TTA Tools: Friday, 11 am MT, 1 pm ET, 10 August 2018

Logistics

Agenda

• Project Decomposition (All)
• Observing Strategy, Scheduling Blocks, and Preliminary Priorities (All)
• Future Meetings Date/Time (Dana)
• AOB

Minutes

• Project Decomposition.
  • Summary of Previous Discussion: DanaBalser summarized our previous discussions. We had agreed that the structure should start with a Proposal which specifies the science objectives. From this one or more Allocation Requests would be made which specify the purpose, justification, etc. After the review process there would be an Allocation Award. In many cases this would be the same as the Allocation Request. The observatory would then generate one or more Projects which would be divided into Programs and Scheduling Blocks (SBs). There was some disagreement about where the information that specifies the details of how to generate SRDPs is located. One constraint from the SRDP group is that the Proposal should include what products should be generated and that they should only be generated once.
  • Monitoring Example: JeffKern discussed several different scenarios based on a monitoring example from LorantSjouwerman that includes GBT monitoring (open sky and Sponsored), VLBA astrometry (bounding and triggered), and VLA polarization imaging. He discussed two different structures. (1) Projects tied to Products. Here the information on how and when to start the pipeline to produce the SRDP is contained with the Project. The downside is that multiple projects will need to be generated for each of the observations listed above. Each of these will need a separate Program. (Currently, the VLA Programs divide Projects by array configuration and proposal priority.) (2) Programs tied to Products. Another approach is to have the products be tied to the Programs. In this picture the SBs are tied to each Project and the Programs, which contain information pertinent to SRDPs, will provide information on when and how to start the pipeline. Therefore the execution blocks (EBs) are tied to the Program. The downside to this approach is that it is a big change. StephanWitz asked if this example was common. We felt that it was not so unusual given all of the triggered and monitoring projects in 19A and expected in the future.
  • Simple Example: DanaBalser suggested a simple example of imaging 10 sources with the VLA at one configuration and frequency. How many SBs or Programs would be generated? This depends on how how the user specified the SRDPs. If they want images for each source asap, then it depends on how the SBs are organized. For example, maybe three sources in SB1, three sources in SB2, and four sources in SB3. In this case there would be 3 SBs, each with a unique Program.
• Observing Strategy, Scheduling Blocks, and Preliminary Priorities. The process needed to generate SBs for observing is not too different than the process needed to prepare for the TAC meeting. We need to generate an LST pressure plot for each telescope with preliminary Priorities. This is done differently for each telescope with Sessions as input (VLA uses the priorizer, GBT uses DSS simulations, etc.). One requirement for SRDPs is that we have to automatically generate SBs. Therefore, this machinery could be used for the TAC process. ToneyMinter does iterate during the TAC meting, whereas AmyMioduszewski does not, but she could do this with the PHT. The TAC has not made as many changes to the preliminary Priorities as it has done in the past. Currently, the Priorities are tied to Sessions. How should we do it in the future tools? Is this a policy decision? LewisBall said (offline) that the tools do not need to support setting priorities at the source level (with the exception of conflicts), but the the Allocation Request seems to be the right place. ALMA does this at the proposal level. AmyMioduszewski pointed out, however, that it will need to depend on LST. Also, currently, the LST pressure plot generated post TAC meeting is rather different than the SB distribution with LST. We could restrict this but LorantSjouwerman worried about gaps in the schedule. It would useful to run through some Use Cases.
• Future Meetings Date/Time. It appears Friday's will not work, so DanaBalser will create a Doodle poll for Tuesdays and Wednesdays.
• Action Items:
  1. Everyone will think of some Use Cases that we can discuss at the next meeting.
  2. DanaBalser will create a Doodle poll for future meetings.
  3. DanaBalser will start to update the memo.