World map small

EVN scheduling

If the successful proposal is scheduled to be observed in the next session, the contact author of the proposal will receive an informative email about how to prepare the observations. This email will require the contact author to provide a schedule using the JIVE pySCHED software (see SCHED/PySCHED software section). If help in the schedule preparation is needed, please contact the EVN Support Scientists at JIVE (usersupport (at)

If stations other than those of the EVN, Arecibo, NRAO, DSN and VSOP (e.g. Ny-Alesund, Matera, Fortaleza) are scheduled in the observations, please consult the code of practice for non-EVN observatories.

Preparation of a schedule

Please use the following template as reference to prepare the schedule.


The following steps will explain the parameters in the different sections of the file that need to be adjusted for your project (highlighted as {PARAMETER}).


SCHED/pySCHED software
EVN schedules can be prepared by modifying the previous template (the following sections detail how to modify it to your own observation). Only this file is required prior de observations. However, processing of the schedule file can be done with these two programs to verify the observations and confirm that they are prepared correctly.

The schedule template (.key) file can be compiled by using the pySCHED software.
pySCHED is a Python wrapper of SCHED developed by JIVE that can be directly installed with e.g. pip install pythonSCHED). For further details, check the GitHub repository of pySCHED. As long as pySCHED has internet, it will retrieve the latest frequency setup for your experiment..
Cover Information
expcode  = '{EXPERIMENT_CODE}'
version  = 1
expt     = '{EXPERIMENT CODE}'
piname   = '{PI NAME}'
address1 = '{PI INSTITUTE NAME}'
address2 = '{PI INSTITUTE STREET}'
address3 = '{PI INSTITUTE CITY}'
email    = '{PI EMAIL}'
phone    = '{PI PHONE}'
obsphone = '{PI PHONE}'

Here please enter the relevant information concerning the experiment to be observed (assigned experiment code), and contact details of the contact author/Principal Investigator.

Correlator Information
obstype  = 'VLBI'
correl   = 'JIVE'
corpol   = 'on'
coravg   = {TIME AVG IN SECONDS}
cornant  = {NUMBER ANTENNAS}
corwtfn  = 'uniform'
corsrcs  = 'standard'
cortape  = 'ftp'

These are the parameters to be used during correlation. Please update the number of antennas that are scheduled for your experiment (CORNANT), and the number of channels per subband (CORCHAN) and the time averaging (CORAVG>). Note that JIVE will confirm the final correlation parameters with the PI before correlation.

Program Control
sumitem = el1, early

In this section you can modify the output information that is written in the .sum file (after running SCHED/PySCHED) per antenna and per observing scan. Often used attributes for SUMITEM are:

  • el1 (el2): Quotes the elevation of each antenna, in degrees, at the beginning (end) of the scan.
  • early: Seconds before the antenna is on source. Negative values mean that the antenna is not on source by the time the scan starts.
  • az1 (az2): Quotes the azimuth of each antenna, in degrees, at the beginning (end) of the scan.
  • slew: Seconds that each antenna requires to move from the previous source to the current source in the scan.
  • dwell: Total duration of the scan (in second) once the antenna is on source.


