The Cauchy-Dirichlet problem for parabolic equations with VMO coefficients (Q815489)

The paper deals with the \(W^{2,1}_p(Q_T)\)-regularity and solvability of the Cauchy-Dirichlet problem for linear second-order, uniformly parabolic equations of the form \[ {\mathcal L}u\equiv \partial_tu- \sum_{i,j=1}^n a_{ij}(x)u_{x'_ix'_j}+ \sum_{i=1}^n b_i(x)u_{x'_i} +cu=f \] with \(x=(x',t)\) lying in the cylinder \(Q_T=\Omega\times(0,T)\subset \mathbb R^{n+1}.\) The principal coefficients \(a_{ij}\) are supposed to belong to the Sarason class VMO of functions with vanishing mean oscillation, while the lower-order coefficients own suitable Lebesgue integrability. Results on regularity (interior and global) and solvability of Cauchy-Dirichlet's problems for the operator \[ {\mathcal L_0}u\equiv \partial_tu- \sum_{i,j=1}^n a_{ij}(x)u_{x'_ix'_j} \] with \(a_{ij}\in\text{VMO},\) are very well known since more than 12 years, see \textit{M. Bramanti} and \textit{M. C. Cerutti} [Commun. Partial Differ. Equations 18, No. 9--10, 1735--1763 (1993; Zbl 0816.35045)]. Indeed, adding to \({\mathcal L_0}\) low-order terms taken from suitable Lebesgue spaces is a trivial trick which does not produce new results and relies on banal applications of the Hölder inequality. Therefore, the eventual reader of the paper under review will find nothing new in contrast to Bramanti-Cerutti's article.