2.5 Spectral Line Observations

SCHED can set observing frequencies for spectral line observations based on velocities provided in the source catalog and rest frequencies provided in a separate type of input. This option is invoked by specifying DOPPLER and can be turned off with NODOP (DOPCAL is an obsolete form that was too easy to confuse with “DO PCAL”). If DOPPLER is invoked, SCHED calculates the velocity of the center of the Earth with respect to the designated reference frame (VREF in the direction of the source at the time of the middle of the project. From this velocity, the source velocity from the source catalog, and the rest frequency, the required observing center frequency is calculated. The antennas need to know the LO settings so SCHED must know the bandwidth. The bandwidth will usually be obtained from the setup file. It may be provided for the scan with the BW parameter for some systems, unfortunately not including the new digital backends such as the RDBE on the VLBA.

The reference frames supported by SCHED are the “Local Standard of Rest” or LSR VREF=L, heliocentric VREF=H, and geocentric VREF=G.

Note that channels assigned to the same BBC will be given the same frequency as the first channel on that BBC, no matter what velocities etc are given for the other channels. This will be the case when there are upper/lower sideband pairs. Their frequencies cannot be set independently. Because of the different sidebands, they will, however, cover different velocity ranges.

The frequencies derived from the Doppler calculations have to be rounded off to a value that can be set on the available synthesizers. For the Mark III/IV and legacy VLBA systems, frequencies can be set to the nearest 10 kHz. For the RDBE with the DDC personality on the VLBA, usable frequencies are multiples of 15.625 kHz. Note that, if there is a mix of systems with 10 kHz and 15.625 kHz steps, one should round to 250 kHz because that is the lowest value common to both. Other systems are different — VSOP, for example, can only be set to the nearest 1 MHz. The RDBE with the PFB personality can only do frequencies in 32 MHz steps, and those have LO dependent offsets — it is probably best to avoid DOPPLER when using the PFB. The parameter DOPINCR can be used to control the rounding.

The Doppler calculations are for the center of the Earth for the middle of the project. This implies that the frequency for each source will be constant for the duration of the project. Experience over the years with spectral line VLBI has shown this to be the preferred observing mode since it minimizes the chances of mistakes at stations that do not have automatic frequency setting. The shifts of the resulting spectra of about a km per second that result from the rotation of the Earth can be removed in post-processing with the CVEL program in the NRAO spectral line software or with a task of the same name and with similar capability in AIPS.

The LO sum used when Doppler calculations are requested are calculated by either the radio definition (VDEF=R in the source catalog ) or the optical definition (VDEF=O or VDEF=Z). With the radio definition, the LO sum is calculated as RESTFREQ * (1 - v/c) - BW/2 where RESTFREQ is the line rest frequency from the line initialization input, c is the speed of light, and BW is the bandwidth (appropriately signed according to the sideband). With the optical definition, the LO sum is RESTFREQ / (1 + v/c) - BW/2. Typically velocities for radio spectral lines in galactic sources are given in the radio definition in the LSR frame. Extragalactic velocities, on the other hand, are typically in the optical definition in the heliocentric frame. The differences in the radio and optical definitions only matter at large (typically extragalactic) velocities.

SCHED will accept redshifts if VDEF=Z in place of velocities, but be very careful that you have adequate accuracy to calculate proper frequencies - the bandwidths are typically very much smaller than the observing frequency so the velocities must be accurate.

Internally, if Doppler calculation is requested, SCHED calculates the desired observing frequencies and puts them in the same array that would be used if FREQ were used. This will override any FREQ specifications in the main schedule and any frequency specifications in the setup files. The frequencies of baseband converters will then be set properly based on the FIRSTLO for the station. Please note that the setup files must still contain a complete, valid frequency specification. This is to allow validity checking and to allow SCHED to pick up default parameters from the frequency catalog. The frequency in the setup should be close enough to the desired observing frequency that only final tuning of the BBCs is needed to get the exact desired frequency. For the VLBA with its 500 MHz IFs, this is not a serious concern.

