ALMA SSR Committee Phone Meeting 2003-07-09

Connection Details

• Call date: 2003-07-09 (Wednesday)
• Call time: 15:00 UT
• Duration: 1 hr
• USA Toll Free Number: 888-795-6990
• Toll Number: +1-484-630-4370
• Passcode: 99640
• Leader: Robert Lucas

Attendance

Draft Minutes

1. Subsystem Scientist Activities (all)
   
   
   • Jeff Mangum -- Control Participated in CDR1 review. ATF activities also directly applicable as ALMA control system proving ground.
     
   • John Richer -- Telescope Calibration Main activity: CDR1 attendance. No major problems, but this is a subsystem with many complex interfaces - ACS, aips++, archive to name but three. The actual computations should be straightforward, but most of the effort is in engineering the software to work with the rest of the system - many parts of which are still fluid.
     
   • Peter Schilke -- Data Reduction User Interface
     
   • Steve Scott -- Correlator Reviewed all correlator documents for CDR1.
     
   • Debra Shepherd -- Off-Line Analysis [see Debra's message on Testing Plan]
     
   • Leonardo Testi -- Observing Program Preparation
     I participated in two subsystem telecons, including the CDR1 one. We discussed a lot the design of the spectrometer setup tool, for which we are developing a detailed design and sort of UseCases. When these will be finalised I will circulate them within the SSR for further comments.
     We also discussed the test plan in several occasions, I started writing test cases in close collaboration with the team and Debra (to build up a consistent test plan). During the CDR1 review it was agreed to reduce a bit (wrt the original plan) the first testing session on R1 in the fall. We now aim at the preparation of a good sample (although not necessary a complete one) of test cases by the end of the summer. These will be helpful both as guidance in the design and for a reduced set of tests that will be performed by myself and Debra possibly with the help of a few other individuals. These test cases will be circulated by Debra together with some other developed for offline and pipeline.
     We are still looking for volunteers to write some of these test cases for areas we are not competent enough to provide good cases:
     • Pulsar observations
     • Solar system observations
     • VLBI observations
     • Programmes with Total Power observations
     • ... whatever you think needs a special test case (meaning a particular setup or parameters or ...)
     We do not need all of them immediately, and we can possibly wait until the Design Reference Science Plan has been completed (which can be helpful in preparing these cases).
     
   • Francois Viallefond -- Archive
     
   • Chris Wilson -- Pipeline Reviewed the four new Pipeline documents and provided detailed comments for CDR1. As part of that process I paid some attention to what work needs to be done to respond to PDR and other comments, and I identified the following possible items (but haven't done any work on any of them yet):
     • Take SSR requirements for Pipeline to another level of detail
     • Performance requirements specifically for the pipeline
     • Pipeline testing plan (similar to Debra's Offline Plan)
     • Identify a single standard observing mode, develop heuristics, implement, and test in a pipeline
     • More attention needs to be paid to requirements, planning, and development work specifically for the Quick Look Pipeline
     • Planning and priorities need to keep the goal of Early Science in mind
     
     
   • Mel Wright -- Scheduling I have run a few more tests of the dynamic scheduler I used for Hat creek, using different mixes of projects, more likely to be scheduled for ALMA. The dynamic scheduler is a MIRIAD task and was originally designed as an off line simulator. I used the weather data from Chajnantor as before. The pass/no pass algorithm seems to work very well for short ~ 1-2 hour SBs.
     Actually it works better than it does for Hat Creek observations where people like longer ~ 8 hour "tracks" rather than shorter SBs. The highest priority projects projects are completed first - as soon as the requested LST range and weather conditions are satisfied. When an SB is scheduled it runs for its allotted time, before the priority of the queue and the weather are re-evaluated, so with longer SBs, the weather can change during the SB and a more suitable (or higher priority) SB cannot be scheduled until the current observation is completed.
   
   
   
2. CDR1 status (R. Lucas, Subsystem Scientists)
   CDR1 is well underway. Nothing significantly affects the SSR requirements so far.
   
   
3. R1 deliverables (R. Lucas)
   1. Revise U. Requirements in Off-Line
      Debra will list those requirements and send the list to the SSR group (mid august)
   2. Observing Modes for Interim Science
      We should prepare this work and propose a list of observing modes, the Science IPT should be also involved.
      It was mentioned in one CDR1 meeting that phased array mode (VLBI) would not be available at start - this should be investigated.
   
   
   
4. Test Plans (Debra Shepherd)
   Debra has sent out her guidelines for the elaboration of test plans for R1. This is a very important activity, possibly the most important for SSR in phase 2.
   While the first part (test use cases) involve mainly the SSR (subsystem scientists) the test plan themselves are a collaboration work with the subsystem leads. The guidelines and examples have been elaborated on experience gained with OffLine and Proposal Preparation. Differences will be encountered depending on the subsystems as they have different deliverables, user interfaces, ... Subsystem Scientists should not hesitate to contact Debra as needed during the next month to get her feedback on the specificities of your subsystem.
   
   
5. Sci. IPT comments on Data Rates (J. Richer) The reply from the Science IPT is presented by John. The Science IPT agrees with all the propositions of Steve's memo. Clearly more discussion is needed (and has already started by email) on point 12, which is an additional proposition to lossily compress the data archived on the long run (while the full data will be available to the observer). As Joe mentioned, the currently budgeted data rate is still 6/60 Mbyte/s, and no increase proposition (e.g. based on the numbers in Steve's proposition) has been made at the project level yet.
   
   
6. Face to Face meeting
   The next face to face meeting will be in Socorro on Nov. 14/15.
   
   
7. Date of next phone meeting According to the schedule nxt meeting was on August 13. As R. Lucas will not be available we decide to skip that meeting. The following one was be on: 2003, September 10th 15:00 UT.