; MERLIN ;--------------------------------------------------------------- ;! Procedure to process MERLIN data ;# ; This software is the subject of a User agreement and is ; confidential in nature. It shall not be sold or otherwise ; made available or disclosed to third parties. ;--------------------------------------------------------------- MERLIN LLLLLLLLLLLLUUUUUUUUUUUU CCCCCCCCCCCCCCCCCCCCCCCCCCCCC MERLIN: Procedure to process MERLIN data. Type HELP MERLIN for more details of inputs and EXPLAIN MERLIN for a description of outputs. All editing and calibration is reversible. PROTOTYPE - full spectral capabilities not yet enabled. Complaints/suggestions to amsr@jb.man.ac.uk Based on P Diamond AIPS procedure AUGUST 2005 INDISK No. of disk drive to use DIR Environment variable for directory containing FITS files. DOLOAD = 0 Data already concatenated and loaded with name TARGET.MULTTB (default) = 1 Sources are in separate files SOURCE_NAME.FITS in DIR area = 2 Input is multi-source file with name TARGET.FITS in DIR area TARGET TARGET source name (required) PHCAL Phase reference source name PACAL Polarization p.a. cal name ACAL Point cal name (required) AFLUX Flux (Jy) of point source (see DOFLUX and HELP MERLIN) DOPLOT = 0 No plots/printout = 1 Send to printer = 2 Use TKPL (interactive) = 3 Print to disc in DIR DOEDIT = 0 Default: edit extreme, isolated points (usually interference) = 1 Run IBLED (interactive). =-1 No editing DOPHSC = 0 No phase reference souce = 1 Assume PHCAL is a point = 2 Three rounds of self-cal starting with point model = 3 Load CLEAN map for model, from DIR area. It must be called PHCAL.ICL001.FITS DOPLZN = 0 No linear polarization = 1 Perform polarization corrections and make polarization images DOTGSC = 0 No TARGET self-cal (see (HELP if DOPHSC 0). = 1 Single pass of phase corr with solint=5 min (for weak sources) = 2 Five cycles of phase corr with decreasing solint (for simple sources) = 3 Ten cycles of phase corr (medium/complex sources) = 4 Ten cycles of phase then 'A&P' (strong sources) DOBP =-1 Do not correct bandpass (default for single- channel data) >=0 Correct bandpass (default for multi-channel data) DOCHANS = 0 Default: data averaged to single channel = 1 Spectral line imaging: Final image is datacube ONLY IF DOPLOT 0 = 2 Wide-field imaging: final single image made from multi-channel data SYSVEL If line source, velocity of TARGET at band centre (km/s) in LSR, radio definition. RESTFREQ If line source, rest freq. of transition (Hz). Use RESTFREQ(1)+RESTFREQ(2). BCHAN Start channel for reference feature. If continuum leave as 1 ONLY CURRENT OPTION ECHAN End channel for reference feature. If continuum leave as 0 ONLY CURRENT OPTION REFANT Reference antenna number (use 0 for MERLIN defaults) DOFLUX = 0 Use AFLUX for ACAL and existing flux scale for all sources. Default, but see HELP MERLIN = 1 At 1.2-7 GHz use ACAL and 3C286 to refine flux scale for all sources. = 2 Compensate for resolved ACAL (NOT IMPLIMENTED) DOUVR If > 0 then will apply uv-range restrictions during TARGET self-cal. If unsure or your TARGET has significant extended structure set DOUVR = 0 DOSEARCH If > 0 then will search for the TARGET within the field. Not allowed if DOPHSC 0 DOREWT = 0 No reweighting (Default) = 1 Reweight the uv-data according to different antenna sensitivities before final TARGET image IMSIZE 0.0 8192. Minimum image size for TARGET BMAJ Major axis of TARGET restoring beam. Default: natural beam size. BMIN Minor axis of TARGET restoring beam. Default: natural beam size. BPA Position angle of TARGET restoring beam. Default: natural beam angle. DOTV =-1 Default: no interactive imaging = 1 INTERACTIVE TARGET imaging = 2 INTERACTIVE PHCAL and TARGET IMAGING ---------------------------------------------------------------- MERLIN Type: Procedure Use: To perform basic editing, calibration and imaging on MERLIN data. You need the runfile MERLIN.ehex where ehex is your AIPS number in 3-digit ehex e.g. for AIPS No. 895 use MERLIN.0OV Before starting AIPS, in the same directory, set environment