# Global Regularity for the 2D Euler Equations of Ideal Incompressible Fluids - Lecture 2 (After Yudovich)

## Speaker:

## Institution:

## Time:

## Location:

# "Global Regularity for the 2D Euler Equations of Ideal Incompressible Fluids - Lecture 1 (After Yudovich)"

## Speaker:

## Institution:

## Time:

## Location:

# Global Regularity for the 2D Euler Equations of Ideal Incompressible Fluids - Lecture 3

## Speaker:

## Institution:

## Time:

## Location:

# EPDiff, a Nonlinear Wave Eqation with Weak Solution

## Speaker:

## Institution:

## Time:

## Location:

EPDiff is short for ``Euler-Poincar\'e equations on the diffeomorphisms.'' EPDiff first arose as a 1D shallow water wave equation, whose weak solutions are solitons, called ``peakons.'' The initial value problem (IVP) for EPDiff in 2D produces emergent soliton-like weak solutions, supported on curves that evolve in the plane. These curves model internal waves in the ocean. Numerical

simulations show that weak solutions supported on ``peakon filaments'' emerge in the IVP of EPDiff, for any confined smooth initial velocity distribution.

Besides dominating the IVP, the weak solutions of EPDiff have three other interesting dynamical properties:

-- they superpose,

-- they form an invariant manifold and

-- their nonlinear interactions allow them to {\it reconnect} with each other in 2D.

The phenomenon of reconnection seen in the IVP for EPDiff is also observed in oceanic internal waves, as seen from the space shuttle using synthetic aperture radar (SAR). Thus, in accord with their original derivation in 1D, weak solutions of EPDiff provide a simplified 2D description of evolving arrays of interacting internal waves in the Ocean.

Remarkably, the same EPDiff equation {\it also arises in image processing} using template matching, an optimization approach in computational anatomy. Here, for example, a 2D measure-valued EPDiff solution optimally interpolates between the outlines, or ``cartoons," of a planar image and its target image obtained by observations at two times. This is template matching. The nonlinear exchange of momentum seen in the interactions of these ``cartoons" introduces the collison paradigm from soliton dynamics into imaging science. Namely, the optimization problem

for template matching corresponds to an evolutionary problem in which image outlines exchange momentum and may reconnect as their positions evolve. In 3D, measure-valued solutions of EPDiff correspond to suface boundaries in 3D images, representing, say, the sequence of shapes executed in a heartbeat.

The existence of these measure-valued solutions of EPDiff is guaranteed -- for any Sobolev norm, and in any number of spatial dimensions. This holds, because the weak solution ansatz is a momentum map for the (left) action of diffeomorphisms on the measure-valued support set of the solutions.

We review these two contexts for EPDiff and show numerical

and analytical results for its solutions in 1D, 2D and 3D.

(EPDiff -- optimization and evolution -- what an equation!)

# Wave Propagation in Periodic Media

## Speaker:

## Institution:

## Time:

## Location:

# A Tale of Two Topologies: Canonical Forms for Ion Channel Data Analysis

## Speaker:

## Institution:

## Time:

## Location:

In this talk I will introduce the manifest interconductance rank (MIR) form and contrast it to another long-known canonical form used in the

data-driven identification of ion channel gating kinetics: the uncoupled model (UCM). (The UCM has every open state connected to every closed state and vice versa). MIR form has significantly fewer parameters and provides more insight into gating kinetics than the uncoupled model. Beyond the new canonical form the principle results to be presented are

(1)All topologies with interconductance rank=1 and with the same number of open and closed states result in identical steady-state statistics

(2)detailed balance is preserved under transformation to either UCM or MIR forms and

(3) a general detailed balance preserving transformation. These results should facilitate maximum likelihood methods for finding models that best fit a given data set.