An optimal variance estimate in stochastic homogenization of discrete elliptic equations

Limit theorems in probability theory (60F99) Homogenization in context of PDEs; PDEs in media with periodic structure (35B27) Difference operators (39A70) Random operators and equations (aspects of stochastic analysis) (60H25)

Abstract: We consider a discrete elliptic equation on the

d

-dimensional lattice

m a t h b b Z^{d}

with random coefficients

A

of the simplest type: they are identically distributed and independent from edge to edge. On scales large w.r.t. the lattice spacing (i.e., unity), the solution operator is known to behave like the solution operator of a (continuous) elliptic equation with constant deterministic coefficients. This symmetric ``homogenized matrix $A_{m a t h r m h o m} = a_{m a t h r m h o m} o p e r a t o r n a m e I d$ is characterized by $x i c d o t A_{m a t h r m h o m} x i = l a n g l e (x i + a b l a p h i) c d o t A (x i + a b l a p h i) a n g l e$ for any direction $x i i n m a t h b b R^{d}$ , where the random field $p h i$ (the ``corrector) is the unique solution of

- a b l a^{*} c d o t A (x i + a b l a p h i) = 0

such that

p h i (0) = 0

,

a b l a p h i

is stationary and

l a n g l e a b l a p h i a n g l e = 0

,

l a n g l e c d o t a n g l e

denoting the ensemble average (or expectation). It is known (``by ergodicity) that the above ensemble average of the energy density $m a t h c a l E = (x i + a b l a p h i) c d o t A (x i + a b l a p h i)$ , which is a stationary random field, can be recovered by a system average. We quantify this by proving that the variance of a spatial average of $m a t h c a l E$ on length scales $L$ satisfies the optimal estimate, that is, $o p e r a t o r n a m e v a r [s u m m a t h c a l E e t a_{L}] l e s s s i m L^{- d}$ , where the averaging function [i.e., $s u m e t a_{L} = 1$ , $o p e r a t o r n a m e s u p p (e t a_{L}) s u b s e t | x | l e L$ ] has to be smooth in the sense that $| a b l a e t a_{L} | l e s s s i m L^{- 1 - d}$ . In two space dimensions (i.e., $d = 2$ ), there is a logarithmic correction. This estimate is optimal since it shows that smooth averages of the energy density $m a t h c a l E$ decay in $L$ as if $m a t h c a l E$ would be independent from edge to edge (which it is not for $d > 1$ ). This result is of practical significance, since it allows to estimate the dominant error when numerically computing $a_{m a t h r m h o m}$ .

Recommendations

Cites Work

Cited In (only showing first 100 items - show all)

This page was built for publication: An optimal variance estimate in stochastic homogenization of discrete elliptic equations

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q533739)