variable(s) for the MERLIN runfile etc. (also see DIR, DOPLOT) e.g. % setenv RUNFIL /scratch/octans_1/visitor1 % setenv DATA /scratch/octans_1/visitor1/FITS In AIPS, type > RUN MERLIN > INP MERLIN and when you are happy with inputs: > MERLIN You can use TPUT and TGET MERLIN but you must not alter inputs to other tasks/verbs after doing RUN MERLIN. If MERLIN fails part- way you should remove all new files and start again or seek help. All calibration and editing is stored in extension tables attatched to the input file, and can be reversed by deleting the appropriate table. See EXPLAIN MERLIN for more details. NB if you type 'RESTORE 0' you will have to re-RUN MERLIN. If you use INTERACTIVE settings, procedure will wait for your input. Author: P. Diamond/A.M.S. Richards Documentor: A.M.S. Richards June 2003 Adverbs: INDISK....AIPS disk to write data to (0 = any) DIR.......Environment variable for directory holding FITS files, e.g. DIR 'DATA' in above example. Can be 'RUNFIL' Default ' ' = PWD. Also see DOPLOT DOLOAD....= 0 One multi-source file already in AIPS called TARGET.MULTTB where TARGET is your TARGET; file must also contain point source cal (ACAL) and (if used) phase reference source (PHCAL) and flux/polarisation angle cal (PACAL,default 3C286). This is normally 'tdproc' output. Default 0 NB All pre-existing CL, SN and BP tables are DELETED = 1 TARGET.FITS, ACAL.FITS, PHCAL.FITS and PACAL.FITS (as required) are in separate files in DIR area = 2 one file (as for DOLOAD=0) called TARGET.FITS in DIR area TARGET....Name of TARGET source REQUIRED All source names must be EXACTLY as used in MERLIN observations PHCAL.....Phase reference source (required if DOPHSC > 0) PACAL.....Polarization position angle calibration source (must be in data file if linear polarization wanted, default 3C286) ACAL......Point source cal, also BP cal if wanted. REQUIRED AFLUX.....= 0 Default value will be used at 1.2 - 8 GHz if ACAL = OQ208, 0552+398 or 2134+004 Use DOFLUX = 1 if 3C286 is present. Otherwise set best known value. See DOFLUX and EXPLAIN MERLIN DOPLOT....= 0 No plots = 1 Hard copy printouts of solution and image plots and text files of scan listings and some calibration details = 2 Plots/text sent to Tek server/screen. INTERACTIVE = 3 Plots/text printed to disc in DIR area. NB AIPS has print environment length limitation, avoid long directory paths DOEDIT....= 0 VPLOT is used automatically to edit visibility amplitudes which are >10 sigma in raw data averaged over ever 3 min. = 1 IBLED will be run on each source in turn. Familiarise yourself with this first. FLAG TIMERANGE flags all sources; other commands flag only current source. STOKES FLAG 1111 avoids spurious polarization results. INTERACTIVE =-1 No editing at all in this procedure. If there is already an FG table it will be applied. DOPHSC....= 0 No phase reference source used. > 0 Options for using phase reference source PHCAL to derive phase and amplitude solutions which will be transferred to the TARGET. = 1 PHCAL is assumed to be point-like = 2 Derive first phase solutions using point model, map PHCAL for 2nd and 3rd (phase, amp & phase) solution rounds. = 3 Load existing image PHCAL.ICLN001.FITS from DIR area for first model. Highest No. CC table is used. For DOPHSC >1, models use merged CC down to first negative. Data are averaged from BCHAN to ECHAN for calibration and solutions are copied back to multi-source data set. DOPLZN....= 1 Use PHCAL to derive corrections for polarization leakage and PACAL to correct polarization angle. DOTGSC....