [0]  [3] [CALIBRATE_AMPLI#UNSPECIFIED,CALIBRATE_BANDPASS#UNSPECIFIED,CALIBRATE_FLUX#UNSPECIFIED,CALIBRATE_PHASE#UNSPECIFIED]
          (nRows = Total number of rows per scan)
Fields: 2
  ID  Code Name                RA              Decl          Epoch  SrcId      nRows
  0    NONE 3C48                01:37:41.299431 +33.09.35.13299 J2000  0        127008
  1    NONE MG0414+0534        04:14:37.800000 +05.34.41.99999 J2000  1        1052730
Spectral Windows:  (1 unique spectral windows and 1 unique polarization setups)
  SpwID  Name          #Chans  Frame  Ch0(MHz)  ChanWid(kHz)  TotBW(kHz) CtrFreq(MHz) BBC Num  Corrs         
  0      EVLA_P#A0C0#17    466  TOPO    386.312        15.625      7281.2    389.9453      12  XX  XY  YX  YY
Sources: 2
  ID  Name                SpwId RestFreq(MHz)  SysVel(km/s)
  0    3C48                0    -              -           
  1    MG0414+0534        0    -              -           
Antennas: 27:
  ID  Name  Station  Diam.    Long.        Lat.                Offset from array center (m)                ITRF Geocentric coordinates (m)       
                                                                    East        North    Elevation              x              y              z
  0    ea01  W24      25.0 m  -107.38.49.0  +33.53.04.0      -2673.3457    -1784.6009      10.4844 -1604008.744400 -5042135.825100  3553403.710800
  1    ea02  W56      25.0 m  -107.44.26.7  +33.49.54.6    -11333.1991    -7637.6832      15.3636 -1613255.391400 -5042613.097800  3548545.906000
  2    ea03  W40      25.0 m  -107.41.13.5  +33.51.43.1      -6377.9680    -4286.7773        8.2312 -1607962.451800 -5042338.218100  3551324.962400
  3    ea04  E04      25.0 m  -107.37.00.8  +33.53.59.7        102.7973      -63.7800      -2.6177 -1601068.806000 -5042051.932700  3554824.838800
  4    ea05  E36      25.0 m  -107.33.20.2  +33.52.34.3      5761.3733    -2704.6731      33.0133 -1596127.730800 -5045193.742100  3552652.419700
  5    ea06  N24      25.0 m  -107.37.16.1  +33.55.37.7      -290.3584    2961.8653      -12.2425 -1600930.069900 -5040316.395500  3557330.390000
  6    ea07  N16      25.0 m  -107.37.10.9  +33.54.48.0      -155.8512    1426.6377      -9.3934 -1601061.954400 -5041175.875300  3556058.026700
  7    ea08  W64      25.0 m  -107.46.20.1  +33.48.50.9    -14240.7524    -9606.2900      17.0885 -1616361.575500 -5042770.516600  3546911.419900
  8    ea09  N40      25.0 m  -107.37.29.5  +33.57.44.4      -633.6056    6878.5897      -20.7984 -1600592.749000 -5038121.341300  3560574.826900
  9    ea10  E20      25.0 m  -107.35.43.6  +33.53.29.9      2082.1759    -987.0087        9.4361 -1599340.800100 -5043150.963000  3554065.231500
  10  ea12  N32      25.0 m  -107.37.22.0  +33.56.33.6      -441.7248    4689.9674      -16.9328 -1600781.044500 -5039347.439100  3558761.527100
  11  ea13  E28      25.0 m  -107.34.39.3  +33.53.04.9      3732.7776    -1757.3308      21.4271 -1597899.895900 -5044068.684700  3553432.450200
  12  ea14  E08      25.0 m  -107.36.48.9  +33.53.55.1        407.8283    -206.0315      -3.2236 -1600801.931400 -5042219.382600  3554706.429400
  13  ea15  N64      25.0 m  -107.37.58.7  +34.02.20.5      -1382.3871    15410.1326      -40.6450 -1599855.687000 -5033332.368600  3567636.606800
  14  ea16  W72      25.0 m  -107.48.24.0  +33.47.41.2    -17419.4641  -11760.2694      14.9442 -1619757.299900 -5042937.656400  3545120.392300
  15  ea17  N56      25.0 m  -107.37.47.9  +34.00.38.4      -1105.2042    12254.2800      -34.2710 -1600128.378100 -5035104.139200  3565024.633900
  16  ea18  W48      25.0 m  -107.42.44.3  +33.50.52.1      -8707.9181    -5861.7682      15.5302 -1610451.901900 -5042471.123800  3550021.073100
  17  ea19  W16      25.0 m  -107.37.57.4  +33.53.33.0      -1348.7121    -890.6209        1.2959 -1602592.853500 -5042054.992400  3554140.702800
  18  ea20  E16      25.0 m  -107.36.09.8  +33.53.40.0      1410.0403    -673.4656      -0.7790 -1599926.104100 -5042772.977200  3554319.801100
  19  ea21  E24      25.0 m  -107.35.13.4  +33.53.18.1      2858.1906    -1349.1352      13.7186 -1598663.082000 -5043581.391200  3553767.014100
  20  ea22  W32      25.0 m  -107.39.54.8  +33.52.27.2      -4359.4399    -2923.1314      11.7614 -1605808.634100 -5042230.084000  3552459.197800
  21  ea23  E12      25.0 m  -107.36.31.7  +33.53.48.5        848.6994    -411.6226      -2.7575 -1600416.518000 -5042462.430500  3554536.041700
  22  ea24  W08      25.0 m  -107.37.21.6  +33.53.53.0      -432.1080    -272.1514      -1.5061 -1601614.083200 -5042001.656900  3554652.505900
  23  ea25  N08      25.0 m  -107.37.07.5  +33.54.15.8        -68.9088      433.1829      -5.0674 -1601147.942500 -5041733.833600  3555235.947000
  24  ea26  E32      25.0 m  -107.34.01.5  +33.52.50.3      4701.6503    -2209.7119      25.1976 -1597053.124400 -5044604.675000  3553058.992700
  25  ea27  N48      25.0 m  -107.37.38.1  +33.59.06.2      -855.2719    9405.9407      -25.9485 -1600374.881000 -5036704.201700  3562667.858900
  26  ea28  N72      25.0 m  -107.38.10.5  +34.04.12.2      -1685.6797    18861.8306      -43.5015 -1599557.928700 -5031396.353400  3570494.736800
</pre></small>
<!-- === Hanning smoothing the data ===
When plotting the data in plotms (plotting amplitude against frequency), it is immediately clear that our observing band contains lots of Radio Frequency Interference (RFI), which is unavoidable at these low frequencies. The situation actually looks worse than it is, because ‘Gibbs ringing’ causes the blending of strong RFI signal into adjacent channels (Fig.1). To reduce the effect of Gibbs ringing, we first Hanning smooth the data (please note that this can take a while to run on the entire 142 Gb data set):
<source lang='python'>
# In CASA
hanningsmooth(vis='MG0414_d1_data.ms', datacolumn='data', outputvis='MG0414_d1_hanning.ms')
</source>
[[File:Pband_hanning.png|400px|thumb|right|Fig.1 Reducing the effects of Gibbs ringing by Hanning smoothing the data. Shown is only a small fraction of the band, both before and after Hanning smooth. The amplitude scale is the same in both plots.]]
