A real‐space cellular automaton laboratory

Geomorphic investigations may benefit from computer modelling approaches that rely entirely on self‐organization principles. In the vast majority of numerical models, instead, points in space are characterized by a variety of physical variables (e.g. sediment transport rate, velocity, temperature) recalculated over time according to some predetermined set of laws. However, there is not always a satisfactory theoretical framework from which we can quantify the overall dynamics of the system. For these reasons, we prefer to concentrate on interaction patterns using a basic cellular automaton modelling framework. Here we present the Real‐Space Cellular Automaton Laboratory (ReSCAL), a powerful and versatile generator of 3D stochastic models. The objective of this software suite, released under a GNU licence, is to develop interdisciplinary research collaboration to investigate the dynamics of complex systems. The models in ReSCAL are essentially constructed from a small number of discrete states distributed on a cellular grid. An elementary cell is a real‐space representation of the physical environment and pairs of nearest‐neighbour cells are called doublets. Each individual physical process is associated with a set of doublet transitions and characteristic transition rates. Using a modular approach, we can simulate and combine a wide range of physical processes. We then describe different ingredients of ReSCAL leading to applications in geomorphology: dune morphodynamics and landscape evolution. We also discuss how ReSCAL can be applied and developed across many disciplines in natural and human sciences. Copyright © 2013 John Wiley & Sons, Ltd.