Special Session 48: Recent Advances in Nonlinear PDEs and Inverse Problems

An Inverse Problem for a Nonlinear Monodomain System

Andrea Aspri
Universit\`a degli Studi di Milano Statale
Italy
Co-Author(s):    Elena Beretta, Elisa Francini, Dario Pierotti, Sergio Vessella
Abstract:
In this talk, I will discuss an inverse problem for the nonlinear monodomain system. After briefly recalling the well-posedness of the direct problem for a reaction--diffusion equation coupled with an ordinary differential equation in the presence of perfectly insulating regions, I will focus on the inverse question of determining such regions from partial boundary measurements. Under suitable assumptions on the conductivity tensor, the nonlinear ionic terms, and the initial activation, I will present a uniqueness result showing that a single partial measurement of the transmembrane potential is sufficient to identify the geometry and location of the insulating inclusion. The analysis relies on a reformulation of the problem as a parabolic integro-differential equation and combines regularity estimates, a three-cylinder inequality, and unique continuation arguments. The result provides a rigorous uniqueness theorem for a class of inverse problems associated with nonlinear coupled systems of PDE--ODE type.

Simultaneous Stable Determination of Quasilinear terms for Parabolic equations

Zi Him Jason Choy
Chinese University of Hong Kong
Hong Kong
Co-Author(s):    
Abstract:
We consider the inverse problem of simultaneously recovering two classes of quasilinear terms appearing in a parabolic equation from boundary measurements. It is motivated by several industrial and scientific applications, including problems of heat conduction and population dynamics, and we study the issue of stability. More precisely, we derive simultaneous Lipschitz and H\older stability estimates for two separate classes of quasilinear terms. The analysis combines different arguments including the linearization technique with a novel construction of singular solutions and properties of solutions of parabolic equations with nonsmooth boundary conditions. This is a joint work with Dr. Yavar Kian.

Stable Determination and Reconstruction of a Semilinear Term in a Parabolic Equation

Maolin Deng
The Chinese University of Hong Kong
Hong Kong
Co-Author(s):    Jason Choy, Maolin Deng, Bangti Jin and Yavar Kian
Abstract:
In this work, we investigate the inverse problem of determining a semilinear term appearing in a nonlinear parabolic equation from the measurement of the conormal derivative on the boundary. We first establish a uniqueness result, showing that the nonlinear term is uniquely determined provided the difference between two candidates does not change sign infinitely many times. Subsequently, we derive H\older-type stability estimates for the reconstruction, where the stability exponent depends explicitly on the number of sign changes of the difference between two admissible nonlinear terms. Improved stability estimates are obtained under higher regularity assumptions on the nonlinear term. Finally, we propose a numerical reconstruction algorithm based on an iterative scheme and validate its performance through various numerical experiments with noisy data.

Nonlocal Partial Regularity

Cristiana Filippis
University of Parma
Italy
Co-Author(s):    Giuseppe Mingione, Simon Nowak
Abstract:
The failure of De Giorgi-Nash theory in vectorial problems opens to the study of partial regularity, namely smoothness of solutions outside a negligible, singular set. This is a classical phenomenon for systems and harmonic maps. I'll give an account of recent advances on nonlinear integro-differential systems, with emphasis on the structure and size of the singular set of solutions. From recent, joint work with Giuseppe Mingione (Parma) and Simon Nowak (Bielefeld).

Inverse Problems for Anisotropic Conductivities in Semilinear Models

Elisa Francini
Universit\`a di Firenze
Italy
Co-Author(s):    Elena Beretta, Dario Pierotti, Eva Sincich
Abstract:
In this talk, we discuss an inverse boundary value problem arising in a semilinear elliptic model motivated by cardiac electrophysiology. The model describes the stationary behavior of the transmembrane potential under pacing and involves an anisotropic conductivity tensor together with a nonlinear ionic response. We address the problem of recovering a piecewise constant anisotropic conductivity from boundary measurements encoded by the Neumann-to-Dirichlet map. Assuming that the nonlinear coefficient and the geometry of the inclusion are known, we present a uniqueness result showing that the conductivity is uniquely determined by the boundary data. The analysis is based on a first-order linearization of the nonlinear map around a nontrivial background state, leading to a linear elliptic problem with a strictly positive zeroth-order term. This reduction allows us to combine boundary determination techniques for anisotropic conductivities with unique continuation arguments to achieve identifiability.

Uniqueness for the inverse spectral modified transmission eigenvalue problem for a piecewise continuous spherically symmetric refractive index

DROSSOS GINTIDES
NATIONAL TECHNICAL UNIVERSITY OF ATHENS
Greece
Co-Author(s):    KYRIAKOS STRATOURAS
Abstract:
We consider the uniqueness question for the inverse modified transmission eigenvalue problem from modified transmission eigenvalues. We assume that the refractive index is spherically symmetric and piecewise continuous. The corresponding direct spectral transmission problem refers to a system of two Helmholtz type equations in the same domain where the spectral parameter appears in the one equation, which corresponds to an artificial metamaterial. In addition, the two corresponding fields have equal Cauchy data on the boundary. By using separation of variables to describe the direct problem, we reconstruct the Dirichlet-to-Neumann map for the field that satisfies the Helmholtz equation with the refractive index n(r). By utilizing well-known results from the Borg-Levinson theory for Schrodingers equation we prove the uniqueness of the recovery of n(r). We explore some special properties of the spectrum and present simple examples.