=== Splitting the data ===
Upon further inspection in “plotms” (using ea01, field 0, corr xx,yy, and coloraxis baseline), it is clear that there is a lot of RFI across most of the 64 MHz (or 43,000 km/s) band that we used for the observations. Luckily, the region around the expected HI absorption line at 390.6 MHz is relatively clean of RFI. Because our target is strong enough in the continuum to perform a good self-calibration, we select this relatively clean part of the band centered around the expected HI line. Note that there were two IFs with different spectral resolution included in the observations. We need only one of the two IFs, and choose the one with the highest spectral resolution. Moreover, we only need the sources 3C48 (which we will use for bandpass, gain and flux calibration) and our target MG0414+0534, hence we select field=’3,4’. This approach has the additional advantage that we reduce our data volume a lot, bringing it down to about 15 Gb.
<source lang='python'>
# In CASA
split(vis='MG0414_d1_hanning.ms', outputvis='MG0414_d1_data.ms', datacolumn='data', field='3,4', spw='17:20~485', keepflags=False)
</source>
<font color=red>'''This 15 Gb data set can be downloaded by clicking [https://casa.nrao.edu/Data/EVLA/MG0414/MG0414_d1_data.ms.tgz here].'''</font> The filename should be MG0414_d1_data.ms.tgz (add the ".tgz" if this is missing), and the file can be unpacked by typing <i>tar xzf MG0414_d1_data.ms.tgz</i>
-->
{| style="border-spacing: 2px; border: 1px solid darkgray; float:right;"
|- style="background: white;"
| style="border: 1px solid darkgray;" | [[File:MG0414_plotants_5.4.0.png|350px|thumb]] ||  style="border: 1px solid darkgray" | [[File:Ant_logplot_6.5.2.png  |350px|thumb]]
|- style="vertical-align: top; background: white; border: 1px solid darkgray;"
| Figure 1a. Plotants plot for <br />MG0414. || Figure 1b. Plotants plot <br />for MG0414 in a log scale.
|}
Next we look at a graphical plot of the antenna locations and save a hard copy
in case you want it later using CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.visualization.plotants.html plotant] (see Figures 1a and 1b). This will be useful for selecting a reference antenna. Note that the VLA was in a period of re-configuration when these MG0414 data were taken, so the antenna configuration is not a standard VLA configuration. Typically a good choice is an antenna close to the center of the array (see Figure 1b). Unless it shows problems after inspection of the data, we provisionally choose ea04. (Note: clicking on any image will open up a larger version.)
<source lang="python">
# In CASA
plotants(vis='MG0414_d1_data.ms',figfile='ant_locations.png')
# In CASA
# If you would like to see another view of the antennas you may plot them on a log plot
plotants(vis='MG0414_d1_data.ms',logpos=True,figfile='ant_logplot.png')
</source>
=== Flagging ===
<!-- NOTE: when revising for CASA release 5.3.0, EM wants to revisit how flagging is approached in the guide. Some of the flagging tasks, such as running rflag and flagging out antennas ea09, 14, and 17 should be introduced earlier - perhaps in this section - so that major flagging is done once and then the user can proceed to the self-calibration and imaging section.  Tony -->
[[File:Pband_flagchan.png|760px|thumb|right|Figure 2. Flagging of calibrator 3C48. Shown are the XX and YY polarization on antenna ea01, both before (left) and after (right) flagging the bad channels. The colors represent the different baselines with ea01.]]
We will now perform a more in-depth flagging of the bad data. Our strategy is to first focus on the calibrator source (in this case only 3C48). If the calibrator data are flagged well, we should be able to perform an accurate calibration of the target data. Later in this document we will then flag the calibrated target data using an automated flagger.
When plotting amplitude vs frequency in CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casaplotms.plotms.html#casaplotms.plotms plotms], we see that there are still two ranges of channels filled with RFI throughout the observing run (see Figure 2).
<source lang='python'>
# In CASA
plotms(vis='MG0414_d1_data.ms',antenna='ea01',xaxis='chan',yaxis='amp',correlation='xx,yy',coloraxis='baseline', scan='7',averagedata=True, avgtime='1e9')
</source>
We will flag this data with the CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.flagging.flagdata.html flagdata]. Because we are flagging channels on all scans of the bandpass calibrator 3C48, these channels cannot be calibrated accurately, so we flag these channels also in the data of the target.
<source lang='python'>
# In CASA
flagdata(vis='MG0414_d1_data.ms', mode='manual', spw='0:148~155', action='apply')
flagdata(vis='MG0414_d1_data.ms', mode='manual', spw='0:51~59', action='apply')
</source>
The data on our bandpass+gain+flux calibrator 3C48 now look clean, with no dead or misbehaving antennas. We can repeat the same [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casaplotms.plotms.html#casaplotms.plotms plotms] call as above:
<source lang='python'>
# In CASA
plotms(vis='MG0414_d1_data.ms',antenna='ea01',xaxis='chan',yaxis='amp',correlation='xx,yy',coloraxis='baseline', scan='7',averagedata=True, avgtime='1e9')
</source>
We now start with the calibration.
== Calibration ==
Many of the steps below are based on the CASA tutorial on [https://casaguides.nrao.edu/index.php/Karl_G._Jansky_VLA_Tutorials#3C129_P-band_continuum_imaging_data_reduction_tutorial reducing VLA P-band continuum data].
=== <i>A priori</i> Antenna Position Corrections ===
We start the calibration by obtaining the latest set of antenna-position corrections compared to the stored values that were derived at the start of the observing period. For this you need to run CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.gencal.html gencal] with parameter <i>caltype='antpos'</i>. (Be sure to have internet connection in order to grab the antenna position corrections.)