Antenna names and codes
The antennas that are scheduled for your observation would be quoted in the EVN Block Schedule (you may receive a dedicated email if you are the PI of the scheduled observations). pySCHED does not use the full name of the atennas. Instead, it uses either an abbreviated name of them or the two-letter code that you find in the EVN Block Schedule. In the following we list the correspondence between the full antenna names and the labellings to be used in pySCHED.
Antenna Name pySCHED name pySCHED two-letter code
Arecibo arecibo Ar
Badary badary Bd
Cambridge cambg32m Cm
Effelsberg eflsberg Ef
Hartebeesthoek hart Hh
Irbene irbene or irbene16 Ir or Ib
Jodrell Bank (Lovell Telescope) jodrell1 Jb1
Jodrell Bank (Mark II) jodrell2 Jb2
Kunming kunming Km
Medicina medicina Mc
Metsahovi metsahov Mh
Noto noto Nt
Onsala onsala85 or onsala60 O8 or O6
Sardinia sardinia Sr
Sheshan shanghai Sh
Svetloe svetloe Sv
KVN Tamna kvntn Kt
KVN Ulsan kvnus Ku
KVN Yonsei kvnys Ky
Tianma tianma65 T6
Torun torun Tr
Urumqi urumqi Ur
Westerbork wstrbork Wb
Wettzell wettzell Wz
Yebes yebes40m Ys
Zelenchukskaya zelenchk Zc
Source Catalog
srccat /
    source='{SOURCE NAME}' ra={hh:mm:ss.ssss} dec={dd:mm:ss.sss} vel={VELOCITY} vref={REFERENCE FRAME} vdef={VELOCITY DEFINITION} /
endcat /

List here all sources that you want to observe and are not available in the standard SCHED (calibrator) catalogs. Include the name, coordinates, and, only in case of spectral line experiments, please add the velocity information of the source (vel=, vref=, vdef=):

  • vel: velocity of the source, in km/s, within the vref reference frame.
  • vref: reference frame for vel (L = Local Standard of Rest, H = heliocentric, G = geocentric).
  • vdef: velocity definition (R = radio definition, O = optical definition).


Frequency Setup

setup = {SETINI SETUP NAME}  ! For example:  setup = evn6cm-2Gbps-32MHz.set

Please add the general frequency setup for the observation. You will receive this information from JIVE prior to the observation. The setup written here refers to the setups that are stored in ~/.pysched/setup/.

Note that if you use a non-standard setup or you are using the NRAO SCHED program instead of pySCHED, you will need to put explicitly here the full information concerning your setup.

Start of Observation
DAY   = {DAY}


Specify when the observation starts (year, month number, day, and time of the day in UTC format), and the scheduled stations. All this information is available from the EVN Block Schedule.

The Scans

! Simple continuum observation

source={FRINGE-FINDER} gap=0:00 dur={5:00} /

source={TARGET} gap={3:00} dur={3:00} /
group 1 rep {X}
	source={TARGET} gap={0:30} dur={14:30} /

source={FRINGE-FINDER} gap={3:00} dur={5:00} /

! Phase-referenced continuum observation

source={FRINGE-FINDER} gap=0:00 dur={5:00} /

source={PHASE-CALIBRATOR} gap={3:00} dur={2:00} /
group 4 rep {X}
	source={TARGET} gap={0:00} dur={3:00} /
	source={PHASE-CALIBRATOR} gap={0:00} dur={2:00} /
	source={TARGET} gap={0:00} dur={3:00} /
	source={PHASE-CALIBRATOR} gap={0:30} dur={1:30} /

source={FRINGE-FINDER} gap={3:00} dur={5:00} /

! Phase-referenced  spectral-line observation

source={FRINGE-FINDER} gap={0:00} dur={5:00} dopsrc={TARGET NAME} /

source={PHASE-CALIBRATOR} gap={3:00} dur={2:00} dopsrc={TARGET NAME} /
group 4 rep {X}
	source={TARGET} gap={0:00} dur={3:00} dopsrc={TARGET NAME} /
	source={PHASE-CALIBRATOR} gap={0:00} dur={2:00} dopsrc={TARGET NAME} /
	source={TARGET} gap={0:00} dur={3:00} dopsrc={TARGET NAME} /
	source={PHASE-CALIBRATOR} gap={0:30} dur={1:30} dopsrc={TARGET NAME} /

source={FRINGE-FINDER} gap={3:00} dur={5:00} dopsrc={TARGET NAME} /

! Polarization observation

source={FRINGE-FINDER} gap=0:00 dur={5:00} /

! Every 1 hour
source={POLARIZATION-CALIBRATOR} gap={0:00} dur={5:00} /

source={PHASE-CALIBRATOR} gap={3:00} dur={2:00} /
group 4 rep {X}
	source={TARGET} gap={0:00} dur={3:30} /
	source={PHASE-CALIBRATOR} gap={0:00} dur={2:00} /
	source={TARGET} gap={0:00} dur={3:30} /
	source={PHASE-CALIBRATOR} gap={0:30} dur={2:00} /