An inverse source problem for a quasilinear elliptic equation

Shubham Jaiswal
University of Jy\\"askyl\\"a, Finland.
Finland
Co-Author(s):    Tony Liimatainen
Abstract:
In this talk, we consider the inverse source problems for quasilinear elliptic equations of the form \[ \left\{ \begin{array}{ll} \nabla \cdot (\gamma(x,u,\nabla u) \nabla u) = F & \text{in } \Omega, \ u = f & \text{on } \partial\Omega, \end{array} \right. \] where $\Omega \subset \mathbb{R}^n$, $n \geq 2$, is a simply connected bounded domain. We consider the specific nonlinearity $\gamma(x,u,\nabla u) = \sigma(x) + q(x) u$, with $q$ assumed to be known. By exploiting the nonlinearity to break the gauge invariance of the problem, we establish unique recovery of both $\sigma$ and $F$ from the associated Dirichlet-to-Neumann (DN) map under the structural conditions $q$ and $\nabla(\sigma/q)$ are nowhere vanishing in $\overline\Omega$. In the absence of these conditions, in particular in the linear case, we demonstrate that the inverse problem admits a gauge obstructing the uniqueness. We use higher order linearizations to obtain a complicated coupled system for the unknowns. The complexity of this system arises in part from the gauge freedom of the linearized equation, which is new in this context. We solve the system by constructing suitable complex geometric optics solutions and applying the unique continuation principle for nonlinear elliptic systems. We anticipate that the solution method developed here will prove useful in other inverse problems as well. This is a joint work with Tony Liimatainen.

Recovery of nonlinear material parameters in a quasilinear Lam\`e system

David Johansson
Aarhus University
Denmark
Co-Author(s):    Yavar Kian
Abstract:
I will discuss the identifiability of a nonlinear tensor in a quasilinear elliptic system, which is inspired by elasticity theory. As measurements we use point measurements at finitely many boundary points of the conormal derivative (or traction) of solutions generated by Dirichlet data (or boundary displacement) in the space of first order polynomials. By a linearization argument, we show that this data uniquely determines a nonlinear but space-independent tensor. Measurements in a single point is sufficient in the case of an isotropic tensor, while several points are used in the anisotropic case.

Extension of the Boundary Control Method to Elliptic and Parabolic Problems, with Applications to the Calder\\`{o}n Problem

Dimitra Kyriakopoulou
Biomedical Research Foundation Academy of Athens
Greece
Co-Author(s):    
Abstract:
We extend the Boundary Control method beyond the hyperbolic setting to certain elliptic and parabolic inverse boundary value problems. The analytic framework is based on Schulze's edge calculus, with the parabolic case reduced to an elliptic problem. The reconstruction step is formulated through Boundary Control algebras generated by boundary-induced solutions. Under suitable separation assumptions, these algebras recover the underlying manifold, or in the parabolic case a compact spacetime strip, as a Gelfand spectrum. We then discuss applications to the Calder\`{o}n problem. In dimension \(n \ge 3\), this gives a real-analytic conductivity-metric result, while in dimension two it yields a Boundary Control approach based on holomorphic trace algebras and recovery up to the natural conformal gauge. We also outline an ongoing attempt to treat the higher-dimensional smooth anisotropic case within the same framework.

The minimal surface transform

Tony Liimatainen
University of Jyv\"askyl\"a
Finland
Co-Author(s):    
Abstract:
Minimal surfaces are fundamental objects in geometry, defined as critical points of the area functional. They also play a central role in the AdS/CFT correspondence in theoretical physics, where the Ryu-Takayanagi formula equates entanglement entropy on the boundary to the area of a minimal surface in the bulk. Bulk reconstruction---a central problem in physics---is precisely the inverse problem of recovering geometric information from areas of minimal surfaces. I will first briefly recall results on this inverse problem based on a PDE approach for the minimal surface equation. These results rely on methods for the Calder\`{o}n problem and are mostly restricted to two dimensional minimal surfaces and one can at best expect logarithmic stability. I will then introduce an integral geometry approach, viewing the first variation of area as a generalized Radon transform---the minimal surface transform. Interpreting this transform as a Fourier Integral Operator yields injectivity results for analytic metrics in dimensions three and higher. This framework achieves two major advances: it extends to higher-dimensional minimal surfaces and bulks, and it provides H\{o}lder stability rather than logarithmic. These mathematical developments have improved upon the best known results in the physics literature, representing a rare instance where techniques from mathematics advance theoretical physics.

