Examples of deliverable LOFAR products


This page describes example characteristics of typical LOFAR observations in its early operational phase.  

The performance described here should be seen as the ACTUAL capabilities of LOFAR for 2012. 


Interferometry: Calibration and Imaging Performance

A) Interferometric Observation with stations in the NL

(20 January 2012)

The LOFAR synthesis calibration and imaging pipeline is constantly being refined and improved. It is  therefore expected that the quality and nature of the products will steadily improve during the course of 2012. This text will be updated regularly to reflect the products that can be delivered. Please consult this page before submission of a proposal.


The most important aspects to consider when requesting LOFAR synthesis imaging observations are the array configuration, the field-of-view and the angular resolution, all of which are related.  Two types of products can be requested: calibrated images/imagecubes and calibrated MS. In addition, the sky model that was used to calibrate the image will be provided.  The visibility data (Measurement Sets) will consist of a series of up to 244 subband MS's with a specifiable integration time and frequency resolution. Each of these parameters will be briefly discussed below.  For more information we refer to other sections on these web pages. 

The specification of many parameters can be done through the MoM interface during the preparation of the observations with defaults offered. Here is brief overview of the most important parameters.

Array configuration:

Throughout the first half of 2012 LOFAR will consist of the following array: 24 core stations (CS, 24x2 HBA sub-stations or 24 LBA stations) and 9 remote stations (RS). In addition, 8 international stations (IS) will be available. However, since the input to the BG/P is limited to 64 stations, if International stations are requested up to 8 HBA sub-stations can not be recorded.

Frequency coverage

Up to 244 subbands of each 0.195 MHz (assuming 200 MHz sampling) can be provided.

The default reccomended frequency range will cover contiguous bands from 30-78 MHz in the LBA and 120-168 MHz in the HBA. However, in the HBA (low and mid) band it is also possible to use a wider, and therefore non-contiguous frequency band (e.g. 115-185 MHz).

Still higher frequency, and therefore higher angular resolution, can be obtained in the HBA-high band; the suggested frequency range in the HBA-high band is 215-245 MHz.

It should be realized, however, that the sensitivity in the LBA and HBA bands varies significantly with frequency (see the section on sensitivity).


Integration time and frequency resolution

The field of view of LOFAR is a strong function of various observing and processing parameters, as well as the state of software to deal with wide field calibration, imaging and ionospheric correction schemes.  The default integration time and frequency resolution will probably be 1s (LBA) and 3s (HBA) and 3.1 kHz (both bands). For experiments involving European stations these parameters will have to be specified by the user because they depend on the applications.

In most, if not all, LBA observations it will be crucial to remove the effects of CasA and CygA. We will use the time-consuming demixing scheme. Following this the data will be averaged to (default) 10x the original intervals, decreasing the datavolume by a factor 100.

If calibrated visibilities (MS) are requested the default averaging time will be 10s and the default frequency resolution will be 12 kHz (=1/16th  of an original subband)

Number of digital beams

LOFAR will allow the user to exchange bandwidth for sky coverage (beams).  Currently their product is measured in units of MHz-beams.

A limited choice will be provided initially: varying from 1 beam of 244 subbands (48 MHz) to 8 beams of ~31 subbands (6 MHz).  The location of these beams on the sky can be specified by the user, taking due consideration of the size of the dipole and tile beams.  The subband coverage can be specified within the ranges of the passbands set by the sampling mode and filters (see section...) . Different beams can have the same or non-overlapping subbands which need not be contiguous.

Image(cube) parameters

A limited range of parameters can be provided for the FOV, pixel size and frequency resolution of the output image cube(s).  These obviously will depend on the requested array configuration and frequency range.

Expected image noise levels

The noise levels indicated below are based on the current performance of the pipeline and are still not close to the theoretical noise.

Furthermore static cable-delay station calibration is continously improving and it is expected that the quoted sentitivities  will improve steadily in 2012. 

For a single subband the following noise figures have been achieved for a 6 hour observation typically using in the LBA 17 Core and 8 Remote stations and in the HBA 38 Core (sub-)stations and 7 remote stations in the first hald of 2011:

LBA:   30-78 MHz, 1x6h :   80 mJy/psf
HBA:  120-168 MHz, 1x6h :   6 mJy/psf

With the improvement in the station calibration and a larger number of stations now available (24 core stations and 9 remote stations), an improvement in the noise by a factor of 1.5  can be expected.

Averaging a large number of subbands will obvious lower the achievable noise level. However, the noise levels are not expected to go down as rapidly as SQRT(#subbands). In due time more typical numbers will be provided.

Brief description of the Imaging  pipeline:

After the observations have finished the imaging pipeline (see "Software Pipelines") will be started. This involves the following standard steps: RFI-flagging, demixing (if required), data averaging, BBS-calibration and imaging. All of these processed can, in principle,  be steered via a set  of user specifiable parameters. However they will have to comply with what the RO considers reasonable parameters, which are set by the calibratability and available processing resources.

We expect that the various parameters will depend on location of the target (in/out Galactic plane), angular distance from CasA and CygA, day/night time etc. A lot of ptractical experience will be gathered during the spring of 2012 when we process the data taken for the MSSS.

