supernova triggering part i needs and questions
play

Supernova Triggering Part I: Needs and Questions Amanda Weinstein - PowerPoint PPT Presentation

Supernova Triggering Part I: Needs and Questions Amanda Weinstein Iowa State University Hierarchy of Needs Most critical to least critical Caveat: This is my (which does not imply items approximation and should at top not important)


  1. Supernova Triggering Part I: Needs and Questions Amanda Weinstein Iowa State University

  2. Hierarchy of Needs • Most critical to least critical Caveat: This is my (which does not imply items approximation and should at top not important) not be taken as “blessed” • Ideally, automatically enable by the SNB group items at top when designing Externally trigger (e.g. for the items on the bottom. SNEWS) a SNB at d < • We need a similar pyramid 15 kpc and record for DAQ requirements physics data Self-trigger a SNB at d > 15 kpc Externally trigger (e.g. SNEWS) a SNB at d < 15 kpc and record physics data Self-trigger a SNB at d < 15 kpc and record physics data 10/9/2017 Amanda Weinstein DUNE DAQ

  3. Supernova Flowchart End-of-life mass > 1.4 Binary No White companion M ◉ ? (Birth ? mass > 8 M ◉ ) dwarf End-of-life mass Yes > 1.4 M ◉ ? (Birth Yes mass > 8 M ◉ ) Possible thermo- nuclear supernova Explosion Core mass succeeds > 3 M ◉ ? ? We want these CC SNe, neutron star Black hole

  4. What physics do we want to capture? Potentially Potentially detect Si- useful burning information phase at later before times (up to explosion in ~30s) neutrinos Intertwined info about shock development and neutrino physics

  5. Spectral Features • Different oscillation physics in neutrino’s journey from proto- neutron star to us - MSW effects (r > 200 km) - “Collective oscillations:” (r < 200 km) - Vacuum oscillations (once away from star) • Flavor-specific burst evolution carries information about mass ordering and SN processes • Key requirements: - Energy resolution <10% (in our Dasgupta, Dighe, Raffelt and Smirnov control) - Energy threshold ~ 5 MeV (mostly - Well-determined light curves in our control) - Statistics (only partly in our - Time-integrated and time- control) resolved energy spectra

  6. What physics do we want to capture? • We want self-triggering and whatever pointing we can get (alerts to other experiments) - Recall that in early period EM radiation does not escape, but neutrinos do - Neutrinos crucial with weak or dust- obscured SNe - “Failed” SNe: have a distinct neutrino signature (e.g. continuously hardening spectrum, abrupt cutoff) • Pointing information: allows optical, infrared telescopes to be ready and on target - Elastic scattering events are best for this but we don’t get many

  7. Factors controlling statistics Distance (inverse square law) Failure to trigger/record Does the 90% • Energy resolution degradation events near threshold(can number Detector inefficiencies • ameliorate by dumping include both • Not having prompt light (worse t0) raw data in long window) of these? • Other losses (e.g. neutrons) 10/9/2017 Amanda Weinstein DUNE DAQ

  8. SNB Trigger • What can this mean? • External burst trigger :a vetted alert we send to other experiments • Internal burst trigger: triggering on the supernova as a whole, i.e on a multi-event signature that occurs over a period of time - Pros: could be as simple as looking for characteristic rate changes • Increases robustness against “wiggles” of radiological and other backgrounds • - Cons/challenges: potential for severe model-dependence • Triggering at the event level, i.e. reacting to the present a cluster of “SN-like” interactions - This raises questions like: what exactly do we mean by an “SN-like” event? • These two things are not mutually exclusive—the second is essential to accomplishing the first. • Question : how do DAQ operations change, locally and globally, if an SN trigger is received? 10/9/2017 Amanda Weinstein DUNE DAQ

  9. Background :Key Numbers • Prototype radiological • Note: Ar 39 rejection requires background: Ar 39 some level of time information - 1 Bq/kg, so 10 7 Hz per 10 kt - Time resolution better than detector 100 ticks (between 20/30 -100 ~2000 hits per drift window from • all Ar 39 in 1x2x6 vs. ~30 from depending on binning) actual neutrino - Individual Ar 39 events low energy and produce isolated • Note: other less-studied hits backgrounds may be higher - # channels needed for rejection energy and tougher to (variable) (K. Warburton) disambiguate. Full disambiguation: 2 to 3 • Raw ADC threshold (2-6, - Don’t yet have hard numbers • threshold dependent) on these. Individual Ar 39 • 10/9/2017 Amanda Weinstein DUNE DAQ

  10. Burst over time • Early time structure of SN neutrino flux sensitive to mass hierarchy. Good news for physics, bad news for triggering 10/9/2017 Amanda Weinstein DUNE DAQ

  11. Timescale problems I • Neutronization burst spike is clean, and low-latency, but far from guaranteed! • SNB-like event density over a longer time period could be characteristic. BUT longer latency (in normal hierarchy, increase in events may not be recognizable for as long as 0.1 seconds) . • Do we try for a single robust model independent criteria, or just multiple burst triggers based on different scenarios? 10/9/2017 Amanda Weinstein DUNE DAQ

  12. Timescale problems II: TPC vs PD • On a completely different note, we have a second timescale problem in the mix • TPC and PD operate on very different timescales • PD efficiency s.t. we likely need to trigger most SN events off TPC • Question: How do we ensure we get max # events with PD information and hence good t0s? 10/9/2017 Amanda Weinstein DUNE DAQ

  13. Questions in Summary • What defines a single-event SNB trigger? - How local is local, in space and time? I.e. how many channels and how large a time bucket do we need to define a single-event SN trigger without other (e.g. radiological) backgrounds swamping us? - We’ve made some progress on this with Ar 39 but still fear the unknown unknowns. - How do we reconcile the timescales of the different systems (TPC vs. PD)? • What defines a burst-level trigger? - How big a chunk of the burst do we need to accumulate data from before the burst- level trigger is robust? - How do we reduce our model dependence in the burst-level trigger definition? - What do we tolerate in terms of fake rate internally vs. in terms of alert rate? • What precisely does DAQ do in response to an SNB trigger? (see part II) • What is the maximal amount of compression we can tolerate before losing 5 MeV threshold and/or degrading energy resolution at low energies? (relevant to external triggering) 10/9/2017 Amanda Weinstein DUNE DAQ

Recommend


More recommend