Lefschetz Center for Dynamical Systems Seminar
Brown University Center for Statistical Sciences Seminar
Department of Ecology & Evolutionary Biology | |
1st Floor Conference Room |
Abstract: The fixation of a beneficial allele in a population leaves a well-characterized signature in patterns of nucleotide variation at linked sites in the genome. This signature can be used to estimate the time since fixation from patterns of polymorphism in extant individuals. I introduce a method to assess the support in polymorphism data for a recent episode of directional positive selection and to estimate the time since fixation. When the approach is applied to the gene tb1 in maize, the data strongly support the recent fixation of a favored allele, consistent with what is known about the importance of tb1 in the domestication process of the species. I also present applications to other genes in the genus zea, about which less is known independently, as well as to FOXP2, a gene involved in speech and language production in humans.
The Fluids, Thermal And Chemical Processes Group
Of
The Division Of Engineering
And
Center for Fluid Mechanics
Abstract: The Rayleigh-Taylor instability (RTI) develops at the interface between two fluids, when a light fluid accelerates a heavy fluid. The turbulent mixing produced by the instabilities is of extreme importance in inertial confinement fusion, astrophysics, and many other applications. To obtain a reliable description of the mixing process, the evolution of the large-scale coherent structure, the dynamics of small-scale structures, and the cascades of energy should be understood. The dynamics of RTI is governed by a system of conservation laws, which are nonlinear partial differential equations with the initial conditions and the boundary conditions at the fluid interface. Singular aspects of the interface evolution cause theoretical difficulties and preclude elementary methods of solution.
We suggest a new theoretical approach to the problem. It is based on group theory, applies the separation of scales in the governing equations, and accounts for the non-local properties of the flow that has singularities. Asymptotic solutions describing the large-scale coherent dynamics are found. The analysis yields a non-trivial dependence of the interface evolution on the density ratio and the acceleration history, and determines the key properties of the spatial RT flow. Our results show that a balance between the inverse and direct cascades is required to keep an isotropy of the coherent structures. The theory explains existing observations, predicts new universal properties of the interface dynamics, and identifies sensitive diagnostic parameters for future observation.
Distinguished Lecture Series
Reception at 4:00 p.m. in Kassar House Lobby |
Distinguished Lecture Series
Reception at 4:00 p.m. in Kassar House Lobby |
Distinguished Lecture Series
Reception at 4:00 p.m. in Kassar House Lobby |
Scientific Computing Seminar
Abstract: I will discuss the formulation and the properties of the moment implicit particle in cell (PIC) method developed at Los Alamos National Laboratory.
The talk will be divided into three parts.
First, I will discuss the challenges of multiple scale problems in plasma physics. Plasmas host a variety of processes, often some are of more interest than others. Often the processes of interest are on long space and time scales. The implicit approach is an excellent way to handle this situation. It focuses on the long scales of interest, with proportionate resolution, without needing to resolve smaller scales accurately. The method implicitly averages over the smaller and faster scales. I will discuss the general properties of the implicit method.
Second, I will discuss how the implicit moment method is designed and turned into a computer code. I will summarize the actual formulation we currently use in our CELESTE3D code. I will spend a little more time discussing the most recent advances in this area: the formulation of the Maxwell's equations and the boundary conditions for them.
Lastly, I will discuss practical issues and provide additional tools for implicit PIC, useful for multiple scale problems. I will describe the methods for grid adaptation and particle adaptation that we have recently developed.
PDE Seminar
Department of Mathematics Colloquium
<--- 2004 Index