<source lang='python'>
# In CASA
gencal(vis='MG0414_d1_data.ms', caltype='antpos', caltable='antpos.cal')
</source>
The output is a list of antenna position corrections that are written to the calibration table <i>antpos.cal</i>:
<small style="font-size:85%;"><pre>
2017-07-25 19:54:51 INFO gencal Determine antenna position offests from the baseline correction database
2017-07-25 19:54:52 INFO gencal offsets for antenna ea01 :  0.00200  0.00000  0.00200
2017-07-25 19:54:52 INFO gencal offsets for antenna ea02 :  0.00210  0.01800  0.00430
2017-07-25 19:54:52 INFO gencal offsets for antenna ea03 : -0.00700  0.01600  -0.02000
2017-07-25 19:54:52 INFO gencal offsets for antenna ea06 :  0.00200  -0.00200  0.00500
2017-07-25 19:54:52 INFO gencal offsets for antenna ea08 : -0.01290  0.01280  0.02920
2017-07-25 19:54:52 INFO gencal offsets for antenna ea09 : -0.01200  -0.01300  0.02400
2017-07-25 19:54:52 INFO gencal offsets for antenna ea15 : -0.00500  -0.01000  0.01100
2017-07-25 19:54:52 INFO gencal offsets for antenna ea16 :  0.01100  -0.01500  -0.01200
2017-07-25 19:54:52 INFO gencal offsets for antenna ea17 : -0.01300  0.00000  0.05000
2017-07-25 19:54:52 INFO gencal offsets for antenna ea18 : -0.00600  0.02200  -0.02000
2017-07-25 19:54:52 INFO gencal offsets for antenna ea22 :  0.00600  0.00000  0.00000
2017-07-25 19:54:52 INFO gencal offsets for antenna ea25 : -0.00140  0.00140  -0.00140
2017-07-25 19:54:52 INFO gencal offsets for antenna ea27 : -0.00400  -0.00500  0.03100
2017-07-25 19:54:52 INFO gencal offsets for antenna ea28 : -0.02100  -0.02900  0.01900
2017-07-25 19:54:52 INFO calibrater Beginning specifycal-----------------------
2017-07-25 19:54:52 INFO  Creating KAntPos Jones table from specified parameters.
2017-07-25 19:54:52 WARN  NB: This EVLA dataset appears to fall within the period
2017-07-25 19:54:52 WARN +   of semester 16B during which the online tropospheric
2017-07-25 19:54:52 WARN +   delay model was mis-applied.
2017-07-25 19:54:52 WARN  A correction for the online tropospheric delay model error WILL BE APPLIED!
2017-07-25 19:54:52 WARN  Marking antpos caltable to turn ON the trop delay correction.
</pre></small>
Note that these observations were taken during a period in which the atmospheric delay terms were calculated incorrectly. In CASA versions 4.7.1 and up a [https://science.nrao.edu/facilities/vla/data-processing/vla-atmospheric-delay-problem correction] for this term is taken into account automatically when running CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.gencal.html gencal] with parameter <i>caltype='antpos'</i>.
=== Ionospheric calibration ===
At frequencies below 5 GHz, ionospheric effects may need to be corrected. These corrections become important for frequencies < 1 GHz. CASA’s strategy is to obtain information on the total electron content (TEC) for the date of observations, which is based on data from the [https://science.nrao.edu/facilities/vlba/publications/memos/sci/gps_ion/node3 global navigation satellite system (GNSS)]. A series of CASA images in the form of a 24 hour movie of the TEC, as function of longitude and latitude, is then generated and stored as ''filename.IGS_TEC.im'', with a corresponding TEC error movie named ''filename.IGS_RMS_TEC.im'' (these movies can be viewed in CASA’s [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.visualization.viewer.html viewer] (please note that this task is deprecated, and in the future CARTA should be used for the visualizations). In order to download the data, you need to be connected to the internet:
<source lang='python'>
# In CASA
from casatasks.private import tec_maps
 
tec_image, tec_rms_image, plotname = tec_maps.create(vis='MG0414_d1_data.ms', doplot=True)
</source>
If, for any reason, the file cannot be retrieved from the server automatically, you can also download this specific TEC file for the tutorial [http://casa.nrao.edu/Data/EVLA/MG0414/igsg2580.16i here].
The information regarding the TEC is sparse and either active or direction-dependent ionospheric conditions may not be corrected very well. Also, the online TEC information improves with time, hence the quality of these data is best about two weeks after the observations.
[[File:Pband_TECIM.png|400px|thumb|right|Figure 3. Total Electron Content (TEC) as functions of time for the day and location of the observations. The time of the observing run is shown in red. We can also see this plot on the log scale, depending on the last plotants inputs.]]
An ionosphere correction table is subsequently generated using CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.gencal.html gencal], in which the projected line-of-sight TEC (which depends on the zenith angle) is sampled for all times in the observation and stored in a standard CASA caltable. Figure 3 shows a plot that is generated for this caltable and can be opened using [https://gitlab.com/DavidGriffith/xv xv].
<source lang='python'>
# In CASA
gencal(vis='MG0414_d1_data.ms', caltype='tecim', caltable='tecim.cal',infile=tec_image)
</source>
See the CASA Cookbook 4.7.2 for a [https://casa.nrao.edu/casa_cookbook.pdf#subsection.4.3.9 detailed description] on ionosphere corrections in CASA (thanks are due to Jason Kooi, University of Iowa, for his contributions).
=== Calibration of requantizer gains ===
Next we will calibrate the requantizer gain levels, which are the visibility amplitudes that were set as the input of the WIDAR correlator. Requantizer scans are added to a P-band observation to optimize the digital power in each spectral window which, in turn, maximizes the signal-to-noise of each window. Because there is a significant variation in power across the 240 MHz of P-band, and because some spectral windows may experience a high power due to strong RFI, setting the requantizer levels during the observations may improve the quality of the data, including the shape of the bandpass across multiple spectral windows. As part of the test observations of MG0414+0534, the very strong source Cygnus A was observed at the start of the run. Re-setting the requantizer levels both before observing Cygnus A and before targeting the other (much weaker) sources was essential to optimize the digital power and avoid correlation errors. Although we do not use Cygnus A as part of this tutorial, and we use only a single spectral window, performing calibration of the requantizer gains is still good practice for P-band spectral line reduction using CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.gencal.html gencal].