= 0 No self-calibration of TARGET, one image is made after applying PHSREF solutions. Will be rubbish if DOPHSC 0. = 1 After applying PHSREF solutions, TARGET is imaged and CC down to first negative are used as model for phase-only self-calibration, 5 mins solution interval. Use for weak sources (typically peak snr >20 but <100). If DOPHSC 0 a default point model is used, see EXPLAIN = 2 After 1, *four more cycles of phase-only self-cal with decreasing solution intervals (for simple sources). = 3 After 1, *nine more cyles of phase-only self-cal with decreasing solution intervals (for medium/comples sources). = 4 After 3, *amplitude and phase self-cal (strong sources, beware introducing symmetric or other artefacts if uv coverage is poor e.g. at low Dec or 22 GHz). All imaging uses completely natural weighting (plus final reweighting if DOREWT 1). All self-calibration uses data averaged from BCHAN to ECHAN; solutions are applied to all TARGET data channels. You can continue manual self-cal; for wide-field TARGETS you might want to use multi-channel data. NB Total intensity (Stokes I) images are used as models. If your TARGET is significantly circularly polarized (e.g. many OH masers) you should set DOTGSC 0 and see MERLIN User Guide for manual self-cal. DOBP......=-1 Do not do bandpass correction (default if data are in single channel) >=0 Derive bandpass corrections using ACAL (default if multi-channel data). DOCHANS...= 0 All except end channels are averaged for all imaging (DEFAULT) = 1 Spectral line imaging: individual channels imaged separately in final image to produce a datacube ONLY AVAILABLE IF DOPLOT 0 = 2 Wide-field imaging - final single image is made from multi- channel data (calibration performed on averaged data) In all cases the channels from ECHAN to BCHAN are averaged and used for all calibration and imaging during calibration. SYSVEL....If line source, enter TARGET velocity (km/s) in band centre (chan 128 for nominal 256 chans etc.) Use radio LSR convention. RESTFREQ..If line source, enter rest frequency (Hz) in double precision BCHAN.....Start channel of reference feature for line TARGET self-cal. Leave as 1 for continuum (default) - first ~5% chans excluded. ECHAN.....End channel of reference feature for line TARGET self-cal. Leave as 0 for continuum (default) last ~5% chans excluded. ** DO NOT RESTRICT CHANS at present please REFANT....Reference antenna. Default is Mk2, Da, Ta or Kn in order of preference. Use same REFANT throughout for polarization imaging, use MK if LO is in the array or use closest antenna with mostly good data. DOFLUX... = 0 Use preset (usually tdproc) flux scale. = 1 Refine flux scale using well-edited 3C286 and ACAL Currently only possible at 1.2-7 GHz; 3C286 must be present. AFLUX must be accurate to ~50% = 2 Use secondary flux-cal to compensate for resolved ACAL NOT YET IMPLEMENTED DOUVR.....= 0 No restrictions on uv range (default) = 1 Start self-cal by giving a low weight to inner portions of uv plane to avoid confusing sources, all data is gradually added in during subsequent rounds of self-cal. Use DOUVR=0 for extended TARGETs (more than 5 - 10 beamwidths). At K-band DOUVR=0 is almost always most suitable. DOSEARCH..= 0 DEFAULT images will be centred on input positions = 1 TARGET images will be shifted to be centred on position found for TARGET. Use ONLY if DOPHSC >0, if the position of your TARGET may be > ~IMSIZ/4 away from the input position. DOREWT....= 1 For the final TARGET image, reweight the data from each antenna in proportion to its sensitivity, see MUG (web ref below). If data are well-edited and well- calibrated this optimises signal to noise but changes the natural beam-size: if Lo is present the beam becomes larger; at 22 GHz if Ca is present the beam becomes smaller. You can adjust the beam-size by running IMAGR manually using suitable values of UVWTFN, ROBUST, BMAJ, BMIN, etc. IMSIZE....Target image size. Minimum size for reliable deconvolution of complex/extended sources is 512x512 pixels (default). Use 512x1024 or larger at low Dec, use larger for larger TARGETS/poor uv coverage. BMAJ......Major axis (mas) of beam to be used for final TARGET image. BMIN......Minor axis (mas) of beam to be used for final TARGET image. BPA.......Position angle (N->E) of beam to be used for final TARGET image. Default for BMAJ, BMIN, BPA is beam fitted using natural weighting. If you want to change by more than ~10% consider running IMAGR manually with different UVWTFN, ROBUST etc. DOTV......