VS05q Results
Submitted checklist.txt and checklist.tex is also available |
[Summary | AIPS | Difmap ] |
Abstract
Fringes to HALCA were found during only part of the NZ pass (about one hour in total). No fringes in IF-2 at Shanghai and very low amplitudes at Kashima, suggesting a serious problem. Kashima data were eventually flagged. Analysis of the calibrator source showed Shanghai gain to be underestimated. Multiplying Shanghai gains by a factor of ~1.7 resolved this problem. Usuda gains were also found to be low by a factor of ~1.6. An image and modelfit of the survey source was made. EBF: 07-OCT-2005: Some editing of data core is 0.3 +/- 0.1 |
|
|
in |
in |
out |
out |
|
FITLD | douvcomp = 1 |
|
|
|
|
Loaded previously calibrated data for re-analysis |
EXTDEST |
|
|
|
|
Removed previous CL, SN, TY, GC, NX tables to return data to an uncalibrated state | |
MSORT |
|
|
|
|
Not really necessary, but did it anyway. | |
INDXR | cparm(3) 0.25 |
|
|
|
|
|
UVFLG | antennas = 2,0 bif 2; eif 2 opcode 'flag' reason 'Shanghai IF 2 bad' |
|
|
|
|
Flagged shaghai IF 2 as it is known to contain invalid data. |
POSSM | docalib -1 |
|
|
|
|
These plots and DQA show Mopra to be best reference antenna |
ANTAB |
|
|
|
|
Gain curve information from VSOP calibration web page. Nominal Tsys used for Kashima, the rest from log files. Limited data available from Mopra and Usuda. | |
APCAL |
|
|
|
|
||
CLCAL | refant 4 |
|
|
|
|
Amplitude calibration solutions applied |
TACOP | inext 'cl' |
|
|
|
|
copy CL2 to CL3 to protect CL2 in case of a mistake in CLCOR |
CLCOR | opcode 'gain' clcorprm=0.906,0 gainver 3 |
|
|
|
|
Penticton gain correction |
CLCOR | opcode 'sbdl' antennas = 1,0 clcorprm -480,-480 antennas = 2,0 clcorprm 1480,1480 antennas = 3,0 clcorprm 1845,1845 antennas = 5,0 clcorprm -1700,-1700 antennas = 6,0 clcorprm -1888,-1888 |
|
|
|
|
Known delay offsets from DQA report applied before fringe-fitting. |
KRING | docalib = 1 gainuse = 3 dofit = 0 refant = 0 soltyp 'nols' search = 6, 4, 1 solmode = 'SBD' bparm(5) = 0.5 cparm = 0, 0, 2, 1000, 100, -1, -1, -1, 0, 0 antwt = 0 solint = 6 prtlev 1 |
|
|
|
|
fringes found at most times on all baselines except to HALCA on the second tracking pass, despite the projected baselines crossing some ground baselines. |
KRING | docalib = 1 gainuse = 3 dofit = 0 refant = 0 soltyp 'nols' search = 6, 4, 1 solmode = 'SBD' bparm(5) = 0.5 cparm = 0, 0, 2, 2500, 1000, -1, -1, -1, 0, 0 antwt = 0 solint = 12 prtlev 1 timerang 0,23,30,00,1,1,30,00 |
|
|
|
|
concentrating on second tracking pass. Increased solint to 12min but still no convincing fringes. Return to FRING and 6 min solint with smaller windows for final fringe-fit. |
FRING | docalib = 1 gainuse = 3 refant = 4 search = 6, 1 dparm = 3, 200, 100, 0.5 solint = 6 aparm = 0, 0, 0, 0, 0, 1, 0.1, 0, 1, 0 |
|
|
|
|
Final fringe-fit. |
TBOUT | outfile 'survey:sn5.txt' docrt = 1000 |
|
|
|
|
Wrote fringe fit SN table out for editing in snedit.pl |
snedit.pl | output file SN6.TXT |
|
|
|
|
Used to flag bad solutions. Much easier than SNSMO |
TBIN | infile 'survey:sn6.txt' |
|
|
|
|
Edited SN table read in |
CLCAL | refant 4 |
|
|
|
|
Amplitude calibration solutions applied |
POSSM |
|
|
|
|
Plot 10min at two times: 0/21:00 to 21:10 and 1/00:20 to 1/00:30. Shows big phase variation near edge of bandpass for Shanghai and smaller variations for Usuda. Chans 30 to 220 OK | |
SPLIT | aparm = 2, 0.5, 0 chansel 30, 220, 1 flagver 1 docalib 1 |
|
|
|
|
|
FITTP |
|
|
|
|
Data written out for analysis in Difmap | |
CLCOR | opcode 'gain' antennas 2,0 clcorprm 1.8,0 |
|
|
|
|
Gain correction for Shanghai after Difmap examination. |
SPLIT | aparm = 2, 0.5, 0 chansel 30, 220, 1 flagver 1 docalib 1 |
|
|
|
|
|
FITTP |
|
|
|
|
Data written out for analysis in Difmap again | |
CLCOR | opcode 'gain' antennas 6,0 clcorprm 1.7,0 |
|
|
|
|
Gain correction for Usuda after second Difmap examination. |
SPLIT | aparm = 2, 0.5, 0 chansel 30, 220, 1 flagver 1 docalib 1 |
|
|
|
|
|
FITTP |
|
|
|
|
Data written out for analysis in Difmap again! | |