:
<source lang='python'>
# In CASA
gencal(vis='MG0414_d1_data.ms', caltype='rq', caltable='rq.cal')
</source>
=== Initial absolute flux density scale calibration ===
The next step is to set the absolute flux levels of our flux calibrator 3C48 (field='0' from our [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.information.listobs.html listobs] output), for which a well-known model exists using CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.imaging.setjy.html setjy]:
<source lang='python'>
# In CASA
setjy(vis='MG0414_d1_data.ms', field='0', scalebychan=True, standard='Perley-Butler 2017', listmodels=False, usescratch=True)
</source>
It is crucial for spectral line work to set the parameter ''scalebychan=True'' to ensure that the absolute flux level is calculated per channel and correctly interpolated across the observing band. If ''scalebychan=False'', then only a single value per spectral window is calculated, resulting in a step function in flux between spectral windows.
=== Delay calibration ===
The delay of each antenna, for each polarization and each spectral window, is now determined. Doing this on a short scan of the primary calibrator is generally sufficient and will correct for both internal (e.g., electronics, cables) and external (e.g., ionosphere) effects. We select ea04 as reference antenna as it produced good data, has baselines to all other used antennas in the array, and is located in the center of the array. We also need to apply, "on-the-fly," the calibration tables for the antenna positions, the total electron content, and the requantizer levels using the CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.gaincal.html gaincal].
<source lang='python'>
# In CASA
gaincal(vis='MG0414_d1_data.ms', caltable='delays.cal', field='0', selectdata=True, timerange='06:43:48~06:48:42', solint='inf', refant='ea04', gaintype='K', gaintable=['antpos.cal','tecim.cal','rq.cal'])
</source>
[[File:Pband_delaycal_5.4.0.png|500px|thumb|right|Figure 4. Delays for each antenna and each polarization. Plot made with CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casaplotms.plotms.html#casaplotms.plotms plotms].]]
The calibration solutions can be plotted using CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casaplotms.plotms.html#casaplotms.plotms plotms] (see Figure 4):
<source lang='python'>
# In CASA
plotms(vis='delays.cal',xaxis='antenna1',yaxis='delay',field='0',coloraxis='corr')
</source>
Note that any delays deviating by more than 30 nsec should be treated with caution and, if needed, flagged.
=== Bandpass calibration ===
Next step is to calibrate the bandpass, which corrects for the frequency dependent gain variations. Before determining the frequency dependent gains, we first calibrate the phases in time, deriving a solution for each integration, to make sure that we can derive accurate bandpass solutions when integrating over a full scan on the bandpass calibrator 3C48. This is done using CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.gaincal.html gaincal]:
[[File:396px-Pband BPcal.png|1500px|thumb|right|Figure 5. Effects of bandpass calibration on the frequency dependent amplitudes (top) and phases (bottom) across the observing band. (For corrections applied using the CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.applycal.html applycal] see Figure 8a below.) The various colors show the different baselines with reference antenna ea04. The plots were made in CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casaplotms.plotms.html#casaplotms.plotms plotms] and in the ''Axes'' tab choosing the option ''Data Column: data''.]]
<source lang='python'>
# In CASA
gaincal(vis='MG0414_d1_data.ms', caltable='bpphase.gcal', field='0', spw='0:250~300', refant='ea04', calmode='p', solint='int', minsnr=3.0, gaintype='G', gaintable=['antpos.cal','tecim.cal','rq.cal','delays.cal'])
</source>
We use these phase solutions to perform the actual bandpass calibration. In this case, we have scans of our bandpass calibrator 3C48 taken at three different times during the observation. By interpolating the bandpass solutions in time, we try to minimize any possible time-varying bandpass effects and improve our spectral dynamic range. Bandpass calibration is performed in CASA by the task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.bandpass.html bandpass]:
<source lang='python'>
# In CASA
bandpass(vis='MG0414_d1_data.ms', caltable='bandpass.cal', field='0', spw='0', solint='inf', combine='', refant='ea04', solnorm=False, bandtype='B', gaintable=['antpos.cal','tecim.cal','rq.cal','delays.cal','bpphase.gcal'])
</source>
To interpolate between bandpass scans, it is essential to specify parameter <i>combine=' '</i>. The reason being that the default of <i>combine='scan'</i>, when coupled with the parameter <i>solint='inf'</i>, would cross the scan boundaries to form one single solution for the bandpass. Only when combining <i>solint='inf'</i> with <i>combine=' '</i> is a bandpass solution obtained for each individual scan on 3C48. The bandpass plots can be viewed using CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casaplotms.plotms.html#casaplotms.plotms plotms] (see Figure 5).
<source lang='python'>
# In CASA
plotms(vis='MG0414_d1_data.ms',field='0',antenna='ea04', xaxis='frequency',yaxis='amp',scan='7',correlation='xx,yy', coloraxis='baseline')
plotms(vis='MG0414_d1_data.ms',field='0',antenna='ea04', xaxis='frequency',yaxis='phase',scan='7',correlation='xx,yy', coloraxis='baseline')
</source>
=== Gain calibration ===
After the bandpass calibration, we perform a phase calibration of our data. Because we apply the bandpass solutions "on-the-fly," we can use the full bandwidth to determine the phase solutions. We will create two different calibration tables to correct for the time-varying phases. First we calibrate the phases on each interval with the CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.gaincal.html gaincal] parameter of ''solint'' set to ''int'':
<source lang='python'>
# In CASA
gaincal(vis='MG0414_d1_data.ms', caltable='intphase.gcal', field='0', spw='0', refant='ea04', calmode='p', solint='int', minsnr=3.0, gaintable=['antpos.cal','tecim.cal','rq.cal','delays.cal','bandpass.cal'])
</source>
[[File:NewFig6_plot.png|500px|thumb|right|Figure 6. Effects of phase calibration. (For corrections applied using the CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.applycal.html applycal] see Figure 8b below.) The various colors show the different baselines with reference antenna ea04.]]
The sole purpose of generating phase solutions on each 10 second interval is to be able to accurately correct for the time-varying amplitudes as we will see below. This method, however, is not ideal for interpolating the phase solutions in time. For that we do a second phase calibration, this time averaging the signal across the phase calibrator scans with the CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.gaincal.html gaincal] parameter of <i>solint='inf'</i>:
<source lang='python'>
# In CASA
gaincal(vis='MG0414_d1_data.ms', caltable='scanphase.gcal', field='0', spw='0', refant='ea04', calmode='p', solint='inf', combine='', minsnr=3.0, gaintable=['antpos.cal','tecim.cal','rq.cal','delays.cal','bandpass.cal'])
</source>
Note that an alternative approach to deriving phase solutions per scan is to use the CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.smoothcal.html smoothcal] to average the phase corrections obtained every 10s with <i>solin='int'</i> over the full duration of a single scan. Generate a phase plot using CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casaplotms.plotms.html#casaplotms.plotms plotms] with the parameters below (see Figure 6):
<source lang='python'>
# In CASA
plotms(vis='MG0414_d1_data.ms',field='0',antenna='ea04', correlation='xx,yy',avgchannel='1e8', xaxis='time',yaxis='phase',coloraxis='baseline')
</source>
We have selected an arbitrarily high number for the average channel to ensure it will include all of them. It will led to a warning message on the terminal screen.
