Netherlands Serviced and Flexible Programme

It is a long-term goal of the JCMT to fully switch to S&F scheduling because it maximizes efficiency by matching projects to local current conditions. The experiences gained from the current Netherlands programme will hopefully prove helpful in this respect.

For those who are not clear on what S&F observing means exactly, for M95B we came up with the following rules (which not necessarily will be taken over as such by the JCMT as a whole!):

1. All observations will be carried out by local JAC staff (Fred Baas or myself) or by a designated observer under close supervision of a local staff member. (unless the P.I. explicitly chooses to come out for the observations)

2. On any particular NL night those NL observations which best fit weather, available equipment, etc. will be carried out, irrespective of when they have been scheduled for on paper. (unless again the P.I. has chosen to come out).

3. Projects are deemed completed when they have reached their allocated time or aimed-at goals, whichever comes first. To qualify the latter: if the stated goal is reached at or during the final shift (e.g. because we could use on-the-fly mapping or there are better receivers) the remainder of the project's time will convert to "low-priority" until PATT projects 'competing' for the same LST-range have been finished.

Common sense should prevail: we want to avoid having to fill part-shifts with finished projects simply to reach their allotted PATT time when other unfinished projects are in need of that same time. For each project not more than several hours (up to perhaps one shift for the longest allocations) will be re-assigned to low-priority.

Semester 95B

The general experiences with S&F scheduling during Semester 95B were very positive and observing was quite efficient. A full 100% (!!) of the scheduled low-frequency (A,B) observations was completed and we only lacked in the C-band because of the weather (there were very few CI nights in spite of a reasonable number of CO(4-3) nights). This was true in particular during the latter half of the semester. In January we already started on some 96A A/B-band observations to compensate for the lack of C-band time (the affected 95B project got PATT time for 96A).

Of the approximately 10 projects, 7 participated fully in the S&F scheduling. Three P.I.' s (1 novice, 2 experienced; two flew in from the USA mainland rather than the Netherlands) chose to do their own observations, although all opted to trade time with the flex observations rather than carrying out their backup programme in case of adverse conditions. In addition one observer was asked to come out to assist with a 15 day observing block. In the end, for the majority of the projects the (remote) P.I. was closely involved in 'real-time' during the observations.

One feature of S&F observing that was appreciated was the ability to compensate for pilot/equipment/weather failure by re-scheduling the missed observations at other times (i.e. day-time or time won due to increased efficiency). We managed to recover all clear failures without losing time from other projects. Also, on one occasion we could combine two projects very efficiently reducing the number of tunings and pointings needed.

A number of deficiencies were noticed, though:

the lack of easy ways for 'remote eaves-dropping', by which I mean that observers in NL look at the observations as they are being carried out by the local staff while providing timely feedback where necessary.

For us this is the preferred mode of operation for long projects. Not only does it provide another set of eyes to spot problems, it also means of course that the P.I. has real-time control over the project even when not physically present at the telescope.

clear observation prescriptions. This is mostly for our benefits. If the occasion arises that a quick change of plans is needed, the PATT proposals do not have the correct format to enable easy decisions: LST-range, receiver and observation sequence need to be pieced together from the tables and text and compared with the overall conditions.

Also, PATT proposals are written to get the observing time, not as recipes for the actual observations and, on occasion, the difference is quite noticeable! Almost always there is a need of one or more communications with the P.I. to clearify details.

We are trying to deal with these two issues during semester 96A. Already during 95B a 'prototyping' log tool was installed which could generate a standard T.O.-type log showing a large number of parameters for each observation (read from the observation file header) and to which the observer could add comments. Even while the implementation was relatively crude, this enabled the remote P.I.(' s) to remain quite well informed about the observations without putting too much of a burden on the observer. Since the log entry can be created as soon as a particular observation starts inspection of the listing approximates looking at the real-time displays (which are less easily 'exported').

By the way, we experienced first-hand that the task of

1) directing the observations,

2) keeping an accurate log,

3) inspecting the data,

4) communicating with remote observers

is almost impossible to accomplish by a single person when the observations are a bit more complicated than trivial (to no surprise to the Director). At least (2) has been taken care of and (3) and (4) will hopefully improve soon. Also, a second real-time display (showing the incoming spectra) was installed in the control room the enable the observer to remain in front of main data reduction machine without loosing track of the quality of the spectra.

Work on a html (Netscape) based 'previewing' tool is almost completed by Tim Jenness. This tool will allow remote observers to look at (das-merged) spectra without even having to logon to the summit computers. The basic tool is working and what remains is solving security and access issues. It is my hope that the log tool and this preview tool will develop into an efficient html based 'Remote observing desk' which will show the log as it is being created by the on-site observer and enable the remote observer to 'click' on scans to list and plot them.

Semester 96A

The Netherlands S&F programme was expanded somewhat for the current semester. One additional staff (Göran Sandell) became part-time member of the so-called 'NL-flex team' so that it is no longer necessary to ask observers to come out from the Netherlands. All semester 96A projects are participating in the S&F programme, although two observers will still come out to carry out the bulk of their programme.

For all projects in this semester we have asked the P.I.' s to submit an extensive 'observation' recipe not only giving detailed information on the equipment setup but also on the observing strategy itself. Graeme Watt is preparing a similar form for the future JCMT S&F programme and we hope that between our form and the forms used for the traditional service programmes a well-designed form will result.

During semester 95B the novice observer decided to come out to gain experience and we also hope to be able to better serve that aspect of observing with the Netherlands S&F programme. Rather than coming just for the few days of their project we hope that advanced students will come for a few weeks and participate in all aspects of the S&F observations under supervision of local staff gaining a broader foundation in sub-mm astronomy.

In conclusion:

In spite of its somewhat 'ad hoc' implementation the Netherlands S&F programme is functioning very well. We clearly have gained a higher degree of efficiency in delivering results, while at the same time decreasing the burden on the observers. There is, of course, a larger burden on the local staff but this is in part offset by an increase of the number of support astronomers and the increased flexibility within the flex- team which in practise operates as a single unit. From a personal point of view, I find the much closer involvement in the observations and astronomy much more gratifying than being peripherally present during the first shift only.

Remo Tilanus, JAC