Adaptive discontinuous Galerkin methods for non-linear reactive flows (Q266290)

The book under review is devoted to adaptive discontinuous Galerkin (dG) methods (in space and time) for non-stationary semi-linear advection/diffusion reaction equations (ADR) with internal boundary layers. In Chapter 1 ``Introduction'' the author discusses the geological and computational background and the outline of the book. Chapter 2 ``Discontinuous Galerkin methods'' contains a description of interior penalty discontinuous Galerkin (IPG) methods. The author studies the effect of the penalty parameter on spatial accuracy, and gives an IPG formulation for a general steady-state linear ADR equation with upwinding for advection as a key part of the nonlinear elliptic model of the next chapter. Chapter 3 ``Elliptic problems with adaptivity'' introduces the steady-state (elliptic) form of the non-stationary model. Existence and uniqueness results for stationary semi-linear ADR equations are given. Then an adaptive algorithm for semi-linear ADR equations including a robust a posteriori error estimator is described. The solver of the discrete approximation is based on an iteration method with matrix reordering. The numerical studies demonstrate the efficiency and robustness of the algorithm. The comparison with other solvers is given. Chapter 4 ``Parabolic problems with space-time adaptivity'' deals with semi-discrete and fully discrete approximations of the non-stationary model. The a posteriori error bounds are derived, and a robust adaptive algorithm both in time and space to the non-stationary semi-linear problem is proposed. The adaptive algorithm is capable of catching not only the spatial layers, but also the temporal layers by decreasing the time step. Beside the refinement procedure, the algorithm includes a coarsening procedure to optimize the computational work. Chapter 5 ``Conclusion and outline'' rounds up the book. The bibliography consists of 98 literature items. Appendix A ``Matlab tutorial'' contains the description of MATLAB routines using dG methods for solving and simulating steady-state ADR equations in 2D.