NRAO Home  >  Green Bank  |  Wiki Topic:    GB > Dynamic > Reviews07 > CvAgenda07

Changes | Index | Search | Go

CV Agenda, Talks (45 minutes total)

Talks

All talks, combined

• Intro (Karen 5 min): what, why, timescale - Intro talk
• Ranking, weather, lookahead (Jim/Dana 20 min)
• Software overview (Mike/Mark/Eric 5 min) - Software talk

Discussion

• Hogg - Has the assumption that observers need to conduct their observations been examined?
  • O'Neil - Yes, but we need to have a mature data reduction pipeline and data quality tools in place first before we can accept the responsibility for data quality. We are continuously examining this assumption and expect to handle more and more service observing proposals.
  • Rood - Some experiments require the observer to be present, e.g. experiments dominated by systematic error.
• Lo - Science grade is an important weighting factor.
• Brogan - By allowing observers to tweak observing parameters, can't they manipulate the system to get their time?
  • They could, but run the risk of getting bad weather and consequently bad data. Top ranked proposals still would win though for the good weather. If this becomes an issue, we can set these in the DSS and allow them to be only tweaked at proposal time and require a justification for the tweak. We hope that we won't have to impose this.
• Frail - We need to make "seasonal observing frequencies" graph T-shirts for the DSS.
• Mangum - How does the system look from the user's point of view?
• Frail - How are you handling things like prime focus feed changes and scheduling projects desiring different feeds?
  • O'Neil - We know that we need to do this and we also expect to get good feedback through experience to figure out how best to handle this.
• Ransom - Will longer Telescope Periods be trumped by similarly ranked shorter Telescope Periods in the scheduling algorithm because it is easier to schedule the shorter periods?
  • Condon - We simulated this and found it wasn't much of an issue. And the factors used in determining a score work in such a way (e.g. timing issues and weather) that we don't expect this to be much of an issue.
  • Shelton Note - Plus we are using standard computer science packing algorithms to do the scheduling, which were created with exactly these complications in mind.