PEAK, if specified, will cause commands to be issued to tell some antennas to peak up their pointing. NOPEAK causes SCHED to stop issuing those commands.
As of March 1999, PEAK is no longer used to control VLA pointing. Please see parameter VLAPEAK.
Warning, NOPEAK takes precedence over PEAK if given in the same scan. SCHED has no way to tell that the PEAK is new rather than left over from the previous scan. Often NOPEAK is given right after the peaking scan inputs to ensure that peaking doesn't get left on in the next scan. But if that next scan is meant to be a peaking scan, say on a different set of antennas, you will not get what you want. We have received user files like this.
For the VLBA, if PEAK is greater than 0, a peakup will be done using the channel number specified by the PEAK argument. The peakup begins either when the scan starts, or when the antenna reaches the source, whichever is later. But the on-line system will wait for a maximum of 30 or 40 seconds to reach source before giving up. If it will take longer than that to reach source, a dummy scan or a gap between scans should be inserted. The peakup routine reads the total power after being on each position for 2 seconds and then goes on to the next position. The pattern contains 10 points. Therefore the peakup will take 20 seconds plus slew times which may be as little as 30 seconds - call it 45 seconds to be conservative - at high frequencies. It could take significantly longer at low frequencies but there is no good reason to use reference pointing there. After the peakup is done, the results will be used until either another peakup is done or the project code changes. Hopefully, we'll have some way to turn it off eventually.
It will be common to peak up at a different frequency from that being used for observing (eg peak at 7mm for 3mm observations). It will also be common to peak on a line source during a continuum observation. Therefore a different setup file will be needed for the peaking scans. It is reasonable to use the standard pointing setup files. They have names like pt7mm.set. They use FORMAT=NONE which causes the on-line system on the VLBA to not touch the formatter. Note that this is very likely to mean that any pulse cal data gathered during such scans are likely to be spurious.
For stations other than the VLA and VLBA, adding a PEAK specification to a scan will help trigger reference pointing observations. This is common for the GBT and Effelsberg.
See the section on reference pointing for much more information on easy ways to insert pointing scans.
This command triggers the addition of REFERENCE_POINTING INTENTS to the VEX file. When similar intents are added using VLAPEAK, they are prepended by VLA:. When they are added because of the use of PEAK, they will not have a station identifier. The scans before the first scan in which reference pointing is requested with PEAK will have "INTENT = REFERENCE_POINTING_OFF". All scans for which PEAK is greater than zero will have "INTENT = REFERENCE_POINTING_DETERMINE". After the first determination, any scans with PEAK less than or equal to 0 will have "INTENT = REFERENCE_POINTING_APPLY". The PEAK scheme needs to be enhanced if we are going to treat different stations differently.