=== Scaling the Amplitude Gains ===
[[File:NewFig7_plot.png|500px|thumb|right|Figure 7. Effects of amplitude and flux density scale calibration. The green points are the YY products on an offending antenna ea17, which we will flag in our target data. (For corrections applied using the CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.applycal.html applycal] see Figure 9 below.)]]
We now perform a calibration of the amplitude variations in time. For this, we will apply the phase solution obtained in each 10 second interval via CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.gaincal.html gaincal]:
<source lang='python'>
# In CASA
gaincal(vis='MG0414_d1_data.ms', caltable='amp.gcal', field='0', spw='0', selectdata=False, solint='inf', combine='', refant='ea04', calmode='ap', minsnr=3.0, gaintable=['antpos.cal','tecim.cal','rq.cal','delays.cal','bandpass.cal','intphase.gcal'])
</source>
Normally this procedure will only calibrate the relative amplitudes, and the CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.fluxscale.html fluxscale] is required to place the data on an absolute flux density scale. However, because our flux-calibrator is also our bandpass and gain calibrator, the absolute flux density scale will already be ok, so we can skip this step. Generate an amplitude gain plot using CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casaplotms.plotms.html#casaplotms.plotms plotms] (see Figure 7) with the parameters below:
<source lang='python'>
# In CASA
plotms(vis='MG0414_d1_data.ms',field='0',antenna='ea04', correlation='xx,yy',avgchannel='1e8', xaxis='time',yaxis='amp',coloraxis='baseline')
</source>
=== Applying the calibration ===
We have now finished the calibration of the data. We will apply the calibration tables that were derived from 3C48 to both the calibrator 3C48 itself and our target MG0414+0354. The CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.applycal.html applycal] creates a '''corrected''' data column where the calibrated data are stored.
<source lang='python'>
# In CASA
applycal(vis='MG0414_d1_data.ms', field='0', gaintable=['antpos.cal','tecim.cal','rq.cal','delays.cal','bandpass.cal','scanphase.gcal','amp.gcal'], gainfield=['','','','0','0','0','0'], parang=True, calwt=False, applymode='calflagstrict', flagbackup=True)
</source>
<source lang='python'>
# In CASA
applycal(vis='MG0414_d1_data.ms', field='1', gaintable=['antpos.cal','tecim.cal','rq.cal','delays.cal','bandpass.cal','scanphase.gcal','amp.gcal'], gainfield=['','','','0','0','0','0'], calwt=False, applymode='calflagstrict', flagbackup=True)
</source>
Sometimes after calibrating the data it is easier to see any offending antennas or baselines. Replotting with <i>ydatacolumn='corrected'</i> will produce the corrected calibration plots (Figures 8a-d).
<source lang='python'>
# In CASA for corrected amplitude and phase vs frequency, Figures 8a and 8b
plotms(vis='MG0414_d1_data.ms',field='0',antenna='ea04', xaxis='frequency',yaxis='amp',scan='7',ydatacolumn='corrected',correlation='xx,yy', coloraxis='baseline')
plotms(vis='MG0414_d1_data.ms',field='0',antenna='ea04', xaxis='frequency',yaxis='phase',scan='7',ydatacolumn='corrected',correlation='xx,yy', coloraxis='baseline')
</source>
<source lang='python'>
# In CASA for corrected amplitude and phase vs time, Figures 8c and 8d
plotms(vis='MG0414_d1_data.ms',field='0',antenna='ea04', correlation='xx,yy', xaxis='time',yaxis='amp',ydatacolumn='corrected',coloraxis='baseline')
plotms(vis='MG0414_d1_data.ms',field='0',antenna='ea04', correlation='xx,yy',avgchannel='1e8', xaxis='time',yaxis='phase',ydatacolumn='corrected',coloraxis='baseline')
</source>
The calibrated data of the primary calibrator 3C48 show that the YY amplitudes of ea17 are off during the second part of the run (see the green points with low values in figure 7). Because calibration solutions are antenna based, to get the best calibration, flag any poor data and re-run the calibration process. A currently known issue is that CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.gaincal.html gaincal] will not find good solutions if any correlation, including cross-correlation, in the data is completely flagged. In this case, there is no difference between flagging only the YY polarization of ea17 or the entire antenna.
Antenna ea17 will be flagged using CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.flagging.flagdata.html flagdata]:
<source lang='python'>
# In CASA
flagdata(vis='MG0414_d1_data.ms', mode='manual', antenna='ea17', action='apply')
</source>
{| style="border-spacing: 2px; border: 1px solid darkgray; float:right;"
|- style="background: white; vertical-align: top;"
| style="border: 1px solid darkgray;" | [[File:Pband_BPcal_corr_amp_scan7_6.5.2.png|400px|thumb]] || style="border: 1px solid darkgray;" | [[File:Pband_BPcal_corr_phase_scan7_6.5.2.png|400px|thumb]] || style="border: 1px solid darkgray;" | [[File:Pband_BPcal_corr_amp_6.5.2.png|400px|thumb]] || style="border: 1px solid darkgray;" | [[File:Pband_BPcal_corr_phase_6.5.1.png|400px|thumb]]
|- style="vertical-align: top; background: white; border: 1px solid darkgray;"
| Figure 8a. Effects of bandpass calibration on the frequency dependent amplitudes across the observing band. The plots were made in CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casaplotms.plotms.html#casaplotms.plotms plotms] and in the ''Axes'' tab choosing option '' Y Data Column: corrected''. || Figure 8b. Effects of bandpass calibration on the frequency dependent phases across the observing band. The plots were made in CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casaplotms.plotms.html#casaplotms.plotms plotms] and in the ''Axes'' tab choosing option '' Y Data Column: corrected''. || Figure 8c. Effects of phase calibration on the time dependent amplitudes across the observing band. The various colors show the different baselines with reference antenna ea04. The plot was made in CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casaplotms.plotms.html#casaplotms.plotms plotms] and in the ''Axes'' tab choosing option ''Y Data Column: corrected''. || Figure 8d. Effects of phase calibration on the time dependent phases across the observing band. The various colors show the different baselines with reference antenna ea04. The plot was made in CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casaplotms.plotms.html#casaplotms.plotms plotms] and in the ''Axes'' tab choosing option ''Y Data Column: corrected''.
