06-20-19 ajroeth daqmeeting · 2019. 6. 20. · strange star formation • possible neutrino fluxes...
TRANSCRIPT
Supernova Pointing
AJ RoethDAQ Physics Performance WG meeting
06-20-19
�1
Motivation for pointing
• Supernova Early Warning System (SNEWS) to warn astronomers
• Finding progenitor in case of black hole formation
�2
Pointing resolution requirements• Any pointing information is useful to
astronomers, but the more precise, the better
• I found preliminary current pointing resolutions of DUNE for SNe • Pointing resolution = Angle at
which 68% of reconstructed SNe directions are closer to truth
�3
?
�4
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1Energy (GeV)
0
2000
4000
6000
8000
10000
Even
ts
eν
eν
xν
Events vs neutrino energy
• Wrote ES event generator • Chose neutrino interacted energy and flavor
(GVKM model distribution from SNOwGLoBES)
• Chose electron energy and calculated angle • Simulated 500 isotropic supernovae • 260 elastic scattering events each in same
direction (GVKM model 10kpc SN) • No noise or radiologicals • Elastic scattering events = ~7% of total,
charged current ~93%
Supernova samples of elastic scattering events
Alex Nikrant, Ranjan Laha & Shunsaku Horiuchi, Phys. Rev. D 97, 023019 (2018)
Supernova samples of ES + CC events• Used a pre-existing CC sample (made
using MARLEY generator) • Sampled according to GVKM model SN
energy distribution from SNOwGLoBES • Rotated to match ES neutrino directions • Ignoring detector anisotropy for now
• Added to ES sample to get 500 ES+CC SN samples • Each SN has 260 ES and 3350 CC
events (GVKM model 10 kpc SN) • No noise or radiologicals
�5
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1Energy (GeV)
0
10
20
30
40
50
60
Even
t rat
es
Event rates vs neutrino energy, nueCC
Directionality
• CC events less directional, but still have directional information
• Can look at how CC/ES ratio changes over signal and its effect on pointing resolution in the future
�6
1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1Cos(Angular Difference)
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
Frac
tion
of e
vent
s
CC
ES
Cosine of angular difference, reco energy 24-26 MeV
1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1Cos(Angular Difference)
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
Frac
tion
of e
vent
s
CCES
Cosine of angular difference, reco energy 24-26 MeV
Addition to current reco: daughter flipping
• There is error between true and reconstructed directions of electron tracks
• Angles between true start electron direction and reconstructed direction (θ) - show ambiguity in reconstructed electron direction - two peaks
�7
1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1Cos(Angular Difference)
0
0.05
0.1
0.15
0.2
0.25
0.3
Frac
tion
of E
vent
s
Cos of error angle, selecting longest track, after PM dQ/dx flippingCos of angular difference
• Can use daughter tracks (from bremsstrahlung gammas compton scattering, ionized electrons, etc.) to reconstruct direction of electron track • Correlate with electron direction
• Calculates average cos of angles between reconstructed track direction and location of daughter tracks, for each end of electron track
• Choose starting point of track to be the side for which avg(cos(θ)) is largest
�8
Addition to current reco: daughter flipping
Reconstructed supernova direction• Wrote likelihood function to reconstruct
supernova direction from all electron directions and energies
�9
• Minimum of negative sum of log likelihood function is reconstructed supernova direction
• Used 250 supernovae to make probability distribution functions for likelihood function, then found pointing resolution of remaining 250
• PDFs only represent ES events for now • Adding CC events would make them more realistic
Supernova pointing resolution• 10kpc SNe with no noise or radiologicals • Next step is studying the effect of noise and radiologicals on
pointing resolution • Future step is studying pointing resolution as function of
number of events • The events happen over ~10s (GVKM model)
�10
ESonly
Noextraflipping,maximumbinisSNdirection
ESonly
Daughterflipping,maximumbinisSNdirection
ESonly
Daughterflipping+likelihoodfunction
ES+CC
Daughterflipping+likelihoodfunction
Pointingresolution 147.2° 30.6° 9.7° 7.5°
Noextraflipping,nolikelihoodfunction
1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1Cos(Angular Difference)
00.05
0.10.150.2
0.250.3
0.350.4
0.45
Frac
tion
of S
Ne
Cos of angle btwn true and reco'd SN direction
ES+CC,daughterflipping,usinglikelihoodfunction
1− 0.8− 0.6− 0.4− 0.2− 0 0.2 0.4 0.6 0.8 1Cos(Angular Difference)
00.10.20.30.40.50.60.70.80.9
Frac
tion
of S
Ne
Cos of angle btwn true and reco'd SN direction
Strange star formation• A time dependent signal worth considering:
strange star formation • Hypothesis that strange quark matter is ground
state of nuclear matter • Possible existence of stars made of strange quark
matter • Drago & Pagliara, Physical Review C, 92, 045801
(2015) - model of conversion of neutron star to strange star
• Initial seed of strange quark matter, star combusts from inside out, ending with hybrid star
• Neutrino signal
�11
image:universetoday.com
Strange star formation• Possible neutrino fluxes for strange
star formation (Drago & Pagliara 2015)
• Occurs over ~20-120 s total, depending on aQ*max
• aQ*max - “the maximum value of the strangeness unbalance at the interface between hadronic matter and quark matter in order for the process of conversion to still proceed because of its exothermicity”
�12
Drago & Pagliara, Physical Review C, 92, 045801 (2015)
Events in argon• Digitized plot, ran through
SNOwGLoBES • No flavor or energy spectrum
info yet - Kate contacted authors • Assumed equal flavor
distribution and T = 5 MeV for now
• Total events given on right (for 10 kpc)
• DAQ buffer length determines how many of these are recorded
• Pointing resolution as a function of number of events to be studied in the future
�13
Time (s)0 20 40 60 80 100 120
Even
ts (/
0.25
s)
0
100
200
300
400
500
aQmax = 0.3
aQmax = 0.5
aQmax = 0.7
Core combustion
Events vs Time
Pointing latency• Pointing information is useful at any time, but
the sooner the better • Time we want pointing information depends
on many factors: when other detectors provide info, when the SN is visible via light, when the astronomers find it • Tens of seconds to hours between neutrino
burst and optically observable signal - want to warn astronomers ahead of this
• Even if it’s been found, pointing information still useful as a confirmation of the origin of the burst and to see DUNE's pointing capabilities
�14
Thank you
�15