The number of channels desired is set with NCHAN in the setup file (NCHAN in the main schedule is an obsolete parameter and is not used.). To calculate a frequency, SCHED must have, for each channel, a bandwidth, a velocity from the source catalog, and a rest frequency from the line initialization input information. If a value is missing for any channel of any parameter, the value of that parameter for channel 1 will be used. This avoids the need, for example, of specifying lots of bandwidths when they are all the same.

For continuum sources mixed in with line sources, specify NODOP for that scan to avoid the Doppler calculations.

Often it is desired to observe a continuum source at the same frequency as a line source for bandpass calibration. This can be done by specifying the line source with DOPSRC in the continuum source’s scan. The DOPSRC will be used for the Doppler calculation.

The important parameters in the SCHED keyin file for Doppler frequency calculation are listed below. Detailed descriptions are given with the descriptions of other SCHED parameters.

DOPPLER and NODOP turn the Doppler calculations on and off.

DOPSRC is used to specify the source, if it is different from the scan source, for which the Doppler calculations should be made. This is useful for bandpass calibration. Warning - as with nearly all SCHED variables, it defaults to the previous scan. After using it, be sure to set it to blank or to the next source.

DOPINCR is used to determine the level of rounding of frequencies that is used.

LINENAME specifies which group of rest frequencies to use. It must match one of the sets of lines named with LINESET in the LINEINIT input.

BW sets the bandwidth for the scan and overrides the setup file value. This was useful for switching between wideband observations on calibrators and narrow band observations on line sources. The value in the setup file will be used if it is not specified. In general, it is a good idea to use a new setup file when changing bandwidths because quite a few other parameters also change. With the new digital back ends, oversampling is now allowed so BW basically doesn’t work.

LINEINIT indicates that after the next “/”, information on the rest frequencies of spectral lines will be given. If invoked, the rest frequencies will be read, and SCHED will return to reading input for the same scan that it was on before the “/”.

The rest frequencies are specified in separate keyin inputs in the SCHED keyin file following a “/” if LINEINIT was specified. There can be one rest frequency per channel, although any not specified default to the first which is often the desired behavior. There should be one velocity per channel in the source file for each source to be observed (other than continuum calibrators). Each group of lines has a name which is then referred to using LINENAME in the scan input. The group can change with each scan, but be careful to change the setup file, too, if necessary. Up to 10 groups of lines are allowed. The parameters in the LINEINIT group are:

Name of the group of lines. LINENAME in the scan input will be used to invoke this group.

Argument: Name of up to 8 characters.

Options: Any name.

Default: None

Usage: Defaults to previous value - don’t do it!

Example: LINESET=’H2O’

Rest frequencies.

Argument: Up to one real number per channel. The first value will be used for any channels for which a value is not given.

Options: Any value.

Default: 0

Usage: Defaults to previous value.

Example: RESTFREQ=22235.08

This parameter, on a separate line with a “/”, ends the restfrequency inputs much like the ENDCAT parameter ends in-line source and station catalog inputs.

Argument: None.

Options: None.

Default: Not specified.

Usage: Defaults to previous value, but this has no effect since this will be the last line of the rest frequency input to be read.

Example: ENDLINES /

A fairly extensive list of possible rest frequencies is given below. These frequencies are not guaranteed. If anyone finds an error, please notify Craig Walker (cwalkernrao.edu).