|}
<font size="+1">'''<br /><br /> '''</font>
<font size="+1">'''Then run the calibration process again starting with the <i>Delay calibration</i>:'''</font>
<source lang='python'>
# In CASA
gaincal(vis='MG0414_d1_data.ms', caltable='delays.cal', field='0', selectdata=True, timerange='06:43:48~06:48:42', solint='inf', refant='ea04', gaintype='K', gaintable=['antpos.cal','tecim.cal','rq.cal'])
gaincal(vis='MG0414_d1_data.ms', caltable='bpphase.gcal', field='0', spw='0:250~300', refant='ea04', calmode='p', solint='int', minsnr=3.0, gaintype='G', gaintable=['antpos.cal','tecim.cal','rq.cal','delays.cal'])
bandpass(vis='MG0414_d1_data.ms', caltable='bandpass.cal', field='0', spw='0', solint='inf', combine='', refant='ea04', solnorm=False, bandtype='B', gaintable=['antpos.cal','tecim.cal','rq.cal','delays.cal','bpphase.gcal'])
gaincal(vis='MG0414_d1_data.ms', caltable='intphase.gcal', field='0', spw='0', refant='ea04', calmode='p', solint='int', minsnr=3.0, gaintable=['antpos.cal','tecim.cal','rq.cal','delays.cal','bandpass.cal'])
gaincal(vis='MG0414_d1_data.ms', caltable='scanphase.gcal', field='0', spw='0', refant='ea04', calmode='p', solint='inf', combine='', minsnr=3.0, gaintable=['antpos.cal','tecim.cal','rq.cal','delays.cal','bandpass.cal'])
gaincal(vis='MG0414_d1_data.ms', caltable='amp.gcal', field='0', spw='0', selectdata=False, solint='inf', combine='', refant='ea04', calmode='ap', minsnr=3.0, gaintable=['antpos.cal','tecim.cal','rq.cal','delays.cal','bandpass.cal','intphase.gcal'])
</source>
[[File:Pband_ampcal_corrected_CASA5.5.0.png|thumb|500px|thumb|right|Figure 9. Effects of amplitude and flux density scale calibration. The green points (the YY products of antenna ea17 clearly seen in Figure 7) have been flagged. The plot was made in CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casaplotms.plotms.html#casaplotms.plotms plotms] and in the ''Axes'' tab choosing option ''Y Data Column: corrected'']]
The new calibration tables have overwritten the previous tables. Once you've re-run the calibration, you need to apply it.
<source lang='python'>
# In CASA
applycal(vis='MG0414_d1_data.ms', field='0', gaintable=['antpos.cal','tecim.cal','rq.cal','delays.cal','bandpass.cal','scanphase.gcal','amp.gcal'], gainfield=['','','','0','0','0','0'], parang=True, calwt=False, applymode='calflagstrict', flagbackup=True)
applycal(vis='MG0414_d1_data.ms', field='1', gaintable=['antpos.cal','tecim.cal','rq.cal','delays.cal','bandpass.cal','scanphase.gcal','amp.gcal'], gainfield=['','','','0','0','0','0'], calwt=False, applymode='calflagstrict', flagbackup=True)
</source>
And then, plot the corrected amplitudes vs time with ea17 flagged (see Figure 9):
<source lang='python'>
# In CASA for corrected amp vs time with ea17 flagged, Figure 9
plotms(vis='MG0414_d1_data.ms',field='0',antenna='ea04', correlation='xx,yy',avgchannel='1e8', xaxis='time',yaxis='amp',ydatacolumn='corrected',coloraxis='baseline')
</source>
We now split the calibrated data of our target into a new measurement set, including only the XX and YY polarization products. Using CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.manipulation.split.html split] the calibrated target data are placed into the '''data''' column of the new measurement set:
<source lang='python'>
# In CASA
split(vis='MG0414_d1_data.ms', outputvis='MG0414_d1_calibrated.ms', datacolumn='corrected', field='1', correlation='xx,yy', keepflags=False)
</source>
== Imaging and Self-calibration ==
We now have a calibrated data set that has all calibration applied except for possible self-calibration. Self-calibration relies on the target itself to better calibrate antenna-based gains (phases and amplitudes) as a function of time. It can be applied to targets that are strong in the continuum or have a strong enough spectral line to derive good phase and, if desired, also amplitude solutions on short (order of minutes) timescales. The procedure relies on an initial model of the target, which then can be used iteratively to improve the gains and hence the quality of the image itself. Here we show the various steps in self-calibrating our data based on the strong continuum emission of MG0414+0354. Please refer to a [https://casaguides.nrao.edu/index.php?title=First_Look_at_Self_Calibration_CASA_6 dedicated CASA guide] on self-calibration for further details.
If strong RFI is present in the target, it may be necessary to flag your data before proceeding with self-calibration. This is not the case with MG0414+0354.
We create an initial model of our target field using CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.imaging.tclean.html tclean]. Notice that we will exclude from the continuum image the channels that we have flagged and the channels where we expected to detect the line:
<source lang='python'>
# In CASA
tclean(vis='MG0414_d1_calibrated.ms', datacolumn='data', imagename='MG0414_d1_cont_R03', spw='0:4~50;60~147;156~250;350~460', specmode='mfs', niter=100, threshold='1.5Jy', imsize=[256,256], cell='1arcsec', weighting='briggs', robust=0.2, savemodel='modelcolumn')
</source>
The information in the MODEL column consist of visibilities that present a model of the target field. CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.imaging.tclean.html tclean] produces a ''.model'' image of the model that you can visualize with [https://cartavis.org/ CARTA] software. It is essential that only components that you know for sure represent the continuum in the target field are taken into account in self-calibration. If this is not the case, you probably cleaned too deep and included noise features or artifacts (e.g., from the beam-pattern) in the model. Self-calibration would permanently lock these as features in your data, so this should be avoided. On the other hand, it is equally crucial to ensure that the model is not empty, as this will result in your target being shifted to the phase center.
You can see the tclean results using the CARTA tool, as we show in figure 10. Please refer to the [https://casadocs.readthedocs.io/en/stable/notebooks/carta.html CASA documentation] on how to load the image into CARTA.
