On this years international conference, the UCP2019 in Berlin I presented a poster with latest news and developments on the radiative transfer solver for the next generation numerical weather prediciton model, ICON. UCP stands for Understanding Clouds and Precitipation and it was the de-facto wrap up conference for the HDCP2 project which had as its aim the development and usage of high-resolution simulations with ICON in LES mode, i.e. 150m horizontal resolution over germany. The key features of the poster examine

• why we care about 3D radiation: 3D vs. 1D radiation shows to alter the boundary layer dynamics interacting with the land surface, the boundary layer dynamics, the convection and clouds As a result we would like to study these interactions in a more realistic setup and hence started out to develop a 3D radiation scheme for ICON’s wedge mesh
• panel 2 depicts the concept for the new radiation solver. The key concept is the same as for the TenStream solver, i.e. discretize radiation spatially and angularly in a number of streams, determine the transport coefficients in a single wedge via MonteCarlo raytracing, couple all cells of the mesh globally in a huge sparse linear equation system and solve it with iterative solvers from PETSc
• the right panel introduces a MonteCarlo model (RayLi), capable to trace natively in the ICON-Mesh which is used as a benchmark model
• the broad lower panel shows output of the current implementation for various solvers in the TenStream radiative transfer package, once with a 1D \(\delta\)-Eddington Twostream, and once with the new 3D solver
• Lastly the two rightmost scenes which show the same scene but with vacuum boundary conditions which are needed to benchmark the TenStream solution against RayLi simulations