lineinit /  
!   Frequencies from Reid and Moran’s Annual Reviews article preprint.  
!   Do not keep more than 10 lines in a SCHED input file.  
lineset=’OH1612’  restfreq=1612.231   /  
lineset=’OH1665’  restfreq=1665.402   /  
lineset=’OH1667’  restfreq=1667.359   /  
lineset=’OH1720’  restfreq=1720.530   /  
lineset=’OH4660’  restfreq=4660.42    /  
lineset=’OH4765’  restfreq=4765.562   /  
lineset=’OH6030’  restfreq=6030.747   /  
lineset=’OH6035’  restfreq=6035.092   /  
lineset=’CH3OH’   restfreq=6668.5192  /  Breckenridge and Kukolich ApJ 438.  
lineset=’CH3OH’   restfreq=12178.597  /  Breckenridge and Kukolich ApJ 438.  
lineset=’OH13’    restfreq=13441.417  /  
lineset=’NH3’     restfreq=18499.393  /  Pratrap Preprint.  
lineset=’CH3OH’   restfreq=19967.3961 /  Menton preprint.  
lineset=’H2O’     restfreq=22235.08   /  
lineset=’CH3OH’   restfreq=23121.0242 /  Menton preprint.  
lineset=’CH3OH’   restfreq=25124.87   /  
lineset=’SiO425’  restfreq=42519.3    /  
lineset=’SiO428’  restfreq=42820.54   /  
lineset=’SiO431’  restfreq=43122.03   /  
lineset=’SiO862’  restfreq=86243.35   /  
lineset=’SiO868’  restfreq=86846.89   /  
lineset=’CH3OH’   restfreq=44069.43   /  Bachiller preprint (SAO)  
lineset=’CH3OH’   restfreq=97980.97   /  Plambeck preprint (SAO)  
lineset=’SiO1293’ restfreq=129363.26  /  
! some more from VLA OBSERVE  
lineset=’H’        restfreq=1420.405752   /  
lineset=’H2CO4830’ restfreq=4829.656900   /  
lineset=’H2CO145’  restfreq=14488.475000  /  
lineset=’NH3(1,1)’ restfreq=23694.495500  /  
lineset=’NH3(2,2)’ restfreq=23722.633600  /  
lineset=’NH3(3,3)’ restfreq=23870.129600  /  
lineset=’NH3(4,4)’ restfreq=24139.416900  /  
lineset=’NH3(5,5)’ restfreq=24532.988700  /  
endlines /  

The example below shows the SCHED input for a spectral line It is a modified version of the file used for a project by Phil Diamond in Dec 95. The modifications are to adjust for some of the new features of SCHED that were not available at the time the file was used. Note that it is not necessary to repeat the DUR and GAP specification every scan. However some users prefer to show these details and it certainly doesn’t hurt. Also, the bandwidth specification is the same as the setup file so it is not required (it used to be).

