Change page style: 

Visiting Instrument Policy

The following policy was developed in discussion with the Gemini Science and Technology Advisory Committee (STAC) and Gemini Board of Directors; it describes the principles on which a visiting instrument may be accommodated and used, and describes two routes via which the PI of an instrument team may seek access for their insturment. It was approved by the Board at its November 2012 meeting.

Principles

  1. Visiting Instruments must be scientifically productive; hence observing runs are normally allocated time by the National Time Allocation Committee (TAC)/International TAC process. Where performance needs to be determined via a short run prior to wider offering, either (i) a small amount of Director’s Discretionary time may be requested or (ii) the STAC may ask the Observatory to remove a sufficient block of time from the normal schedule.
  2. Visiting Instrument observing runs are carried out by the visiting instrument team. A visiting instrument Principal Investigator (PI) may invite Gemini staff onto their team.
  3. Support of visiting instruments is subject to Gemini staff availability.
  4. Visiting instruments whose support requires a major effort on the part of Gemini staff, or which are expected to visit multiple times thus requiring a significant cumulative effort, must be offered to the Gemini user community via one of the methods identified in the next section. Instruments which can be accommodated without much effort may be used for a PI's own programme only. In the latter case, PIs may invite collaborations; this is optional on the first visit.
  5. No visiting instrument will normally be offered to the community on its first visit. However if significant interest is anticipated even before the first visit, the Observatory may optionally make known that there may be a visiting instrument and other PIs can contact the PI to request collaboration.
  6. If an instrument appears to offer a popular capability and demonstrates competitive performance, the observatory will normally pursue the appropriate level of offering if staff effort is available to do so.
  7. The work required to offer an instrument to the community should not be allowed to be a cause of the effort requirement becoming "major". (For example, if incorporation into the Phase I Tool (PIT) and Observing Tool (OT) would require a large effort for an instrument which otherwise classed as "small", other ways should be found to make it possible for PIs to apply.)
  8. Full integration of a visiting instrument which will see limited use must be avoided. Partial integrations may make sense, e.g. for observing efficiency or scheduling.
  9. If an instrument requires particular observing conditions and would attract a lot of proposals, it should be queue scheduled, either by integration into the standard queue planning process or via a simpler alternative.
  10. As Gemini is funded via the United States National Science Foundation, all data taken using visiting instruments must be made available to the community. Access does not have to be via the Gemini Science Archive (GSA).
  11. Visiting Instruments will not be provided for in the Gemini data reduction package or Quality Assessment (QA) pipeline. However, a condition of any visit with external PI involvement is that the visiting team undertakes either (i) to reduce all data and provide reduced products to the PIs or (ii) provide comprehensive "do it yourself" instructions to PIs and support them in the use of those instructions.
  12. Unless impractical due to the specialized nature of the instrument, data files made available to the community will be in FITS format, whether provided via the GSA or not. Header information will be appropriate for later "archival" use. Where possible without major effort, the Visiting Instrument Interface should be used to enable headers to include accurate values for telescope position, timing etc.
  13. If the Visiting Instrument is to be offered to the community, the method via which the community will access their data must be agreed before any run takes place.
  14. The scheduling of a visiting instrument for PI-only observations must not impact significantly on queue completion with facility instruments (e.g. by taking an entire month in peak RA time.) If the visiting instrument is offered to the community, it is acceptable to pre-schedule such a block and adjust the queue filling with facility instruments accordingly.
  15. Visiting instrument runs can be interrupted by rapid Targets of Opportunity (ToOs), provided that summit staff can carry out the ToO observation and the instrument switch can be carried out efficiently enough. In the event of disagreement, the Head of Science Operations will make the decision.
  16. Visiting instrument PIs agree to follow Gemini safety policies and will not put people or facilities at risk.

Routes for Instrument PIs

Instruments Requiring Minor Effort and/or Cost

For an instrument which may be expected to require a small amount of effort or money, the following process should be followed:

  1. Before a first visit, the PI engages with Gemini staff to determine the scale of work required to support a visit. The following points assume that that effort is confirmed to be small (e.g. an installation day and deinstallation day for the site crew, minimal software and cabling, limited night support etc.).
  2. The PI submits a proposal to their own partner TAC to support a visiting-instrument run, stating in the technical justification the outcome of the consultation with Gemini staff.
  3. If the TAC awards time, the run is accommodated.
  4. If the PI subsequently wishes to re-apply, appropriate methods to make the instrument available to the community will be determined: collaboration in a single merged proposal, separate proposals to the TAC process, etc.
  5. If in the course of the visit, the work required grows significantly above the anticipated level, Gemini staff will make a best-efforts attempt to ensure the success of the run but the Observatory retains the right to cancel it.

Instruments Requiring Major Effort and/or cost

For an instrument which will clearly require a lot of Gemini effort or money:

  1. Before a first visit, the PI engages with Gemini staff to determine the scale of work required to support a visit. The following points assume that that effort is confirmed to be large (e.g. site crew and other engineering for >1week, significant software just to get the instrument to function on the telescope, etc.).
  2. The STAC is consulted, with the following summary information: (i) the effort estimate, (ii) a statement of what other work would be adversely affected and (iii) a statement of the range of possible science that the instrument could support. The STAC is invited to weigh the effort required against other priorities, and hence to determine (i) whether to proceed, (ii) if so whether the instrument should be formally offered to the community at a subsequent visit (subject to performance being demonstrated in the first) and (iii) whether the PI should apply via the normal TAC process or the Observatory should block time off the normal schedule, for a PI-only run to enable performance to be determined with a view to future wider offering.
  3. The STAC makes a recommendation to the Board (possibly out of step with the normal Board meeting cycle, depending on the timing relative to semester calls).

Depending on decision (iii) above, either

  1. The PI asks for time via the TAC process, and if successful the run is accommodated, or
  2. The observatory accommodates the run within a defined block of time, removed from the schedule for the semester as agreed by the STAC and Board.

In either case, if in the course of the visit the work required grows significantly above the agreed level, Gemini staff will make a best-efforts attempt to ensure the success of the run but the Observatory retains the right to cancel it.

If the run is successful, the STAC is consulted again to determine whether it wishes to see the instrument back and offered formally. If the STAC decides that the effort is worthwhile and believes that the instrument should be offered to the community, the observatory determines the best method to make the instrument available. This may include incorporation into the OT, PIT, queue planning tool, observing system etc. in the case of an instrument deemed likely to return many times, to attract a good number of proposals, and to require thorough queue scheduling to maximize science return.

If the decision is to offer formally through the PIT, the proposal software is modified in time for a call for proposals and the operational software in time for the semester.

If the decision is not to offer formally through the PIT, the PI and the observatory work out a "lightweight" method for community members to gain access (e.g. via collaborative proposals with the PI, or via use of a generic "visitor" instrument in the PIT).