[[File:CARTA-MG0414_d1_cont_R03-6.5.2.png|800px|thumb|right|Figure 10. Image obtained after the first run of tclean task (MG0414_d1_cont_R03.image), visualized with CARTA.]]
Next, we derive phase corrections to this model. Because our target is strong enough to do self-calibration on it, we should get enough S/N per baseline on a 60 sec interval, using CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.gaincal.html gaincal]:
<source lang='python'>
# In CASA
gaincal(vis='MG0414_d1_calibrated.ms', caltable='sc1.gcal', field='0', spw='0', refant='ea04', calmode='p', solint='60s', combine='', minsnr=3.0)
</source>
This creates a new phase-calibration table that we call ''sc1.gcal''. We now apply this calibration table to the calibrated data in order to improve the phase-only solutions using CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.applycal.html applycal]:
<source lang='python'>
# In CASA
applycal(vis='MG0414_d1_calibrated.ms', field='0', gaintable=['sc1.gcal'], interp='linear')
</source>
Because self-calibration is an iterative process, we want to repeat this step again in order to further improve the gain calibration. For that, we first split the data calibrated with the self-calibration solutions applied to the DATA column to a new measurement set, again using CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.manipulation.split.html split]:
<source lang='python'>
# In CASA
split(vis='MG0414_d1_calibrated.ms', outputvis='MG0414_d1_sc1.ms', datacolumn='corrected', field='0', keepflags=False)
</source>
Now we can use this newly created measurement set to make a better continuum image and subsequently perform the next round of self-calibration:
<source lang='python'>
# In CASA
tclean(vis='MG0414_d1_sc1.ms', datacolumn='data', imagename='MG0414_d1_cont_sc1.R03', spw='0:4~50;60~147;156~250;350~460', specmode='mfs', niter=500, threshold='500mJy', imsize=[256,256],cell='1arcsec', weighting='briggs', robust=0.2, savemodel='modelcolumn')
</source>
<source lang='python'>
# In CASA
gaincal(vis='MG0414_d1_sc1.ms', caltable='sc2.gcal', field='0', spw='0', refant='ea04', calmode='p', solint='30s', combine='', minsnr=3.0)
</source>
<source lang='python'>
# In CASA
applycal(vis='MG0414_d1_sc1.ms', field='0', gaintable=['sc2.gcal'], interp='linear')
</source>
<source lang='python'>
# In CASA
split(vis='MG0414_d1_sc1.ms', outputvis='MG0414_d1_sc2.ms', datacolumn='corrected', field='0', keepflags=False)
</source>
<source lang='python'>
# In CASA
tclean(vis='MG0414_d1_sc2.ms', datacolumn='data', imagename='MG0414_d1_cont_sc2_R03', spw='0:4~50;60~147;156~250;350~460', specmode='mfs', niter=1000, threshold='100mJy', imsize=[256,256], cell='1arcsec', weighting='briggs', robust=0.2, savemodel='modelcolumn')
</source>
Now that we have a good model of the continuum in the field of MG0414+0354, we will attempt a final self-calibration using both the phases and amplitudes via CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.calibration.gaincal.html gaincal] :
[[File:CARTA-MG0414_d1_cont_sc3ap_R03-6.5.2.png|800px|thumb|right|Figure 11. Image obtained after self-calibration. Compare with image shown in Figure 10.]]
<source lang='python'>
# In CASA
gaincal(vis='MG0414_d1_sc2.ms', caltable='sc3ap.gcal', field='0', spw='0', refant='ea04', calmode='ap', solint='30s', combine='', minsnr=3.0, solnorm=True)
</source>
<source lang='python'>
# In CASA
applycal(vis='MG0414_d1_sc2.ms', field='0', gaintable=['sc3ap.gcal'], interp='linear')
</source>
<source lang='python'>
# In CASA
split(vis='MG0414_d1_sc2.ms', outputvis='MG0414_d1_sc3ap.ms', datacolumn='corrected', field='0', keepflags=False)
</source>
[[File:MG0414_sc3ap_msview_6.5.2.png|right|thumb|800px|Figure 12. Display of time vs baseline for the MG0414+0354_d1_sc3ap.ms measurement set in [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.visualization.msview.html msview], with the animation axis set to 'channel'.]]
<source lang='python'>
# In CASA
tclean(vis='MG0414_d1_sc3ap.ms', datacolumn='data', imagename='MG0414_d1_cont_sc3ap_R03', spw='0:4~50;60~147;156~250;350~460', specmode='mfs', niter=1000, threshold='100mJy', imsize=[256,256], cell='1arcsec', weighting='briggs', robust=0.2, savemodel='modelcolumn')
</source>
The output continuum image from the last iteration of CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.imaging.tclean.html tclean] continues to be an improvement over previous imaging, as you can see using [https://cartavis.org/ CARTA] again (Figure 11 showing results after three passes of self-calibration).
However, now it also shows that we are not likely to gain further improvement with additional rounds of self-calibration (you can inspect this yourself by comparing the images created across the different iterations in [https://cartavis.org/ CARTA]). This indicates that we now have a calibrated data-set of MG0414+0354 that is almost ready for final imaging. Before imaging, we need to flag the target data to remove poor-quality data. First, we run an automated flagging routine using CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.flagging.flagdata.html flagdata] with the parameter <i>mode='rflag'</i>:
<source lang='python'>
# In CASA
flagdata(vis='MG0414_d1_sc3ap.ms', mode='rflag', datacolumn='data')
</source>
After having ran this automated flagging, we inspect the data via the CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.visualization.msview.html msview] by displaying ''time'' vs ''channel'' for the different baselines, or you may display ''time'' vs ''baseline'' for the different channels:
<source lang='python'>
# In CASA
msview('MG0414_d1_sc3ap.ms')
</source>
Keep in mind that when the data is properly calibrated, it is easier to recognize low-level imperfections which may have to be flagged to improve the image quality. In this case, the noisier baselines are seen with lighter colors (see Figure 12). Hovering over these baselines will report a number pair such as 8-16, 8-17, etc. in the [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.visualization.msview.html msview]. From this we can tell that baselines 8-# and 12-# (which correspond to antennas ea09 and ea14) have significantly higher noise than the other antennas, something which was not picked up by the automated flagging routine. Additionally, for a number of baselines, the data quality looks relatively poor. We flag these antennas and baselines via the CASA [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.flagging.flagdata.html flagdata] task before imaging the data.