An inverse problem for the prescribed mean curvature equation

Janne Nurminen
University of Jyv{\"a}skyl{\"a}
Finland
Co-Author(s):    Tony Liimatainen
Abstract:
In this talk I will formulate an inverse source problem for the prescribed mean curvature equation (PMC) \begin{equation*} \nabla\cdot\left[\frac{\nabla u}{\left(1+|\nabla u|^2\right)^{1/2}}\right] = H(x)\quad\text{in }\Omega \end{equation*} for a smooth bounded set $\Omega\subset R^2$. The question is if from measurements done on the boundary $\partial\Omega$ one can determine the mean curvature $H$ in $\Omega$. The talk is based on joint work with Tony Liimatainen (https://arxiv.org/abs/2509.22078) and we show that it is indeed possible to recover $H$. The proof relies on the higher order linearization method and asymptotical analysis in an integral identity using complex geometric optics solutions from the work of Guillarmou and Tzou in 2011.

Interior decay of solutions to elliptic equations and applications to inverse problems

Luca Rondi
Universit\`a degli Studi di Pavia
Italy
Co-Author(s):    
Abstract:
We investigate the decay in the interior of solutions to elliptic equations with respect to the boundary data. In particular, the decay rate is linked to the so-called frequency of the boundary datum and is in general stronger when coefficients are smoother. We show applications of these decay results to the electrical impedance tomography. A key ingredient in the proof is to study the distance function from the boundary for a Riemannian manifold. We show that, up to a conformal change of the metric, it coincides with the distance in the Euclidean case, thus inheriting its regularity properties.

Inverse problems for quasi-linear elliptic systems modeling electrolysers

Matteo Santacesaria
University of Genoa
Italy
Co-Author(s):    G.S. Alberti, W. Gerner
Abstract:
Optimizing electrolyser cells for green hydrogen production requires a precise understanding of their internal electrochemical processes. We model these dynamics using a coupled system of quasi-linear elliptic PDEs and investigate the inverse problem of reconstructing non-linear diffusion coefficients and electric potential relations. In this talk, I will show that boundary measurements alone are insufficient for unique reconstruction. By generalizing a known linearization technique to systems with non-local nonlinearities, we prove that combining boundary and interior measurements resolves this issue. I will outline the mathematical framework and explain why interior data is the essential key to freezing the coefficients and successfully solving the inverse problem.

Inverse scattering problems for nonlinear wave equations on Lorentzian manifolds

Teemu Tyni
University of Oulu
Finland
Co-Author(s):    Spyros Alexakis, Hiroshi Isozaki, Matti Lassas
Abstract:
We discuss some recent results on inverse scattering problems for semi-linear wave equations. The inverse scattering problem is formulated on a Lorentzian manifold equipped with a Minkowski type infinity. We show that a scattering functional, which roughly speaking maps measurements of solutions of a semi-linear wave equation at the past infinity to the future infinity, determines the manifold, the conformal class of the metric, and the nonlinear potential function up to a gauge. The main tools we employ are a Penrose-type conformal compactification of the Lorentzian manifold, reduction of the scattering problem to the study of the source-to-solution operator, and the use of higher order linearization method to exploit the nonlinearity of the wave equation. This is a joint work with S. Alexakis, H. Isozaki, and M. Lassas.

Boundary regularity for nonlocal equations

Marvin Weidner
University of Bonn
Germany
Co-Author(s):    Serena Dipierro, Xavier Ros-Oton, Enrico Valdinoci
Abstract:
The boundary behavior for solutions to nonlocal equations with exterior Dirichlet boundary conditions has been extensively studied in recent years and it is well known that, in general, s-harmonic functions are not better than $C^s$. In contrast, the Neumann problem for nonlocal equations has received much less attention, and the optimal boundary regularity of solutions remains unknown. In this talk, I will present recent progress on this question, based on a new classification result for solutions to general nonlocal equations in 1D. This is joint work with Serena Dipierro, Xavier Ros-Oton, and Enrico Valdinoci.

Generic regularity of the free boundary in the Alt-Caffarelli-Phillips problem

Hui Yu
National University of Singapore
Singapore
Co-Author(s):    Xavier Fernandez-Real
Abstract:
For generic boundary data, we discuss an improvement of the estimate of the singular dimension in the Alt-Caffarelli-Phillips problem.

Wave scattering by close-to-touching inclusions

Sanghyeon Yu
Korea University
Korea
Co-Author(s):    
Abstract:
Concentrating light at the nanoscale is one of the greatest challenges in optics. Using conventional optical devices, light cannot be focused into a spot smaller than a few hundred nanometers due to the diffraction limit. Nano-optics overcomes this fundamental difficulty by using optical nanostructures with close-to-touching inclusions. In such structures, light can be strongly localized in the narrow gap regions between the inclusions. In this talk, we discuss how to characterize the singular behavior of waves scattered by close-to-touching inclusions, and its implication for the inverse design of nano-optical devices.