Table 4 Convergence of various metrics describing the solution to the new subduction zone benchmark as a function of degrees of freedom in the heat equation \({T}_\text {ndof}\). The employed meshes have grid refinement in the wedge above and near the coupling point. The factor \(f_m\) is representative of the element size near the coupling point. TH Taylor–Hood, PF penalty function method. P2P1P2 indicates a discretization that has quadratic shape functions (P2) for velocity and temperature and linear shape functions for pressure (P1). P2P2 is for velocity and temperature only because pressure is eliminated from the Stokes equation in the penalty function method (Cuvelier et al. 1986). In this case, \(z_\text {io}\) = 139

From: An introductory review of the thermal structure of subduction zones: II—numerical approach and validation