At the upcoming meeting in November, the JCMT Board will be addressing two linked themes: the construction of new instruments and the ongoing operations support.

It is clear that without additional resource into the Development Fund, there is inadequate funding (by a sizeable amount) to support the new projects, SCUBA-2 and CHAMP-D. While the funds required for the CHAMP-D are very much less than for SCUBA-2, the latter was easily the top priority new instrument when the community was polled.

The Board is looking to see if there is an opportunity to change the operating model so that funds saved can be transferred into development. This would then mean that the additional funds needed for the new instruments would be smaller, especially when savings can be integrated over the next decade and the costs are shared between three agencies.

By this note I am informing the user community of the focus of the debate. Users are invited to let their colleagues on the Advisory Panel know their views in advance of the meeting on November 6/7. The most obvious focus is the question of 16-hour or 12-hour shifts. This crudely could be put as a reduction of science now (less hours observing) in exchange for more science later (i.e. by having new instruments like SCUBA-2).

Users should remember that the International Review and the JCMT Board stressed that the top-priority for the JCMT was completion of the most scientifically exciting and high profile programmes. This should be borne in mind when making your input; quantity does not trade with quality.

The current model provides support in principle for 24 hour observing (when conditions permit) by having two TSSs on 12 hour shifts each. This specifically provides for a sixteen hour observing night. For all intents and purposes we should ignore 24 hour observing as it is of sufficient rarity to be almost non-existent.

The 16-hour observing night works very well in that a good RA range is accessible at any one time, the 8-hour shifts work well with observers, observers can set up for first shift before the start of the shift, second shift benefits from the facility being `up and running`, potential for extended observing hours when possible; gives the opportunity to support more proposals, etc, etc.

The downside is that transport coordination is more difficult with the TSSs being separated from the observers, there are more vehicles required, there are more TSSs required, there is a question as to whether the first shift really benefits from the early start because of inferior weather, there are more (numeric) proposals to support.

The difficulty comes when trying to quantify the above in terms of a cost-benefit analysis. Some of the costs are easy to calculate, but the science benefits or losses are harder to determine. One of the obvious questions that could give a handle on this is how the weather affects the number of hours, especially those hours in the early evening and morning. We have looked at some statistics of CSO tau and seeing for the past decade and these data are displayed here. The Science group at the JAC have discussed the interpretation of these data, but rather than sway user opinion, I leave it to experienced observers to use the data as they see fit based on their knowledge.

In terms of operating models and cost savings, there are only a limited number of models that can be implemented realistically and the most obvious is to move to a single shift of 12 hours per night. To make savings there could be a trade-off between support astronomers and TSSs, with some of the current support tasks being undertaken by the TSSs. These all have high level qualifications and some of whom have expressed an interest in taking up more varied duties. Savings on support astronomers would be made by not filling vacancies when staff leave.

As examples, the pros of a 12-hour shift are that any fault has less impact on a 12-hour shift than an 8-hour shift; the TSS is linked with the observer (as on UKIRT) simplifying vehicle schedules and having savings in number of vehicles; there are less observing programmes and astronomers to support; there are cost savings in terms of operations and staff. On the other hand, the cons could be seen to be that the shifts would not give 12 hours of full observing maybe because of initial set-ups and hang-ups; any faults have a bigger impact on a 12-hour than a 16-hour night; there are less programmes allocated time; there is less time overall; good weather is ‘lost’; flexibility is lost.

Clearly any reduction in overall observing time implies that the facility needs to operate more efficiently in terms of fault rate (and we have to do much better here) and scheduling. The latter means we need to match the programmes to the conditions and keep track of what observations have been done and how well (and here the Observatory Management Project comes into play for efficient observing queue management using software rather than people).

Alternatives

If we had remote observing, then there might be the possibility of clawing back some of the time, especially good time, by operating the telescope from HP after a 12-hour shift (actual schedule hours to be defined). Perhaps there should be a major push to reconsider some form of remote operation at least for some of the time. There may be the possibility to use some of the TSSs in a ‘fill-in’ mode to claw back periods of good weather, although this has not been explored in any detail.

I understand the difficulty that this question poses for users. I have not been given the remit ‘if you can save ‘x’ on operations then you can have SCUBA-2’. Therefore, users might be wary of what the trade-offs might be and by offering up the 16-hour nights could see them lost, but without getting a SCUBA-2 in compensation. All I can say is that the Board is not looking to reduce the cost of the overall JCMT budget, but rather a transfer of funds from one part to another.

Ian Robson

Director JCMT

27^th September 2000

Note: We have been asked for a list of advisory panel members, should you wish to write to them, their e-mail addresses can be found here.