A new deliverable (D1.8) has been published that complements the list of dwarfs inherited from the original ESCAPE project with novel concepts that include parallel implementation of high order discontinuous finite elements (discontinuous Galerkin, henceforth DG) in spherical geometry, advanced linear solvers to support semi-implicit time discretizations, fault-tolerant versions of linear solvers already employed for the finite volume elliptic solver dwarf and radiation parameterizations based on artificial neural networks (NN).

Each of these dwarfs aims at extending the range of advanced mathematical tools applied in numerical weather prediction in a different direction. In addition, this deliverable provides WP3 with a sample of the techniques and codes representing the core of the final WP1 deliverables, so as to start work on benchmarking the performance of the present implementation of these novel approaches.