The HADES seminar on Tuesday, April 26 will be at 3:30 pm in Room 740.
Speaker: Pierre Germain
Abstract: On a Riemannian manifold, consider the spectral projector on a thin spectral band $[\lambda , \lambda + \delta]$ for the Laplace-Beltrami operator. What is its operator norm from $L^2$ to $L^q$? Or, to put it in semiclassical terms, how large can the $L^p$ norm of a quasimode normalized in $L^2$ be? This is a fascinating problem, which is closely related to a number of fundamental analytic questions. I will try and describe what is known, and some recent progress that have been made. There will be some overlap with my talk at the Analysis seminar, but not much.
The HADES seminar on Tuesday, April 19 will be at 3:30 pm in Room 740.
Speaker: Mitchell Taylor
Abstract: We will discuss some aspects of the nonlinear geometry of function spaces, and how “free” constructions can shed light on such problems. The goal will be to show that certain nonlinear properties of function spaces are actually equivalent to linear ones.
The HADES seminar on Tuesday, April 12 will be at 3:30 pm in Room 740.
Speaker: Gunther Uhlmann
Abstract: Calderón’s problem (also called electrical impedance tomography) asks
the question of whether one can determine the electrical conductivity
of a medium by making voltage and current measurements at the
boundary. I will give a survey of some of the progress made on this
problem, including the more recent progress on solving similar
problems for nonlinear equations and nonlocal operators.
The HADES seminar on Tuesday, April 5 will be at 3:30 pm in Room 740.
Speaker: Nima Moini
Abstract: In this talk, I will sketch a new approach to the study of kinetic equations solely under the assumption of conservation laws. The new idea is based on an uncertainty principle, the introduction of blind cones with respect to an observer and the Galilean invariance of different inertial frames of reference. In fact, as the uncertainty inevitably increases with time, particles will move away in an asymptotically radial manner from any fixed observer thereby establishing a new notion of dispersion. The generality of this approach reveals a mathematical relationship between the Landau and Boltzmann equations in the context of “the grazing collisions”, which until now was solely phenomenological. Moreover, I will discuss a new scattering theory for the kinetic equations and demonstrate its utility in the case of the Boltzmann equation for hard spheres. The new framework improves upon the existing results by proving the asymptotic completeness of the solutions of the Boltzmann equation near an equilibrium in the $L^\infty$ setting. In particular, for any solution to the transport equation, there are arbitrarily close in $L^\infty$ norm, scattered solutions of the Boltzmann equation, this implies that solutions of the Boltzmann equation defined over the whole space will not converge to the state of thermodynamic equilibrium.
The HADES seminar on Tuesday, March 29 will be at 3:30 pm in Room 740.
Speaker: Yuchen Mao
Abstract: Unlike many other equations, initial data for the Einstein equation have to solve the constraint equations, which makes it an interesting problem to construct asymptotically flat localized initial data. Carlotto and Scheon proved the existence of gluing construction of such initial data supported in a cone through a functional analytic approach. We give a simpler proof by explicitly constructing a solution with conic support that achieves the optimal decay conjectured by Carlotto, and lower regularity. Another conjecture made by Carlotto is whether we can construct initial data localized in a smaller region without violating the positive mass theorem. As an application of our solution operator, we prove this is possible for the case of a degenerate sector. This is a joint work with Zhongkai Tao.
The HADES seminar on Tuesday, March 15 will be at 3:30 pm in Room 740.
Speaker: Jeffrey Galkowski
Abstract: We discuss the typical behavior of two important quantities on compact Riemannian manifolds: the number of primitive closed geodesics of a certain length and the error in the Weyl law. For Baire generic metrics, the qualitative behavior of both of these quantities has been well understood since the 1970’s and 1980’s. Nevertheless, their quantitative behavior for typical manifolds has remained mysterious. In fact, only recently, Contreras proved an exponential lower bound for the number of closed geodesics on a Baire generic manifold. Until now, this was the only quantitative estimate on either the number of geodesics for typical metrics, and no such estimate existed for the remainder in the Weyl law. In this talk, we give stretched exponential upper bounds on the number of primitive closed geodesics for typical metrics. Furthermore, using recent results on the remainder in the Weyl law, we will use our dynamical estimates to show that logarithmic improvements in the remainder in the Weyl law hold for typical manifolds. The notion of typicality used in this talk will be a new analog of full Lebesgue measure in infinite dimensions called predominance.
Given recent results of myself and Canzani on the Weyl law, all of these estimates are reduced to a study of the closed geodesics on a typical manifold. We will recall these results on the Weyl law and discuss the ideas used to understand closed geodesics on typical manifolds.
Based on joint work with Y. Canzani.
The HADES seminar on Tuesday, March 8 will be at 3:30 pm in Room 740.
Speaker: Arthur Touati
Abstract: In this talk, I will present recent work on high-frequency solutions
to the Einstein vacuum equations. From a physical point of view, these solutions
model high-frequency gravitational waves and describe how waves travel on a fixed
background metric. There are also interested when studying the Burnett conjecture,
which addresses the lack of compactness of the family of vacuum spacetimes. These
high-frequency spacetimes are singular and require to work under the regime of
well-posedness for the Einstein vacuum equations. I will review the literature on
the subject and then show how one can construct them in generalised wave gauge
by defining high-frequency ansatz.
The HADES seminar on Tuesday, February 22 will be at 3:30 pm in Room 740.
Speaker: Yonah Borns-Weil
Abstract: Quantum dynamics is concerned with quantum analogues of classical dynamical systems. A common situation is scattering theory, in which the Hamiltonian dynamics of scattered particles are replaced by wavefunctions obeying the Schrödinger equation. When time is discretized, the analogues are open quantum maps, which are Fourier integral operators arising from phase space diffeomorphisms that are then “opened” by sending some regions to “infinity.” In this talk, we analyze a simple open quantum map, based on the classical Arnol’d cat map. We shall show using the method of Grushin problems that the spectrum has a very simple form in the semiclassical regime as h approaches 0. Emphasis will be given to motivation and interpretations of the result.
The HADES seminar on Tuesday, February 15 will be at 3:30 pm in Room 740.
Speaker: Jeffrey Kuan
Abstract: In this talk, we present a well-posedness result for a stochastic fluid-structure interaction model. We study a fully coupled stochastic fluid-structure interaction problem, with linear coupling between Stokes flow and an elastic structure modeled by the wave equation, and stochastic noise in time acting on the structure. Such stochasticity is of interest in applications of fluid-structure interaction, in which there is random noise present which may affect the dynamics and statistics of the full system. We construct a solution by using a new splitting method for stochastic fluid-structure interaction, and probabilistic methods. To the best of our knowledge, this is the first result on well-posedness for fully coupled stochastic fluid-structure interaction. This is joint work with Sunčica Čanić (UC Berkeley).
The HADES seminar on Tuesday, February 8 will be at 3:30 pm in Room 740.
Speaker: Izak Oltman
Abstract: When computing eigenvalues of finite-rank non-self-adjoint operators, significant numerical inaccuracies often occur when the rank of the operator is sufficiently large. I show the spectrum of Toeplitz operators, with a random perturbation added, satisfy a Weyl law with probability close to one. I will begin with numerical animations, demonstrating this result for quantizations of the torus (a result proven by Martin Vogel in 2020). Then give a brief introduction to Toeplitz operator (quantizations of functions on Kahler Manifolds). And finally outline the main parts of the proof, which involve constructing an `exotic calculus’ of symbols on a Kahler manifold.