BEGIN:VCALENDAR
VERSION:2.0
CALSCALE:GREGORIAN
PRODID:UW-Madison-Physics-Events
BEGIN:VEVENT
SEQUENCE:2
UID:UW-Physics-Event-4788
DTSTART:20180426T210000Z
DURATION:PT1H0M0S
DTSTAMP:20260415T113356Z
LAST-MODIFIED:20180418T171602Z
LOCATION:5280 Chamberlin Hall
SUMMARY:Can a tool developed for hurricane prediction be taken to pred
 ict neutrino flavor evolution?\, NPAC (Nuclear/Particle/Astro/Cosmo) F
 orum\, Eve Armstrong\, University of Pennsylvania
DESCRIPTION:We assess the utility of an optimization-based data assimi
 lation (D.A.)<br>
\ntechnique for treating the problem of nonlinear ne
 utrino flavor<br>
\ntransformation in core-collapse supernovae.  D.A. 
 was invented for<br>
\nnumerical weather prediction\, and it shares so
 me features of machine<br>
\nlearning for the purposes of predictive p
 ower.  Within the D.A. framework\,<br>
\none uses measurements obtaine
 d from a physical system to estimate the<br>
\nstate variable evolutio
 n and parameter values of the associated model.<br>
\nFormulated as an
  optimization procedure\, D.A. can offer an<br>
\nintegration-blind ap
 proach to predicting model evolution\, which offers an<br>
\nadvantage
  for models that thwart solution via traditional numerical<br>
\ninteg
 ration techniques. Further\, D.A. performs most optimally for models<b
 r>
\nwhose equations of motion are nonlinearly coupled. In this explor
 atory<br>
\nwork\, we consider a simple steady-state model with two mo
 no-energetic<br>
\nneutrino beams coherently interacting with each oth
 er and a background<br>
\nmedium. As this model can be solved via nume
 rical integration\, we have an<br>
\nindependent consistency check for
  D.A. solutions.<br>
\n<br>
\nWe find that the procedure can capture k
 ey features of flavor evolution<br>
\nover the entire trajectory\, eve
 n given measurements of neutrino flavor<br>
\nonly at the endpoint\, a
 nd with an assumed known initial flavor<br>
\ndistribution. Further\, 
 the procedure permits an examination of the<br>
\nsensitivity of flavo
 r evolution to estimates of unknown model parameters\,<br>
\nlocates d
 egeneracies in parameter space\, and can identify the specific<br>
\nm
 easurements required to break those degeneracies.
URL:https://www.physics.wisc.edu/events/?id=4788
END:VEVENT
END:VCALENDAR