English

Reduced order modeling for elliptic problems with high contrast diffusion coefficients

Analysis of PDEs 2023-04-24 v1

Abstract

We consider the parametric elliptic PDE div(a(y)u)=f-{\rm div} (a(y)\nabla u)=f on a spatial domain Ω\Omega, with a(y)a(y) a scalar piecewise constant diffusion coefficient taking any positive values y=(y1,,yd)]0,[dy=(y_1, \dots, y_d)\in ]0,\infty[^d on fixed subdomains Ω1,,Ωd\Omega_1,\dots,\Omega_d. This problem is not uniformly elliptic as the contrast κ(y)=maxyjminyj\kappa(y)=\frac{\max y_j}{\min y_j} can be arbitrarily high, contrarily to the Uniform Ellipticity Assumption (UEA) that is commonly made on parametric elliptic PDEs. Based on local polynomial approximations in the yy variable, we construct local and global reduced model spaces VnV_n of moderate dimension nn that approximate uniformly well all solutions u(y)u(y). Since the solution u(y)u(y) blows as y0y\to 0, the solution manifold is not a compact set and does not have finite nn-width. Therefore, our results for approximation by such spaces are formulated in terms of relative H01H^1_0-projection error, that is, after normalization by u(y)H01\|u(y)\|_{H^1_0}. We prove that this relative error decays exponentially with nn, yet exhibiting the curse of dimensionality as the number dd of subdomains grows. We also show similar rates for the Galerkin projection despite the fact that high contrast is well-known to deteriorate the multiplicative constant when applying Cea's lemma. We finally establish uniform estimates in relative error for the state estimation and parameter estimation inverse problems, when yy is unknown and a limited number of linear measurements i(u)\ell_i(u) are observed. A key ingredient in our construction and analysis is the study of the convergence of u(y)u(y) to limit solutions when some of the parameters yjy_j tend to infinity.

Keywords

Cite

@article{arxiv.2304.10971,
  title  = {Reduced order modeling for elliptic problems with high contrast diffusion coefficients},
  author = {Albert Cohen and Matthieu Dolbeault and Agustin Somacal and Wolfgang Dahmen},
  journal= {arXiv preprint arXiv:2304.10971},
  year   = {2023}
}
R2 v1 2026-06-28T10:13:43.614Z