Description of the EVN Pipeline Feedback Pages

The following is a description of the plots and information that are available on the pipeline feedback page. The pipeline feedback for EVN user experiments can be obtained from the EVN Data Archive (choose experiment, select 'pipeline' and then select 'pipeline plots'). Further details are available from the pipeline homepage.

Please note that some plots may not be available for all experiments. Also, some of the plots will not be available for particular sources (e.g. dirty maps are not useful for bright sources and so are not shown, weak sources are not self-calibrated, etc.). In general, the plots of (u,v) data show all baselines to a single station only, and each AIPS IF is plotted separately.

General Comments. (Brief data summary and scan listing)
The JIVE support scientist responsible for pipelining the experiment will note here general particulars of the experiment. Also given are a brief summary of the data (scan listing, list of participating telescopes with their AIPS antenna table numbers, the number of visibilities on each baseline, etc.). Also given is the EVN Reliability Indicator, which gives an estimate of the ratio of observed (good) visibilities to the number of visibilities expected at scheduling time. ERI* is also given, and is the ERI modified to account for data lost due to the weather (or other 'acts of God'). Thus ERI* >= ERI.

Plots of the autocorrelations.
The autocorrelation against frequency. These plots use 22 minute or scan averages (whichever is smaller) and cover the whole experiment. They are useful for identifying RFI.

Plots of the uncalibrated amplitude and phase against time.

This shows the raw data averaged across each AIPS IF and plotted as a function of time for the whole experiment. Each AIPS IF is plotted separately. This can be used, for example, to determine times when stations were not producing data or to identify IFs with problems. Note that flags from telescope monitoring data (that flag data taken when antennas are off source) will have been applied. As the data have not yet been fringe-fitted some decorrelation is likely due to averaging across the IFs.

Plots of the uncalibrated amplitude and phase against frequency channel
Each plot shows a scan average. All scans are plotted (but only baselines to the reference antenna). Each channel is plotted so bandpass shapes and the sensitivity of different telescopes and IFs can be compared.

The uncalibrated amplitude and phase of the crosshand correlations against frequency channel.
If the cross polarizations (RL and LR) were correlated these are plotted here for information (no polarization calibration is done by the pipeline).

TSYS against time
A priori Tsys values as a function of time for all AIPS IFs.

Telescope sensitivities from the a priori TSYS and Gain curves
This is the result of combining the Tsys with the station gain curves. The units are square root of the SEFD (nominal values of the SEFD of EVN telescopes can be found on the EVN status table).

Fringe-fit phase solutions
The phase solutions produced by fringe-fitting the data for all telescopes and all AIPS IFs.

Fringe-fit delay solutions

Fringe-fit rate solutions

Telescope bandpasses
Plots of the bandpass (amplitude and phase) determined for each telescope using the AIPS task BPASS.

Calibrated amplitude and phase against time
This shows the data averaged across each AIPS IF and plotted as a function of time for the whole experiment. Each AIPS IF is plotted separately. A priori amplitude calibration, fringe-fitting results and bandpass corrections have been applied. Data at times when antennas were known to be off source (from telescope monitoring data) will have been flagged. Flags may also be applied to remove data with low SNR at the edges of the AIPS IFs.

Calibrated amplitude and phase against frequency channel
Each plot shows a scan average with calibration and flagging applied (see description of previous plot). All scans are plotted (but only baselines to the reference antenna).

Naturally weighted dirty map produced before self-cal
These are the dirty maps of all the pipelined sources, using natural weighting. These maps may be useful to see whether faint sources have been detected, but are unlikely to be useful for bright sources (see, instead, the clean maps below).

Uniformly weighted dirty map produced before self-cal
As above, but produced using uniform weighting (data are weighted according to the local density of u,v points).

Phase corrections applied to a priori calibrated and fringe-fitted data by self-calibration
These are the solutions produced by the first iteration of phase-only self-calibration of each source.

Amplitude corrections applied to a priori calibrated and fringe-fitted data by self-calibration
These are the amplitude corrections derived from amplitude-and-phase self-calibration of the sources (done after an initial phase-only self-calibration). Corrections should be small if the a priori amplitude calibration was good. Please note that the error given here is the square root of the error in the original Tsys value.

Telescope sensitivities (the total AMP gain applied during both a priori and self calibration)
This is similar to the telescope sensitivities plot given previously, but modified to account for the amplitude self-cal solutions given in the previous plot. Again these numbers can be compared to the values in the EVN status table to indicate the antenna performance.

Residual closure phase
This is the residual closure phase after a model of each source has been subtracted from the (u,v) data. This value should be close to zero, assuming the pipeline maps of the sources are reasonable.

Calibrated visibilities and the source model
This is similar to the calibrated visibilities against time plots given above, except that all calibration (including self-calibration where relevant) has been applied, and each source is plotted separately along with a line indicating the best model of the source produced by the pipeline. Discrepancies between the data and the model can indicate errors.

Calibrated visibilities against u,v distance
The calibrated (including self-calibration) visibility amplitude is plotted as a function of u,v distance for each source.

u,v coverage
This is a conventional u,v coverage plot for each source, using only the data which remains after the data have been edited and self-calibrated by the pipeline.

Crude map
This is the final, clean, map that was produced by the pipeline procedure. Due to the incomplete nature of the pipeline processing (few iterations of self-calibration), and in particular the lack of interactive data editting, spurious features are likely in these maps.