! EXAMPLE - spectral line observations.  
! Note added Jan. 17, 2013 - GBT needs pointing which isn’t shown here.  
! Nov. 1, 2012.  Use GBT instead of VLA which can no longer  
! use the legacy system.  Eventually switch to RDBE/DDC and WIDAR.  
! Also switch to 2 bit and removed tpmode.  RCW.  
! ==========================================================  
! =================  Cover Information  ====================  
! ==========================================================  
EXPT = ’BD27 VLBA format 7mm line 1995 DEC 29 18:00 -> DEC 30 18:00 UT’  
EXPCODE  = ’BD027’  
VERSION  = 1  
PINAME   = ’P.J.Diamond’  
ADDRESS2 = ’P.O. Box O’  
ADDRESS3 = ’Socorro, NM 87801, USA’  
PHONE    = ’1-505-835-7365 (work) or 1-505-835-2095 (home)’  
OBSPHONE = ’1-505-835-7365 (work) or 1-505-835-2095 (home)’  
FAX      = ’1-505-835-7027’  
EMAIL    = ’pdiamond@nrao.edu (internet)’  
! ==========================================================  
! ==============  Correlator Information  ==================  
! ==========================================================  
correl   = ’Socorro’  
coravg   = 4  
corchan  = 256  
cornant  = 10  
corpol   = ’off’  
corwtfn  = ’uniform’  
corsrcs  = ’standard’  
cortape  = FTP  
corship1 = ’Athol Kemball’  
corship2 = ’P. O. Box O’  
corship3 = ’Socorro NM 87801’  
! ==========================================================  
! ====================  Source Catalog  ====================  
! ==========================================================  
!  This has sources with positions in mixed equinoxes.  
!  It is normally recommended to use J2000.  
!  Note most of the continuum sources below could be picked up  
!  from $SCHED/catalogs/sources.gsfc  
SOURCE=’3C273’    RA=12:26:33.2480 DEC= 02:19:43.290 EQUINOX=’B1950’ /  
SOURCE=’3C279’    RA=12:53:35.8380 DEC=-05:31:08.040 EQUINOX=’B1950’ /  
SOURCE=’3C345’    RA=16:41:17.6080 DEC= 39:54:10.820 EQUINOX=’B1950’ /  
SOURCE=’3C454.3’  RA=22:51:29.52   DEC= 15:52:54.35  EQUINOX=’B1950’ /  
SOURCE=’OJ287’    RA=08:51:57.2530 DEC= 20:17:58.440 EQUINOX=’B1950’ /  
SOURCE=’1823+568’ RA=18:23:14.9490 DEC= 56:49:18.050 EQUINOX=’B1950’ /  
SOURCE=’1334-127’ RA=13:34:59.8150 DEC=-12:42:09.900 EQUINOX=’B1950’ /  
SOURCE=’1633+382’ RA=16:33:30.6280 DEC= 38:14:10.050 EQUINOX=’B1950’ /  
SOURCE=’0420-014’ RA=04:20:43.5400 DEC=-01:27:28.660 EQUINOX=’B1950’ /  
SOURCE=’1749+096’ RA=17:51:32.8185 DEC= 09:39:00.728 EQUINOX=’J2000’ /  
SOURCE=’RAQR’     RA=23:41:14.269  DEC=-15:33:42.89   EQUINOX=’B1950’  
                  VEL = -27.0, -27.0 /  
SOURCE=’RLEO’     RA=09:44:52.2    DEC= 11:39:40.8   EQUINOX=’B1950’  
                  VEL = -2.0,  -2.0 /  
SOURCE=’VYCMA’    RA=07:20:54.6    DEC=-25:40:12.2   EQUINOX=’B1950’  
                  VEL = 20.0 /  
SOURCE=’VXSGR’    RA=18:05:02.9    DEC=-22:13:55.6   EQUINOX=’B1950’  
                  VEL = 8.0 /  
SOURCE=’UHER’     RA=16:23:35.0    DEC= 19:00:18.0   EQUINOX=’B1950’  
                  VEL = -15.0, -15.0 /  
SOURCE=’IKTAU’    RA=03:50:43.7    DEC= 11:15:31.8   EQUINOX=’B1950’  
                  VEL = 34.0, 34.0  /  
SOURCE=’TXCAM’    RA=04:56:41.4    DEC= 56:06:29.9   EQUINOX=’B1950’  
                  VEL = 9.0, 9.0 /  
SOURCE=’NMLCYG’   RA=20:46:25.59   DEC= 40:06:58.3   EQUINOX=’J2000’  
                  VEL = -5.0, -5.0 /  
SOURCE=’SPER’     RA=02:19:15.1    DEC=+58:21:34.0   EQUINOX=’B1950’  
                  VEL = -40.0, -40.0 /  
! ==========================================================  
! ====================  Station Catalog  ===================  
! ==========================================================  
stafile  = ’$SCHED/catalogs/stations.dat’  
freqfile = ’$SCHED/catalogs/freq.dat’  
! ==========================================================  
! ==============  Spectral line rest frequecies  ===========  