source={POLARIZATION-CALIBRATOR} gap={0:00} dur={5:00} /

source={FRINGE-FINDER} gap={3:00} dur={5:00} /

In this section, specify the list of scans to be observed. Commonly, a scan is defined as follows:
source={SOURCE-NAME} gap={0:00} dur={2:00} /,
where dur is the total duration of the scan and gap the time gap between the end of the previous scan and the start of this one. Multiple scans can be compacted in a loop (group N rep X, where N defines the number of lines to be considered within the loop and X the number of times that the loop will be repeated).
The dwell parameter is an alternative way to specify the duration of a scan. The difference to dur is that the start time of the scan is delayed until all antennas are on-source (taking into account the slewing and settle time of each antenna). This implies that telescopes will be idle waiting for the slowest antenna to reach the source position instead of recording once they individually reach the source. For EVN observations this must not be used since the different telescopes of the EVN have significantly different slewing times, and thus a significant loss of on-source time can be encountered. In addition, due to the flexibility of the EVN (specially in the e-EVN mode) some antennas may be added in the last minute, producing that the full schedule would change if dwell is in use. Therefore, please always use the parameter dur instead.

In every observation we recommend to schedule at least two scans on fringe-finder sources (strong and compact calibrator sources). These scans will be used for the clock search at correlation time, and often for bandpass calibration (see Calibration section).

In the following we explain the basic examples of different types of observations that are provided in the template file:

Simple continuum observation

In case the target source is strong enough on all baselines (implying that there are unresolved or barely resolved strong source components), the calibration can be performed directly on the target, and it can be observed in subsequent scans. As explained in the System Temperature Measurements section, gaps of at least 10 seconds with antennas on source must be scheduled every 10-13 minutes to provide accurate system temperature measurements.

NOTE that with this approach, no astrometric information can be recovered from the target.

Phase-referenced continuum observation

When the target source is too faint (less than ~100 mJy on any baseline), it cannot be used to conduct the main calibration of the data. Then, one needs to rely on a nearby strong and unresolved source (phase calibrator; see Calibration section). The phase calibrator must be observed alternately with the target to interpolate the solutions from the calibrator to the target.

Phase-referenced spectral-line observation

For this kind of observations a Doppler correction source must be specified with dopsrc in each scan. For target and phase calibrator this must be the same source (typically the target) to apply the same velocity correction to these sources.

Polarization observation

To properly calibrate the polarization D-terms, a polarization calibrator must be scheduled periodically along the observation (likely every hour; see Calibration section).


System Temperature Measurements
The EVN telescopes regularly measure system temperatures using noise diodes. Some telescopes can measure the temperatures continuously. However, the other stations measure them only when there are long enough time gaps between observing scans. This refers to time gaps that guarantee at least 11 seconds on source for these stations at the beginning of the scan. One of these long gaps must be scheduled every ~10-13 minutes (SCHED/PySCHED will provide warnings for intervals between gaps of more than 15 minutes). You can check the length of time gaps by looking at the .sum output file (see Output section) using the early attribute in sumitem (see Program Control section).
Effelsberg Pointing
Effelsberg is a 100-m antenna that requires periodic pointing scans. To achieve a good pointing precision, Effelsberg has to have an eight-minute period without recording every four hours (for 6, 5, and 3.6 cm; i.e. C, M, and X band), and every two hours (for 1.3 and 0.7 cm; i.e. K and Q band). No pointing scans are required for 21 cm (L band) or longer wavelengths. This eight-minute period can be achieved by either having a gap of eight minutes in the schedule or taking Effelsberg out of the station list for some scans.
Jodrell Bank (Lovell Telescope) slewing limitations
The Lovell Telescope (typically known as Jodrell1, or Jb1, in EVN schedules) has a slewing limitation of 12 source-changes per hour (which is equivalent to a ten-minute cycle time between two sources). If you want to go to a faster cycle time and/or are using a traditional phase-referencing cycle like the one quoted above, it is required to schedule an observing pattern that omits the Lovell Telescope every other reference-source scan.
A convenient way to perform this is illustrated in the following phase-referencing example:

