ASAM is a numerical solver for atmospheric applications at all scales ranging from the globe up to building resolving simulations. ASAM stands for All Scale Atmsopheric Model. The underlying compressible Euler equations in flux form are solved in an Eulerian framework. ASAM was at first designed for CFD simulation around buildings where the obstacles are includes within a Cartesian grid by a cut cell approach. This approach is now extended to other orthogonal grids like the lat-lon grid. Due to the implicit time integration scheme there are no stability problems with small grid cells resulting from cut cells or cells near the poles. ASAM requires that physical processes can be prescribed as a time continuous process with respect the model variables and is not dependent from the time step. ASAM is a developing research code and has a lot of different options to choose numerical methods, number of variables and physical processes. The code is used for testing new numerical approaches but is also for process studies on different scales. Examples are LES simulations of stratiform clouds, vortex generation in street canyons, moist bubble experiments, orographic rainfall and so on. ASAM is a fully parallelized code and is easily portable between different platforms. The following links provide more detailed information about the model content:
- Test Cases
- Grid Generator
- LES running on GPU’s using OpenGL + GLSL (see ASAMgpu)
ASAMgpu is a model for three dimensional atmospheric simulations. It uses GPU’s to provide a maximum performance that enables 3D-LES on architectures ranging from a notebook, an ordinary gaming pc up to one or more high performance MultiGPU Nodes.
If we take a closer look at the project, it splits in two parts. The first one is the GPGPU-Interface, consisting of just two basic submodules for shader and texture handling. This layer can be used in different models, e.g. Shallow Water Equations, Particle dynamics or Ice-Structure-Simulations. It uses OpenGL + GLSL for communication and Shader implementation. This approach features transparency to all parallelization details like blocksizes, cachesizes or threads. It is platform independent, all you need is a graphics environment that supports hardware accelerated OpenGL and Shaders (any DirectX10 compatible hardware)
How to work with ASAMgpu you can find under ASAMgpu Tutorial
Eulerian Framework for atmospheric multiphase flows
Possible simulations with ASAMgpu are shown under Applications.