Robust error estimates in weak norms for advection dominated transport problems with rough data (Q2930098)

This paper is concerned with a stabilized finite element method for a transient convection-diffusion equation with a velocity vector field with multiscale character and rough data, and aims at taking a first step towards an understanding of what transport problems are computable in the high Péclet regime, beyond the standard assumption of smooth data. The author considers the case that the velocity field has two scales, a coarse scale with slow spatial variation, which is responsible for advective transport and a fine scale with small amplitude that contributes to the mixing. For this problem, the author derives a posteriori and a priori error estimates of a finite element method with symmetric stabilization by using the multiscale decomposition of the advective velocity to improve stability. All derived estimates are independent both of the Péclet number and of the regularity of the exact solution, and are robust with respect to the Péclet number in the sense that one also controls the Sobolev norms of the exact solution in the error constant. The estimates allow for low regularity data and multiscale advection, that may have strong spatial variation on the fine scale under a special scale separation condition.