group 4 rep {X}
    stations = jodrell1, all other antennas 
    source={TARGET} gap={0:00} dur={3:00} /
    source={PHASE-CALIBRATOR} gap={0:00} dur={2:00} /
    source={TARGET} gap={0:00} dur={3:00} /
    stations = all other antennas     ! NOTE THE LACK OF jodrell1 
    source={PHASE-CALIBRATOR} gap={0:30} dur={1:30} /

stations = jodrell1, all other antennas ! Continue with the full array
In this way, the Lovell Telescope (Jodrell1) will not observe every other scan on the phase calibrator. Instead it will keep observing the target source.
Output Sum File
SCHED/PySCHED produce different output files. You may want to look at the .sum file to verify the schedule of your observation.
The example below shows the scan summary section included in the output of the created .sum file for a given experiment using sumitem = el1, early.

Output sum file

This is a summary of all scans that will be observed in the observation, with the relevant data (indicated in the sumitem command) for each station (columns). In this example, the first row of each scan shows the elevation of the source at the start of the scan. The second row quotes the number of seconds that a station is on source before the start of the scan. Negative values show for how long the antenna is still moving from the previous source position until it reaches the current one.
For System Temperature measurements, the values from the second row should be at least 10 seconds every 10-13 min for all antennas with non-coninuous Tsys measurements (see System Temperature Measurements section).
Once the schedule is prepared, review the following points by means of this checklist.
  • Coordinates of all sources are correct.
  • Frequency setup is correct.
  • Single polarization observations are avoided.
  • Date, start and end time match the block schedule information for your observation.
  • Stations in the schedule match block schedule information for your observation.
  • At least one (more than one recommended) fringe-finder visible by each station is scheduled.
  • Sufficient time gaps for system temperature measurements are placed in the schedule (see System Temperature Measurements section).
  • Effelsberg has scheduled enough pointing scans.
  • For phase referencing observations: separation between target and phase reference source is small enough (see "Selection of calibrators" section below).
  • For spectral line observations: targets and their calibrators use the same dopsrc source.
  • For spectral line observations: vel, vvdef, vref parameters are correct.
  • For polarization observations: suitable polarization calibrators are scheduled.

Selection of calibrators

Calibration sources are critical in EVN observations. It is recommended to schedule at least two scans on strong, unresolved sources (so-called fringe-finders) to guarantee a good signal-to-noise ratio for all baselines during the calibration. Fringe finders are required for every observation in order to determine the station clock delay offsets and drift rates. In addition, fringe finders can be used for bandpass calibration due to their high signal-to-noise ratios on all baselines. For the observation of faint targets or astrometry studies, phase-referencing calibrators are required. These sources must have well-known coordinates, be compact and exhibit a strong signal on all baselines (above ~0.3 Jy), and should be located close to the target source (within 2-3 degrees). This guarantees an accurate calibration of the phases for all antennas and a reliable transfer of those results to the target source. These observations require the detection of the phase-referencing calibrator sources within the coherence time to be able to obtain solutions that can be transferred to the target source (see VLBA Scientific Memo 24 for more detail). In the following table we quote the recommended cycling times (total observing time on phase-reference source and target source) for the different frequency bands of the EVN. Note that depending on the source elevation and separation to the phase calibrator, shorter or longer cycling times may be preferred.

