Weather Constraints

The dynamic scheduler uses the current API and wind speed, as well as the predicted wind speed when deciding whether to run an observation, i.e. a SB. If the API, current or future wind speeds are above the constraints set in the SB using the OPT then the SB will not be selected. This can make it difficult to observe high frequency SBs.

The wind speed limit is based on the fact that wind will push the VLA dishes around causing the primary beam to move on the sky. At higher frequencies the primary beam is smaller and so it is easier to move a dish enough to move the target out of the center of the primary beam, therefore the wind speed limit is lower at higher frequencies. Wind speeds higher than the limits will cause the amplitudes to vary which cannot be calibrated after the observation.

However, the recommended API limits are based on estimates of the maximum API limit assuming that the science target could not or should not be self-calibrated. If the science target can be self-calibrated then the API limit can be much higher giving the SB a greater chance to be observed. Starting in December 2022 there are two possible Atmospheric Phase Limits (APLs) to choose from depending on whether or not the science target can be self-calibrated.

The following table lists the default weather constraints available in the Observation Preparation Tool (OPT) for each observing band. An observer may choose between the conservative APL or the less conservative self-calibration APL. The wind limit will remain the same between either APL choice.

How to choose the Atmospheric Phase Limit for your SB

Choose the lower (original) limit if:

• The science target is too faint to self-cal (see below), or
• The flux density of the science target is unknown and likely faint, or
• There is a scientific reason not to self-calibrate, e.g., when the scientific goal is to perform astrometry

Choose the higher limit if:

• It is desirable to self-calibrate the science target, and
• The signal to noise ratio (SNR) of the science target is > 3 in a solution interval using the subband width (for continuum) or the anticipated line/channel width (for spectral line) for a single baseline and polarization.
• a solution interval (solint) should be based on how fast the atmospheric phases are likely to be changing and therefore shorter for higher frequencies and longer baselines.
• for a 25 antenna VLA the SNR of the target peak flux density/rms_solint should be ≥20

Examples:

• Example 1: The science target is a compact source with a peak flux density of 15mJy/beam. Observing occurs  in A-configuration at Ka band continuum. Should we use the higher APL?
• Estimate the solint. The recommended calibration cycle time for A-configuration and Ka band is 3 minutes so we assume that the atmospheric phases will vary faster than that with the higher APL (which allows for observations at less favorable weather during the observation). So let's pick a solint of 30 seconds, which would allow ~6 solutions over 3 minutes to track the variations.
• Determine the rms for 30 seconds and a bandwidth of 128MHz, a single polarization and 25 antennas in Ka band using the VLA Exposure Calculator.
                       rms_solint=0.48mJy/beam
• Calculate the SNR: SNR=15/0.48=31 (peak/rms, all in mJy/beam)
                  As this ratio is over 20, this source is good to self-calibrate and therefore one can use the higher APL
• Example 2: The science target is a maser line in K band with a line width of ~3km/s (~220kHz). The expected peak flux of the maser is 147mJy over 3km/s or in a ~220 kHz channel and the observation occurs in C-configuration.
• Estimate the solint. The recommended calibration cycle time for C-configuration and K band is 6 minutes so we assume that the atmospheric phases will vary faster than that with the higher APL. So let's pick a solint of 1 minute, which would allow ~6 solutions over 6 minutes to track the variations.

• Determine the rms for 1 minute and a bandwidth of 3km/s or ~220 kHz, a single polarization and 25 antennas in K band using the VLA Exposure Calculator.
                       rms_solint=10.5mJy/beam
• Calculate the SNR: SNR=147/10.5=14 (peak/rms, all in mJy/beam)
                  This ratio is not near 20 so this maser source is probably not a good source to self-calibrate; use the lower, more restrictive APL.

Caveats:

• The theoretical rms_solint might be lower than the actual image rms for many reasons (e.g. RFI, overhead unaccounted for) and therefore self-calibration could be difficult. The SNR ≥ 20 is a slight overestimate of the SNR needed because of the likelihood that the rms will be slightly higher than that given by the VLA Exposure Calculator.
• If the target is dominated by extended emission, the flux detected by longer baselines may be much lower than the peak. It then might be difficult to get good self-calibration solutions on the longer baselines, although uv-restrictions in the self-calibration procedure may help.
• The recommendations and examples above are conservative. There are tricks that can be used to self-calibrate in non-ideal situations. If you are an experienced user and you know you can self-calibrate your science target then please use the higher APL; in that case doing these calculations can be skipped.