Coarse graining of an individual-based plant model

The evolution and structuring of plant communities is governed by local plant-to-plant interaction. It has been shown that the individual variability of the organisms and the spatial neighbor constellation is essential for plant system dynamics. Therefore, several spatially explicit individual-based model approaches for plant communities have been developed in the last decades. Among these the field of neighborhood approach is promising as it describes the system at an intermediate level of complexity. A present limitation of this approach, and many other individualbased models, is that it can only be studied by simulation. In general simulations of large individual-based models are computational demanding and reveal only limited information on the long-term dynamics of the system. Here, we report ongoing work that aims to investigate the dynamics of field of neighborhood by a numerical coarse-graining technique. This so-called “equation-free” method, uses short burst of individual-based simulation to extract information that is necessary to investigate the system on the level of plant distributions. This approach allows us to investigate the individual-based model as if macroscopic equations of motion were available, providing more information in shorter time. We emphasize that our approach is well suited for implementation on CELL or GPU systems, as it can be parallelized with high efficiency and most time-critical operations can be performed on the rendering pipeline.