! ==========================================================  
lineset=’ccal’  restfreq=43122.027, 43122.027, 43126.027, 43126.027,  
                         43130.027, 43130.027, 43134.027, 43134.027 /  
lineset=’prog’  restfreq=43122.027, 43122.027 /  
! ==========================================================  
! ====================  Observing setup  ===================  
! ==========================================================  
!  This is a fairly fully specified setup file.  
!  Note the VLBA and GBT setsups are the same in this case.  
setinit = bd27.set  /  
nchan    = 8  samprate = 8.0  bits = 2  bbfilter = 4.0  !  128 Mbps  
format = VLBA1:1  
bbc      = 1, 2, 3, 4, 5, 6, 7, 8  
netside  = U, U, U, U, U, U, U, U  
ifchan   = R, L, R, L, R, L, R, L  
!    Radio Astronomy allocation:  42400-43500  
pcal     = ’off’  
freqref  = 43150.99  
freqoff  = -8,-8,-4,-4,0,0,4,4  
firstlo  = 42400.00  fe(1) = ’7mm’   fe(3) = ’7mm’  
synth(2) = 7.6  synth(3)= 11.6   ! LO = Syn(2) + 3*Syn(3)  
station  = ’VLBA’, ’GBT_VLBA’   rchan = A  lchan = C   /  
endset /  
!  As a demonstration, the following is all that is needed  
!  to get an equivalent setup to the above.  It is  
!  the same.  Note that both turn off the pulse calibration tones  
!  which can mess up spectral line observations.  
!  Neither setup uses the default ’format’ because that would give a  
!  speedup factor on correlation and the correlator output data  
!  rate would be too high.  Note that the format can be forced  
!  with either ’tpmode’ or ’format’.  
setinit = bd27a.set  /  
nchan    = 8  bits = 2  bbfilter = 4.0  
pol      = dual  
pcal     = ’off’  
band     = ’7mm’  
endset /  
! ==========================================================  
! =================  Initial Scan Information  =============  
! ==========================================================  
YEAR   = 1995  
MONTH  = 12  
DAY    = 29  
START  = 18:02:00  
! ==========================================================  
! ========================  The Scans  =====================  
! ==========================================================  
SETUP  = ’bd27a.set’  
SOURCE = ’1749+096’  GAP = 00:02:00 DUR = 00:11:00 DOPSRC ’SPER’/  
SOURCE = ’SPER’      GAP = 00:02:00 DUR = 00:11:00 /  
SOURCE = ’SPER’      GAP = 00:02:00 DUR = 00:11:00 /  
SOURCE = ’SPER’      GAP = 00:02:00 DUR = 00:11:00 /  
SETUP  = ’bd27.set’   !  Try the other one.  
SOURCE = ’3C454.3’   GAP = 00:02:00 DUR = 00:11:00 DOPSRC ’SPER’/  
SOURCE = ’SPER’      GAP = 00:02:00 DUR = 00:11:00 /  
SOURCE = ’SPER’      GAP = 00:02:00 DUR = 00:11:00 /  
SOURCE = ’SPER’      GAP = 00:02:00 DUR = 00:11:00 /  
SOURCE = ’SPER’      GAP = 00:02:00 DUR = 00:11:00 /  
SOURCE = ’3C454.3’   GAP = 00:02:00 DUR = 00:11:00 DOPSRC ’SPER’/  
SOURCE = ’SPER’      GAP = 00:02:00 DUR = 00:11:00 /  
SOURCE = ’SPER’      GAP = 00:02:00 DUR = 00:11:00 /  
SOURCE = ’SPER’      GAP = 00:02:00 DUR = 00:11:00 /  
!     __________________________________  
!    |                                  |  
!    |   Many scans in the same style.  |  
!    |__________________________________|  
SOURCE = ’1749+096’   GAP = 00:02:00 DUR = 00:11:00 DOPSRC ’UHER’/  
SOURCE = ’UHER’       GAP = 00:02:00 DUR = 00:11:00 /  
SOURCE = ’UHER’       GAP = 00:02:00 DUR = 00:11:00 /  
SOURCE = ’UHER’       GAP = 00:02:00 DUR = 00:11:00 /  
SOURCE = ’UHER’       GAP = 00:02:00 DUR = 00:11:00 /  
SOURCE = ’1749+096’   GAP = 00:02:00 DUR = 00:11:00 DOPSRC ’UHER’/  
SOURCE = ’UHER’       GAP = 00:02:00 DUR = 00:11:00 /  
SOURCE = ’UHER’       GAP = 00:02:00 DUR = 00:11:00 /  
SOURCE = ’UHER’       GAP = 00:02:00 DUR = 00:11:00 /  
SOURCE = ’1749+096’   GAP = 00:02:00 DUR = 00:11:00  DOPSRC ’UHER’/  
! ---------------------------------------------------------------------