The Local Sky Model to calibrate the data will probably be extracted from the Global Sky Model provided by MSSS which we expect to be available at the time the first observations get scheduled.
However, it will always be possible for the uses to supply their own sky model.

Initially the calibration and imaging pipeline will only contain one major cycle: i.e there will not be an update of the LSM using the computed image. However, a rudimentary pipeline with more than one major cycle might be offered after the Summer of 2012.

B) Imaging with international stations

Baselines provided by the international array range from 53km (DE601-DE605) to 1292km (SE607-FR606). The corresponding resolution extends far into the sub-arcsec range for the high band and reaches 1 arcsec even in the low band (see table 2 at )

The international stations (IS) devote 90% observing time to ILT observations so that they can be included for most observations. Data connections to Groningen allow the use of all 8 IS at the same time.

It has to be understood, however, that the total number of input data streams is limited to 64, which means that some Dutch stations have to be excluded when HBA_DUAL is used in combination with all international stations.

At the moment, data loss at some stations is quite common even when using less than 64 stations in total. This is being investigated and will be solved as soon as possible.

Even in the high band, ionospheric effects can be quite severe. At night-time and under quiet ionospheric conditions, (differential) ionospheric delays are generally below 0.5 microsec at frequencies around 150 MHz, but these values can easily increase by an order of magnitude at low elevation or at day-time. 5 microsec correspond to about one full phase turn over a subband, so that data have the be kept channelised before calibration. Keeping 16 channels generally seems to be safe for the high band. Phase rates can be below 10 mHz (again around 150 MHz) under good conditions, but increase to 100 mHz during day-time. This implies that time-averaging beyond one or at most a few seconds generally has to be avoided.

Due to both effects, the data volume before calibration has to be considerably larger than on short baselines. Ionospheric effects are even much stronger in the low band.

Besides the total ionospheric delays, long baselines are also strongly affected by differential Faraday rotation (DFR). On short baselines it is often possible to avoid full polarisation calibration and calibrate and image the XX and YY correlations basically independently. Because ionospheric Faraday rotation can be very different at distant stations, this does generally not work well on international baselines. Under optimal conditions, DFR can be small enough to be neglected in the high band, but often (and almost always in the low band) the effect is so strong that it introduces closure errors of the order one. In this general case, DFR has to be solved for explicitly (which will eventually be possible with BBS), or the procedure described below has to be applied to avoid the problem.

The point-source sensitivity of the international array is higher than that of the Dutch array alone. Because of the higher number of antennas/elements of IS, even the baseline-sensitivity is superior for international baselines. Nevertheless, sensitivity is one of the limiting factors in international observations. The reason is that almost all known sources are more or less resolved on international baselines. Due to interstellar scattering, this may even be true for intrinsically compact sources. As a result, fluxes on international baselines are typically reduced by one or even several orders of magnitude compared to shorter baselines. With the exception of very bright sources, the flux per subband and integration time is not sufficient for calibration. Because of this, direct phase solutions are generally not sufficient, but at least linear trends in time (rates) and frequency (delays) have to be used. Contributions from clock offsets (non-dispersive) and the ionosphere (dispersive) generally both have to be taken into account.

BBS will eventually be able to solve for these parameters. For the time being, fringe-fitting with other software packages is the best option for all but the brightest sources.

Until fringe-fitting and DFR-correction are implemented into BBS and are fully operational, the following procedure is used to image long-baseline observations. Note that not all the required software is publicly available at the moment. This will change in the near future.


Transient Buffer Boards


TBB observations work currently with external trigger provided by the particle detector LORA. Plans for the future is to continuously trigger independently on the pulse in the data only. The later improves with the signal to noise ration, meaning a rather low noise level is needed. Transient RFI conditions are such that -until the self-triggering algorithm is improved to suppress these; most self-triggers are due to local RFI sources.

Figure XX: Raw voltages across time.

Figure XX  shows the raw voltages of all antennas of one polarisation of 1 station, as recorded in the TBB, when a cosmic ray (7*10^17 eV) was detected. The trigger to dump these data came from the LORA particle detector.

The data is (phase) shifted in time to correct for delays between the signals of the different dipoles. 


Figure XX+1 shows the different arrival times for the same antennas.


The typical data size is 190 MB per station (for a core station) and two polarizations
for 5 ms of read-out. The data size depends on the number of antennae used.

The data rate is 400 MB/s per antenna

The maximum amount of data than can be dumped accounts for 1.28 seconds (with the current default time resolution of 5 ms for cosmic ray observations). It takes 48 minutes to write the full 1.28 s of  TBB on the storage nodes (With forthcoming improvements this may be reduced to 8 minutes)

The data are written as raw voltages along with some metadata in hdf5 format.

The processing pipeline is not available at CEP. There is a semi-automated CR pipeline in Nijmegen (<1 minute per event per station). For processing  VHECR events using this pipeline one should contact the CRKSP group. The software package to access the data and do some processing (eg. FFT, RFI mitigation, ..) with python scripts, PyCRTools, is available at CEP.


Design: Kuenst.    Development: Dripl.    © 2016 ASTRON