CLCOR | opcode 'gain' antennas 5,0 clcorprm 32.5,0 |
|
|
|
|
Gain correction for Kashima after third Difmap examination. |
SPLIT | aparm = 2, 0.5, 0 chansel 30, 220, 1 flagver 1 docalib 1 |
|
|
|
|
|
FITTP |
|
|
|
|
Data written out for analysis in Difmap again! | |
CLCOR | opcode 'gain' antennas 2,6,0 clcorprm 0.92,0 |
|
|
|
|
Gain correction for Usuda and Shanghai after fourth Difmap examination. |
SPLIT | aparm = 2, 0.5, 0 chansel 30, 220, 1 flagver 1 docalib 1 |
|
|
|
|
|
FITTP |
|
|
|
|
Data written out for analysis in Difmap again! |
AIPS Input Files
ANTAB.TXT TABIN.TXT |
Comments No INDXR.TXT needed as there were no gain jumps during this experiment. |
Additional Files
Name | Description |
POSSM01_AP.PS | Post-calibration spectra (IF 1 and 2) at 0/21:00 |
POSSM02_AP.PS | Post-calibration spectra (IF 1 and 2) at 1/00:20 |
AIPS Output Files
Phase selfcal applied? | Yes |
Global amplitude selfcal (gscal) applied? | Yes |
Amplitude selfcal on non-global timescales used? | No |
Modelfit made and saved? | Yes |
Image Made and saved? | Yes |
Comments on Difmap processing
After initial AIPS extraction, 1334-127 data loaded and examined. Only Hart, Mopra and Shanghai observed this source. By comparison with VLBA data (and accounting for variability), Shanghai gain may be down by a factor of about 1.8. First look at J1316-33 supports this and indicates Usuda may be low too. CLCOR in AIPS used to apply 1.8 Gain correction. Corrected 1334-127 data looks OK now. A simple single circular Gaussian was modelfitted and saved with vs_final_mod. The corrected J1316 data looks a little odd. Ududa gain may be out by about a factor of 1.7. CLCOR used to apply this and data exported again. J1316 data now looks OK. Flux on short Usuda-Shang baseline ~1.5 Jy maybe a bit high (expect about 1.2 Jy) but will see how it goes. Also, Kashima data has very low amplitudes! Assuming Mopra and Usuda gains are now OK, the Mop-Kashima (0/22:30) and Mop-Usuda (0/21:00) baselines are almost the same, so can guess Kashima correction = 1.3 Jy / 0.04 Jy = 32.5. Corrected in AIPS, but resulting Kashima data is very noisy - bad pointing perhaps. Decided to flag Kashima. After best possible image made with just cleam/selfcal, gscal was run. The result of this flagged all of IF2 (because no amplitude closure) and in IF1, Shanghai and Usuda were re-scaled by 0.91 and 0.93 respectively while VSOP and Mopra were increased by 1.04 and 1.07 respectively (no solution for Hart as no closure). This is most likely an over-compensation of Usuda and Shanghai gains, so applying just those corrections (0.92 each) in AIPS with CLCOR. Looked at 1334-127 data again, fitted simple circular gaussian to data and saved using vs_final_mod. J1316 imaged again for last time. Used selfflag to stop flagging of unusable data after phase-selfcal because Mopra-Hart baseline provides constraints that would otherwise be lost with no significant loss of image quality as a result. Saved with vs_final_img. Model fitting: Started with two circular Gaussians. vs_wplot suggests a VSOP weight increase of about 100. This is set using selfant vsop,false,100. Modelfit converged after about 70 iterations with uvstat(rms) = 0.173 Jy. Changed secondary to elliptical Gaussian and re-converged to get uvstat(rms) = 0.171 Jy, so no significant improvement and circulars Gaussians will do. saved with vs_final_mod. EBF: 08-OCT-2005: Some editing. See uv data in J1616_FOMALONT_AIPS.FITS Core 0.3 +/- 0.1 |
Image | ||||
Ampl, Phase vs (u,v) Distance PS (39 kb) JPEG (42 kb) |
Ampl, Phase vs Time PS (250 kb) JPEG 1, 2, 3 |
(u,v) Coverage PS (21 kb) JPEG (26 kb) | Closure Phase vs Time PS (227 kb) JPEG 1, 2 |
Clean Image PS (32 kb) JPEG (46 kb) |
Modelfit | ||||
Ampl, Phase vs (u,v) Distance PS (39 kb) JPEG (42 kb) |
Ampl, Phase vs Time PS (250 kb) JPEG 1, 2, 3 |
(u,v) Coverage PS (21 kb) JPEG (26 kb) | Closure Phase vs Time PS (227 kb) JPEG 1, 2 |
Image with model components PS (35 kb) JPEG (39 kb) |
More details on the Images and Modelfits are found in this page.