Session notes for project planning & preparations for decadal
Adrian Lee: Overview of CDT straw man configuration
Disposition of telescopes to sites was not in the CDT report.
Q. Any thoughts on a more northern site? A. Not a huge improvement over Chile + Pole, so not part of the CDT straw man.
CDT cost $412M, but some options for lower cost by re-using existing telescopes and infrastructure.
Need to develop the straw man into a reference design, including telescope distribution among sites, technical selections.
Kathy Turner comments
Kathy called in for a few minutes.
We are really happy with the results of the CDT. Excellent report. Happy with the different costing studies. This is notional, so changes expected. Charles presented CDT results to AAAC and Particle Physics Science Advisory Committee. Meeting last week with DOE and NSF. Pleased to hear about progress with ICCC, collaboration, the reference design, and plans for the future. Agencies will come back to collaboration with advice.
Jim Yeck: pPDG
Key goals: 1 project; 2 sites, Chile + Pole; joint DOE/NSF oversight.
Boundary conditions. DOE can begin investments based on P5. NSF will await 2020 Decadal Survey. Historically DOE does not build big telescopes.
Implementation. Scope sequencing. Concurrent development of the 2 sites. Pursue the 3 CDT science goals in parallel.
Discussing scope distribution between the 2 sites, agency coordination, phased transition to operations.
Timeline. CD0 late 2018. CD1/3a 2021 (CD1 is a critical step for DOE), CD3b 2023. NSF CDR review 2021, PDR 2022, FDR 2023, MREFC funding 2024. Agency response to timeline on 2/28/18 was generally positive.
Key message. Important to move on technical decisions ASAP. All previous experience points to this. Develop a technically limited schedule, then adjust to suit funding reality later.
Decadal CDR will need to be supported by development of a reference schedule and costing.
Need to define the role of legacy investments, international collaboration, and private investments (e.g., Simons Observatory).
Mark Devlin: Lessons from Simons Observatory
6yr 73M program. 6m cD telescope. 4x42cm refractors. First light 2021. Adjacent to ACT & Simons Array.
Difficult to do the right thing at the right time. Need to keep on track. Some decisions will go against best personal interests. Merging separate teams is a huge challenge. Stage-3 teams are used to running their own show. Advice is helpful, but pushing too hard is not. Don't let the project split into independently run smaller projects. Find tasks that can be solved once for the entire project. Make decisions early and often. Don't study for another month to make it better. Be honest about the limits of your ideas and technologies. Say what is wrong. Don't make decisions for personal/political reasons. Communication is key, but telecoms sap the life out of the team. Skip telecoms whenever you can. Generate post telecom summaries. Don't generate ideas in real time. Stop telecom bullies. Use hack-style targeted meetings to work out problems. Send out a weekly project update. Be realistic about what you can contribute. Deliver what you say you will do. Reward all of the time people put into the project.
Comment from Charles. Don't make the same decision often. A. SO doesn't revisit decisions. Everyone claps when they make a decision.
John Carlstrom: Planning for the Decadal
CMB-S4 will be an MREFC. Enter the queue if and only if the Decadal Survey ranks the project as high(set) and NSF deems ops affordable.
Originating NSF Organization must be prepared to fully fund pre-construction costs.
Design stage lasts 3-5 years, costs 10% of construction cost.
Decadal timeline. White papers Mar 2019. Additional inputs from selected projects in Spring 2019. CATEs in Summer/Fall 2019. Survey executive session 2020.
We need to provide a reference design and cost estimate that is good enough to support CATE.
Action items. New name? Full A&A science case (new science book). Build community support (especially non CMB community). Submit white papers. Coordinate with other MREFC projects (NGVLA< Billion Object Survey). Prepare Decadal CDR to show we are ready to go.
CDR. Transformational science. Conceptual design. Collaboration. Ready to go, not a bunch of options. Glossy graphic. Responsible parties are Yeck, Staggs & Carlstrom, at least for now. Schedule: internal review Nov 2018, ready to submit to Decadal Spring 2019.
Reference design. Goal is to demonstrate a feasible, low risk, costed project. Not the final design. Choosing between options is not a down select. Need to decide major components of reference design ASAP. Need to identify any long-lead design items. More detail is better, but be realistic.
Charges to parallel sessions. Determine as much of reference design as possible. Where is there consensus? What do we need to do to make decisions? What are the viable options? Set the path for drafting new chapters for the science book. Determine chapter titles, outline, scope, sims & forecasting, identify authors. For sims & forecasting identify requirements, schedule, and responsibilities.
Q. Clem is concerned about setting out all the experiment details when it is not really the final design. A. The reference design doesn't have to be super detailed, but it needs to lay out enough detail to convince agencies.
Q. Akito doesn't know how we will build consensus on the design.m A. Parallel sessions tomorrow will help.
Q. Zeesh wonders how the level of detail will be kept uniform. A. We want a fairly uniform final report, so we do have to watch this. Suzanne notes we will have multiple drafts.
Q. How does CATE assess risk?. A. Not clear. Aerospace was not so great at understanding ground-based projects. Need to guide them.
Comment from Devlin. CATE wants to see receipts. SO will have these.
Q. What about other competition for MREFC? A. NGVLA, billion object spectrometer, TMT? Radial velocity program.
Comment from Charles. Aerospace will likely do the CATE again. May be possible to work with other parts of Aerospace to address cost and risk methodologies for ground-based projects.
Tom Crawford: Extended science case
Concept, funding, understanding of power of CMB-S4 have evolved, so CMB-S4 science case has evolved.
Parallel sessions tomorrow will look at what is the extended science case, and how we construct science and measurement requirements for that case.
We are along way from defining what the extended science case actually is.
What do we all this science?
Q. What performance will be used for developing the science case? The details really matter, especially resolution, so we need a rough reference design and cost sensitivities. A. Start with the performance and cost sensitivities from the CDT report.
Comment from Julian. We don't have a lot of time for iteration. Need to fit the science case to whatever we come up with by March. Comment from Carlstrom. If there is a killer science case, we want to know about it.
Q. Ruhl. CDT design was based on r and Neff. Does a new science goal have to rival those? A. Crawford. CDT claimed that requirements for r and Neff would cover all the other science too. CDT required the new science to be cost neutral. Ruhl doesn't want to spin his wheels on a reference design if we are going to change it for the extended science. Suzanne thinks the extended science case is very important for the Decadal, and that it will cause some changes in the instrument design.