21cm / Intensity Mapping. Adrian Liu, UC Berkeley

Transcription

1 1cm / Intensity Mapping Adrian Liu, UC Berkeley

2 A new project cannot be incremental

3 Measure all the modes!

4 But how? And what does that get us?

5 For now, a lot of the details don t matter CHIME

6

7 Next steps

8 Keep going with the current instruments COPSS GBT

9 R&D in conjunction with better understanding of systematics Enhanced simulations and hardware development High data rates; high throughput computing leveraging DoE HEP expertise? Next-generation correlator technology leveraging DoE understanding of radio frequency tech? Real-time ionospheric calibration?

10 Instrument agnostic forecasts given unique properties of 1cm and intensity mapping Given a noisy map (from HI, CO, ) with infinite radial resolution from < z < 6, what does that mean for our ability to measure power spectrum, bispectrum, Given specific measurements of P(k) with no overall normalization, what can we say about early dark energy? About N e? About f NL? What could we measure in conjunction with CMB-S4?

11 Instrument agnostic forecasts given unique properties of 1cm and intensity mapping Higher redshifts: Is reionization/astrophysics actually irrelevant to cosmological science cases? funded: primarily a reionization experiment, but could try for Same instrument agnostic forecasts for prereionization redshifts. What s the argument besides a new redshift regime and more modes?

12 What about the?

13 Rmfp cent) h (with 5 per (multi-z) fid,mfp 1000 fid 100 h (with 5 R per cent) (T min ) log vir (T min log ) vir,fid (40.38) 1.03 (.06) 1.05 (5.43) The is an observatory, 1.03 (11.81) 1.00 (11.99) 1.00 (1.95) an 1.0 (6.11) 1.00 (10.04) 1.00 (0.96) not targeted experiment h (with 5 per cent) Greig & Mesinger Rmfp mfp (Mpc) Rmfp (Mpc) Table. Summary of the median recovered values (and associated 16th and 84th percentile e min and the associated IGM neutral fraction, x 0, Rmfp and Tvir H I for all considered observ a 100 modelling uncertainty). Our fiducial parameter set is ( modelling 40 uncertainty 60 and without 80 which results in an IGM neutral fraction of x H I = 0.48, 0.71, 0.83 at z = 8, 9 and 10 respect Instrument Parameter (% error) min log10 (Tvir ) Rmfp (multi-z) min ) Rfid,mfp 14 log10 (Tvir,fid fid potential gains by shortening the on the EoR model parameters could 1.3 (40.38) 1.03 (.06) 1.05 (5.43) ng the cosmic variance). In e ect, considering alternative statistics of t (11.81) 1.00 (11.99) 1.00 (1.95) 6 (6.11)as the 8 w have the same sensitivity to the statistics, such bispectrum (e (10.04) 1.00 (0.96) 44 6 On smaller scales, the decreased 15) and other non-gaussian probe min 0 (Tvir min min [K]) log (T vir log10 [K]) 10 (Tvir [K]) 5. 4 Table5. 3. from the deep and would likely benefit from increased s medium-deep llow survey ensures these are preintermediate scales and in turn cou Rmfp (Mpc) 4.8R 4.8 mfp (Mpc) mfp pletely overwhelming any potential gains by shortening the on the EoR model p EoR5.constraints in the presence statistics than the 3D spherically ave observing times (decreasing the cosmic variance). In e ect, considering alternati ty. However, all asobserving noted in GM15, strategies now have the same sensitivity to the statistics, such as th PS on large scales. On smaller scales, the decreased 1 cm4. 15) and other non mall scales does not significantly 4.0 noise contribution from the deep and medium-deep thermal would likely benefit s multiple epoch observations, as survey5 ensures CONCLUSION surveys relative to the shallow are preintermediate scales fesc R esc Rthese (Mpc) mfp mfp (Mpc) mfp still4.6 adequately sampled fromthe theeor constraints in the presence ferred for recovering statistics than the 3 The reionisation of a modelling uncertainty. However, as noted in GM15, epoch is astrophy

14 The is an observatory, not an targeted experiment

15 Summary 1cm / Intensity mapping has unique survey properties. Perform instrument-agnostic forecasts for now, respecting these properties. Near-term R&D efforts, particularly to understand systematics. Optimized instruments have the potential to get to science more quickly and precisely than the.