=-1 Default - no interactive imaging = 1 For TARGET imaging, IMAGR starts in interactive mode. = 2 PHCAL and TARGET imaging start in interactive mode. ---------------------------------------------------------------- See the MERLIN User Guide for more information about manual processing. MUG http://www.merlin.ac.uk/user_guide/ AFLUX and DOFLUX If your input flux scale is inaccurate by >~50%, UVMOD should be used to improve it before running the MERLIN procedure. At 1.2 - 8 GHz, if you know the flux density AFLUX of your ACAL on MERLIN scales to ~50%, e.g. extrapolate to your epoch from from http://www.merlin.ac.uk/user_guide/collect_cal_flux.txt, and 3C286 is present, then DOFLUX = 1 will use the shorter baselines to refine the flux scale and apply the solution to all sources. If AFLUX=O & ACAL = OQ208/1404+286, 0552+398/DA193 or 2134+004, a default AFLUX value will be used; you should set DOFLUX 1 if 3C286 is present. This will give a flux scale accurate to a few percent. For other ACALs or very high/poor accuracy consult MERLIN staff. 22 GHz flux densities are available courtesy of H Tersranta, Metsahovi. OUTPUT FILES TARGET.MULTTB contains all the input sources concatenated into one file with OUTPUT TABLES depending on the options selected: AN (anten): If DOPLZN 1 this contains polarization corections FG (flag): Any input tables preserved and used; higher number tables contain accumulated flags, so highest no. contains all. BP (bpass): If doband > -1 and data are multi-channel CL (calib): 1 Blank template 2 Global amplitude correction per antenna/polarisation if DOFLUX 1, otherwise also blank template 3,4 Phase-only PHCAL solutions if DOPHSC > 0 5 Amp&Phase PHCAL solutions if DOPHSC > 0 Also polarization PA correction if DOPLZN 1 6,7,8 If DOTGSC > 0 then these are equivalent to TARGET.MULTSC CL 2,3,4 (see below). All the files will have an HI (history) file which can be inspected using verb PRTHI to see the settings used in its ancestral tasks. TARGET.MULTSC is made by applying any relevant BP, CL and AN corrections and splitting out the TARGET from TARGET.MULTTB. If DOTGSC > 0 then CL 2: First self-cal phase corrections - if DOPHSC > 0 these were derived using the CC from the first TARGET map after applying the PHCAL phase corrections so the TARGET position is as accurate as possible; - if DOPHSC 0 these were derived using a default point model so the TARGET position will appear to be at the input coordinates regardless of the true position. CL 3: CL 2 + final phase-only corrections if DOPHSC > 1 CL 4: CL 3 + amp&phase corrections if DOPHSC = 4 TARGET.SPLIT is the final split file made from TARGET.MULTTB applying all calibration; if DOCHAN >0 then all except the end channels will be kept separate; otherwise they are averaged. If DOREWT > 0 then the appropriate antenna weights are applied, see MUG web ref. above. TARGET.IBM001 is the last beam made TARGET.ICL001.* are the final CLEANed images, the highest number * being the most calibrated if DOTGSC > 0 (see notes above on position). The final image is made from TARGET.SPLIT. TARGET.POLI and .POLA are the polarized intensity and angle images if DOPLZN > 0 PHCAL image names follow the same convention. If DOSEARCH 1, inspect the final ICL001.* header and history file to see if a large shift (>~20 beams) was applied in IMAGR; if so, you are advised to use UVFIX instead of using a shift during imaging. Calibration solutions for the ACAL and PACAL etc. are not all retained (apart from CL2) as they are not applicable to other sources; inspect the plots to check these. Some SN tables are preserved as follows: PHCAL.TASAV SN 1: Phase-only solutions ACAL.TASAV - only present if DOFLUX > 0 SN 4: Time-constant amplitude correction (no phase solutions), used to make TARGET.MULTTB CL 2 and applied to all data thereafter NOTES If you are using an old version of AIPS (e.g. last century) some adverbs e.g. 'nplot' are different which may affect printout and a few tasks e.g. BPASS, CLCAL may not work as desired - you are advised to download the latest frozen version of aips (currently 2005) or check your results very carefully.