In the LCO scheduling system, the airmass of an observation is influenced by a constraint rather than determined by an optimization. The user is allowed to select a maximum airmass for acceptable scheduling. The airmass at which an observation is made affects system throughput and image quality, and while you might think of this constraint as “what is the highest airmass that I can tolerate?”, the limit selected actually plays a more significant role. Once the hard limits for scheduling an observation are computed (determined by target position in relation to twilight and airmass at each site and by the time window specified in the request), the scheduler then tries to put the observation at the earliest time that it can be accommodated. It does this to maximize the chance of success in the event of a technical failure or weather outage. There are two important consequences to the airmass limit at this point. The earliest time may correspond to an observation from the opposite hemisphere, even though it would later be possible to observe the target at a much lower airmass from the same hemisphere. Second, if the target is rising, then it will schedule the observation for the time when it has just passed the airmass limit. Thus, in many cases, the result is an observation done very near the maximum airmass limit.
The downside of making the airmass limit too low is that you are decreasing the span of time during which your observation can be made. This only matters if you make it so low that your observation is not executed because it cannot contend with other requests during the limited number of opportunities that it has.
We understand that using the airmass in the global optimization for the schedule is a desirable upgrade, and we are working on it. In the interim, we are trying to provide a simple improvement and so we are changing the default airmass limit from 2.0 to 1.6 Note that changing the default does not have such a dramatic effect on contention because everyone’s usable window has been shortened. However, there are certainly cases where this is not a good number to use – e.g., if your object is close enough to the poles (above declination 81 degrees) that it will never get below this airmass from our sites. So, remember that you do not have to leave this limit at the default value, and in some cases you should certainly change it.
What is the best value to use for the airmass limit? The answer is complex, and it depends on both predictable things, like the position of the target and the distribution of possible observing sites, and on unpredictable things, like what the level of contention is for the possible places that the scheduler could put the observation. Generally, you should use the visibility tool to understand the predictable things – remember that the scheduler will try to put your observation the first place that it can.
Here are some considerations for setting the airmass limit.
B: 30%; V: 17%, R: 9%, I: 7%
FWHM ~ (airmass)0.6
In the end, there is no substitute for thinking through the implications of the limit you select for each observation request.