Observing Strategy

If you have time scheduled for P band observations, you should obtain the latest version of OBSERVE or JOBSERVE (a GUI-based Java-version of the OBSERVE program). When submitting your observe file to the VLA operators (observe@nrao.edu, email as body of the text) it is recommended to send a copy to the Data Analysts (analysts@nrao.edu) who will look at the file and, if necessary, reply with useful advice. Consider the following guidelines before submitting your schedule.

General Considerations

The prime focus of the VLA antennas is located behind the subreflector even when it is moved as far from the primary reflector as the feed legs allow. Therefore, the P band dipoles are out of focus (by about 40 cm) and the primary beam is somewhat broadened (FWHP ~ 2.5 degrees) and does not have clear nulls. Strong sources can be detected through the sidelobes and aliased into the wanted image unless they are properly represented (using "flanking fields" in the imaging stage, see the discussion on P band data reduction for details).

The sensitivity at P band depends on the location of the target sources because the system temperature is a strong function of the galactic background, which results in a significant decrease in sensitivity toward the galactic plane compared to observations at high galactic elevations (the rms noise can increase by a factor of ~2 due to the higher system temperature and confusion from sources and their sidelobes).

P band observations are best done at nighttime when RFI is minimized. Daytime observations should be successful for fields far from the Sun. At times of significant solar activity targets should be more than 40 degrees from the Sun, although this limit can be relaxed somewhat during quiescent times.

Spectral-line Observations

P band observations should be performed in spectral mode because of RFI and the response of the P band primary beam which detects sources located far from the nominal pointing position. These can contribute significant amounts of flux density and can only be deconvolved properly with spectral observations. RFI at P band is generally narrow-band and both internally and externally generated. Spectral observations allow discarding the contaminated channels, if necessary. The default spectral-line setup for P band uses two IF pairs with a bandwidth of 3.125 MHz each (PP default in JOBSERVE) which is recommended for D and C array observations, although observations with a bandwidth of 6.25 MHz (P1 default in JOBSERVE) are usually successful in the more extended (A, B) configurations. The choice of bandwidth depends on the scientific requirements of the observing program. Because of the large number of sources that are detected in any P band observation (contributing at least ~15 Jy to any observation) it is useful to spread the uv coverage by scheduling observations of each target at several hour angles, if at all possible.

For details on spectral line observations consult the Spectral Line Users Guide.

Choice of Integration time

Time-averaging of visibilities results in amplitude loss for sources not located at the phase center. The magnitude of the amplitude loss depends on the distance of the source from the phase center and the configuration of the VLA. The following Table 9 gives the maximum integration time for a given configuration that prevents the amplitude loss for a point source at the first null of the primary beam from exceeding the value in the table. Deep imaging of the desired field of view requires imaging sources observed through the sidelobes. Under such circumstances the averaging time needs to be correspondingly shorter. The recommended integration time in the A configuration is 5 seconds (but see the notes on Minimum and Maximum available integration times). A 10-second integration time can be used in B, C and D configuration (longer times do not allow the most efficient removal of RFI). Time-averaging loss results in smearing of sources located far from the phase center. At the moment, there is no standard algorithm to correct this effect and estimate the true size of sources. It is best to minimize the averaging time such that the amplitude loss and smearing are insignificant.

Choice of Hour Angle range

The system temperature of the VLA antennas at P band does not increase much with decreasing elevation of the antennas. For example, a 7% increase has been measured at a frequency of 323 MHz when changing the elevation from 90 degrees to 40 degrees (see figure). However, observations at low elevation result in longer paths throught the Earth's ionosphere than those at higher elevation. A balance between good uv-coverage and minimization of phase noise and increased system temperature due to the ionosphere can be obtained by restricting the observations to within 5 hours of transit and elevations above 25 degrees, if at all practical given the declination of the target source.

Calibration

It is best to observe one of the primary flux calibrators (3C48, 3C147, 3C286, 3C295) at least once per observing run. The secondary calibrators, which are used to estimate the overall antenna gains, are observed more often, about once every two hours. Source confusion often requires making images of the secondary calibrators (and optimizing them through self-calibration) before determining calibrations constants that can be applied to the target fields. A bandpass calibrator (3C48, 3C147, 3C196, 3C286, 3C295 or 3C380) should be observed at least twice in order to obtain good calibration of the relative gains of the spectral channels. Often, an attractive option is to use one of the bandpass calibrators as a secondary calibrator as well (even if it is not located within 10 degrees of the target fields) because the large flux density that that is observed in any field allows correction of the phases through self-calibration.

Sample Observe File

A typical observe file in multi-channel continuum mode for wide bandwidth P-band observations is given here. Observations are taken in "line" mode. The observations are recorded in 4IF mode (//DS 4) each of which has a bandwidth of 6.25 MHz (code 3333). The data have on-line Hanning smoothing applied (code H) and have an integration time of 10 seconds). The resulting data set has two pairs of 15 channels (each with a width of 390.6 kHz) with the first set centered at 323.125 MHz (RR and LL correlations) and the second set centered at 328.125 MHz (again RR and LL correlations). The observations consist of two target sources, with 3C48 used as primary calibrator, 3C196 used as bandpass and phase calibrator for the first target field (BFIELD11) and 3C295 used as the secondary and bandpass calibrator for the second target (BFIELD7). Finally, a seconf primary calibrator, 3C286, is observed at the end. Notice that the observations of both targets are interspersed (where allowed by a minimum elevation constraint of 25 degrees) for better uv-coverage. This setup does not allow measuring linear polarization as the cross-products (RL and LR) are not determined. If such data are wanted, replace mode "4 " with mode "PA" or "PB" which will observe RR, LR, RL and LL, albeit for only one frequency setting.

Copyright © 2009 Associated Universities, Inc.
The National Radio Astronomy Observatory is a facility of the National Science Foundation operated under cooperative agreement by Associated Universities, Inc.