Attendance: Shaul, Amy, Julian, Al, Raphael, Bill, Charles, Lloyd
Notes by: Shaul
Shaul reviews schedule for decision
Amy goes over budget
Budget is at cost; Amy recommends going below cost by ~$100M because decadal panel's cost estimators may increase cost, in which case we could be beyond cost. Shaul notes issue for EC to discuss/decide: should we go at cost or below cost?
Line 7: based on EPIC, inflated to today's $$, for 4 years, 50% for 1 year. Al: what drives the cost? This seems too high. Amy: look at EPIC report. Action: look at EPIC report for this cost.
Line 11: based on NASA guidelines. Note: phases A-D = construction; phases E,F - flight,science operations.
Line 13: based on EPIC.
Line 14: reasonable estimate to within 10%. Note that power is a significant driver. Currently assuming 1 kW, 80% in cooling and 20% in readout. Al notes that the margin for the cooler should be in heat lift, not in power. Spec the cooler to x2 the heat lift, but don't inflate the power because once the heat lift of the cooler is specified, it can not take more power. NASA will check whether cooler is 'in family/out of family'. It will be out of family, but we need to explain. Bill notes that 200 W for readout power seems aggressive. Action (for JPL): Need to check whether we can fit this.
Line 17: cooling costs $60-$90M. Shaul: Significant increase relative to previous estimate. Al suggests we get ROM costs from vendors.
Integration and Testing: Bill points to the Planck experience. More integration and testing would have been useful. It only had 3 cool downs of the entire instrument end-to-end. Shaul: when is the time to refine this estimate? Action for Amy - Look at Planck, and the CDF for CORE and see what they assumed and whether we should refine our estimates.
Spectrometer/Imager: with $38M for the spectrometer, we have $30-$60M left for the imager, ~x3 smaller than EPIC IM.
Julian: what should we do with international contributions?
Shaul: this is a report to the decadal panel about what can be done within $1B. We are not writing a proposal. We can highlight that we would have partners, but this is not so relevant now.
SH: What emerges is a 'liteBIRD (LB) class' imager and a mini-spectrometer.
Raphael and Julian: how should our decision depend on LB?
Julian: Should we assume LB happens and we are putting forth the *next* space mission? If so, it doesn't make sense to propose a LB-class imager now (+spectrometer).
Charles: We shouldn't assume stuff that may happen. LB is 2027 at the earliest, not that far from the time scale of a Probe. Proceed with our preferred path.
Kogut: Agree. It would be a mistake to base our choices what may happen many years down the road.
Lloyd + Raphael: We are learning that it is difficult to fit two instruments in $1B. Worried that both instruments are compromised by trying to fit both in the envelope. Neither would appear compelling.
Shaul: we will prepare a concrete smaller imager, with resolutions and sensitivities, and attempt to provide more detailed trade-offs. Will continue next week.
Fundamental Physics delayed.
Hi Shaul,
since there was no time to discuss them, and I didn't include the numbers in the slides, I just thought I would send an update here. Some of the numbers relevant for the STM I find for the strawperson are (the numbers are the uncertainty on the parameter)
| | 55 % | 77% |
| ns | 0.0015 | 0.0012 |
| tau | 0.0021 | 0.0018 |
| n_run | 0.0024 | 0.002 |
| N_eff | 0.039 | 0.033 |
| m_nu | 43 meV | 36 meV |
| Omega_k | 0.0018 | 0.0015 |
Neutrino mass does not include DESI BAO prior here. I should get in touch with Nick to see if we can provide a joint constraint on neutrino mass from lensing and clusters internal to the probe. It seems one can get interesting numbers without using any other data.
I'm still waiting to compare and check these numbers with other people, Jens is still running additional constraints on annihilating dark matter, etc.
f_sky is chosen by hand. My remaining homework is to understand how much of the sky one might hope to use with the strawperson configuration.
My second homework is what we discussed before, turn my forecasting code into a code that allows optimization.
I didn't include r at this point because it requires some thought how we model foregrounds. I wasn't able to join the discussion with Josquin today because of the CDT call, but I should talk to him.
Let me know if there are questions or other things we should do. Otherwise I think it will be interesting to see how things degrade as we change the configuration.
Best,
Raphael