BEGIN:VCALENDAR
VERSION:2.0
PRODID:researchseminars.org
CALSCALE:GREGORIAN
X-WR-CALNAME:researchseminars.org
BEGIN:VEVENT
SUMMARY:Charles Collot (Courant institute of mathematical Sciences)
DTSTART:20200423T140000Z
DTEND:20200423T145000Z
DTSTAMP:20260404T143401Z
UID:IMS/5
DESCRIPTION:Title: <a href="https://stable.researchseminars.org/talk/IMS/5
 /">On the derivation of the homogeneous kinetic wave equation</a>\nby Char
 les Collot (Courant institute of mathematical Sciences) as part of PDE sem
 inar via Zoom\n\n\nAbstract\nThe kinetic wave equation arises in many phys
 ical situations: the description of small random surface waves\, or out of
  equilibria dynamics for large quantum systems for example. In this talk w
 e are interested in its derivation as an effective equation from the nonli
 near Schrodinger equation (NLS) for the microscopic description of a syste
 m. More precisely\, we will consider (NLS) in a weakly nonlinear regime on
  the torus in any dimension greater than two\, and for highly oscillatory 
 random Gaussian fields as initial data. A conjecture in statistical physic
 s is that there exists a kinetic time scale on which\, statistically\, the
  Fourier modes evolve according to the kinetic wave equation. We prove thi
 s conjecture up to an arbitrarily small polynomial loss in a particular re
 gime\, and obtain a more restricted time scale in other regimes. The main 
 difficulty\, that I will comment on\, is that one needs to identify the le
 ading order statistically observable nonlinear effects. This means underst
 anding correlation between Fourier modes\, and relating randomness with st
 ability and local well-posedness. The key idea of the analysis is the use 
 of Feynman interaction diagrams to understand the solution as colliding li
 near waves. We use this framework to construct an approximate solution as 
 a truncated series expansion\, and use in addition random matrices tools t
 o obtain its nonlinear stability in Bourgain spaces. This is joint work wi
 th P. Germain.\n
LOCATION:https://stable.researchseminars.org/talk/IMS/5/
END:VEVENT
END:VCALENDAR