<source lang='python'>
# In CASA
flagdata(vis='MG0414_d1_sc3ap.ms', mode='manual', antenna='ea09', datacolumn='data')
flagdata(vis='MG0414_d1_sc3ap.ms', mode='manual', antenna='ea14', datacolumn='data')
</source>
To flag offending individual baselines, in CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.flagging.flagdata.html flagdata] you may use <i>antenna='antenna&antenna(&antenna...)'</i>.
Flagging data can be an iterative process. Often, the better the data is calibrated, the easier it is to recognize low-level imperfections in it. To optimize the data reduction, you can script the calibration process, and opt to flag the data also in the original data set (i.e., before calibration), after which you run the entire calibration again.
When you flag these antennas on the MS before the self calibration took place, you must go back and rerun the calibration before the final imaging, starting at the '''<i>Delay calibration</i>''' section. If you rerun the calibration you will either have to rename your measurement sets created from split during self calibration, or delete them in order to redo self calibration as split will not overwrite a ms name.
We image the data using the CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.imaging.tclean.html tclean]. A 0.5Jy threshold was chosen for this command by interactively cleaning a single channel of the final self calibrated image and taking the 5σ of the noise levels once the image was cleaned. With an ideal threshold found, we can now clean all channels (this process can take hours to complete depending on the machine you are using):
<source lang='python'>
# In CASA
tclean(vis='MG0414_d1_sc3ap.ms', datacolumn='data', imagename='MG0414_d1_line_sc3ap_vel_R03', spw='0', specmode='cube', outframe='bary', veltype='optical', restfreq='390.600MHz', niter=10000000, threshold='0.05Jy', imsize=[256,256], cell='1arcsec', pbcor=True, weighting='briggs', robust=0.2, perchanweightdensity=True, savemodel='none',interpolation='nearest')
</source>
OBS: From CASA version 6.2.0 onwards, one new input parameter has been added to the task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.imaging.tclean.html tclean]. The psfcutoff parameter allows fitting non-Gaussian PSF. The default value is 0.35, but can vary from 0.01 to 0.99. It is a parameter for advanced users, and it should rarely be changed. If a user needs to use it, the resulting fits should be carefully checked. We have used the default value in all [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.imaging.tclean.html tclean] runs of this tutorial since we have not set it at any time.
We can inspect the resulting line-data cube, again using [https://cartavis.org/ CARTA]. The HI absorption feature is clearly present (see Figure 14). If you did NOT re-run the calibration sequence, you should get a plot that should look similar to Figures 13 and 14. To see your spectral line, open the last image of the source. Go to a channel that shows the source, then zoom in on the source. Click the ''Point marking'' icon, find the brightest pixel in the source and click on it. Now click on the ''spectral profile tool'' button to get the single point profile of the spectrum. Here you will see a spectrum with 2 dips that drop to zero; these dips are a part of the flagging we have done. You will need to zoom in in order to notice any spectral lines due to the auto scaling; zoom in by clicking on the graph and creating a yellow box then release.
The data cube contains both continuum and line emission, which can optionally be separated using the CASA task [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.manipulation.uvcontsub.html uvcontsub]. In principle, this task can be used for continuum subtraction and fitting in the uv domain; however, combining spws for the continuum fitting is not yet supported in the updated [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.manipulation.uvcontsub.html uvcontsub] in CASA 6.5.2, and for that particular functionality the [https://casadocs.readthedocs.io/en/v6.5.2/api/tt/casatasks.manipulation.uvcontsub_old.html uvcontsub_old] should be used for the time being.
[[File:CARTA-cube-6.5.2.png|1250px|thumb|center|Figure 13. Test-data of the redshifted HI 21cm absorption line on top of the radio continuum in MG 0414+0354 (i.e., the continuum was not subtracted); see More, Carilli & Mention 1999 (Apj, 510, 87) for the original detection. It is the final cleaned image here displayed using CARTA. As can be read from the bottom right panel, shown is channel 115, at 388.07 MHz. The profiles on the right show a single point 1D spectrum on x and y axes.]]
{{Checked 6.5.2}}
<!-- NO LONGER RELEVANT, CARTA DOES EVERYTHING IN ONE, PREV. FIG. [[File:MG0414 d1 line sc3ap vel R03 recal 6.2.0.png|1100px|thumb|right|Figure 14. Left: The Single Point Spectral Profile with the auto scaling. Noticed the tip, make a tight horizontal yellow box around the main signal, like shown in figure. On right: The same after we zoom it.]]-->
<!--IMVIEW: Alternative Tip: You can click in the save icon (top left) in the Spectral Profile tool and export the spectral profile as a text file. And then, you can use any program you find appropriate to make the plot. You can see small differences in the continuum flux level, depending on the pixel you have selected to see the line profile. However, you should be able to identify the line in any of the bright central pixels.-->
<!-- --[[User:bemonts|Bjorn Emonts]]   
-- 5.0 corrections: [[User:kradford| Karlee Radford]] (2018-03-21)
-- Edits [[User:tperreau| Tony Perreault]] (26 Apr 2018)
-- Edits [[User:Hmedlin | Heidi Medlin]] (27 April 2018)
-- Yet more edits [[User:tperreau| Tony Perreault]] (2 May 2018)
-- And yet more edits [[User:tperreau| Tony Perreault]] (May the 4th be with you 2018)
-- 5.4.0 updates: [User:kradford| Karlee Radford]] (2018-10-11)
-- 5.5.0 updates: [User:asobotka| Alex Sobotka]] (3 July 2019)
-- 5.7.0 updates: [User:pbeaklin| Pedro P.B. Beaklini]] (29 September 2020)
-- Changed last figure: [User:pbeaklin| Pedro P.B. Beaklini]] (7 October 2020)
-- 5.7.2 updates [User:pbeaklin| Pedro P.B. Beaklini]]. Cheking tec_map (21 December 2020)
-- 6.2.0 updates [User:pbeaklin| Pedro P.B. Beaklini]]. Added before and after selfcal figures (21 June 2021)
-- 6.2.0 updates [User:pbeaklin| Pedro P.B. Beaklini]]. Changed task links, changed robust and added 'nearest' on last clean. (01 July 2021)
-- 6.5.2 updates [User:akapinsk| Anna D Kapinska]]. Exchanged the ms.tgz file for download (error of unexpected EOF when unpacking), and exchanged some figures for clarity, plus some minor corrections. Removed imview, added CARTA. (13 Feb 2023)
-->

Latest revision as of 10:17, 8 February 2024

Last checked on CASA Version 6.5.4

Retrieved from "https://casaguides.nrao.edu/index.php?title=Template:Checked_6.5.4&oldid=36501"