Frequency band Recommended cycling time (minutes)
L (21/18 cm) 4-6
C/M (6/5 cm) 4-6
X (3.5 cm) 2-4
K & Q (1.3 & 0.7 cm)  2


For polarization observations additional polarization calibrators must be scheduled to correct for the D-terms (see Polarisation observations section below). Some good polarisation calibrators at frequencies up to 15 GHz are 3C84 and OQ208 (sub-structured sources on milliarcsecond scales but unpolarised), and DA193 (weakly polarised with very compact structure).

Public catalogues of calibration sources can be found in different places. For example:



Spectral line observations


The velocity range of individual line sources should be taken into account when choosing the frequency setup. Observations are performed with fixed frequencies, and Doppler corrections are to be determined in the scheduling process. SCHED automatically calculates appropriate observing frequencies if, in the key file, the source velocities and the rest-frame frequencies of the lines are specified (see Spectral line observation setup in SCHED). Note that target lines should be located close to the centre of the frequency band because the band edges have lower sensitivity and can have large phase offsets. The bandwidth must be large enough to cover the full extent of the spectral line. If a line source requires a large velocity coverage, it is possible to overlap two frequency bands to guarantee a large enough bandwidth.

When several line sources are observed with different frequency setups in one observation, AIPS requires bandpass calibrators/fringe-finders to be observed with the same setups, so independently for each frequency setup.

The number of spectral channels should be large enough to avoid artificial effects (e.g. RFI, unexpected spurs, etc.) and to divide the true lines into more than a few spectral channels.


Polarisation observations

There are two aspects to proper calibration of polarisation-sensitive VLBI observations: (1) D-term calibration, the calibration and removal of the instrumental polarisations, or D-terms, for each antenna and (2) position angle calibration, the calibration of the absolute orientation of the polarisation position angles (PPAs).

D-term calibration

The D-terms can be determined from either (1) an unpolarised source or (2) a source with a simple polarisation structure.

(1) If a strong unpolarised source is chosen, a single observation scan of 8-10 minutes on that source is sufficient.
(2) In the case of a polarised source, it is necessary to observe it over a wide range of parallactic angles (exceeding about 90 degrees). Usually, scheduling five or six scans with a duration of several minutes is sufficient.

In some cases, one of the other calibrators scheduled in the observation may be suitable as a polarisation calibrator, as long as the source is known to have a simple polarisation structure or is unpolarised. Some good polarisation calibrators at frequencies up to 15 GHz are 3C84 (resolved source with a complex structure but the emission is unpolarised), OQ208 (slightly resolved source but of unpolarised emission), and DA193 (very compact source with weakly polarised emission).

Position angle calibration

The lack of sources with constant polarisation position angles (PPAs) on VLBI scales makes impossible to determine the PPAs with VLBI standalone observations. Thus it is necessary to obtain simultaneous or contemporaneous polarisation measurements with a connected interferometer of the same compact polarization calibrator. By comparing the orientation of the PPAs for the total VLBI-scale polarisation with their known orientation in the connected interferometer observations, it is possible to determine the necessary rotation to calibrate the VLBI polarisation angles. To achieve this, the EVN observations should contain five or six scans on the polarisation calibrator with durations of several minutes spread over the full observation. Instead of conducting dedicated simultaneous or contemporaneous observations with connected interferometers, it is sometimes possible to retrieve the same PPA information from public databases as the University of Michigan Radio Astronomy Observatory database, the VLA/VLBA Polarisation Calibration Page, and the Master EVLA POLCAL Database. However, in this case, there is no guarantee that observations of the selected PPA calibrator sources are conducted close in time to the EVN observations. It is important to try to have the measurements from connected interferometers as close in time as possible to the EVN observations, since the polarisation of compact sources can vary on timescales of days or weeks.

Note that it may be possible to use observations of a compact source with a simple polarisation structure for both D-term calibration and PPA calibration, allowing to spend less total time on the polarisation calibration. Here, it is necessary to find a source that is both relatively highly polarised and has a simple polarisation structure.



If you have any queries regarding the EVN scheduling, please contact the EVN Support Scientists at JIVE (usersupport (at)