SELES: A Spatially Explicit Landscape Event Simulator

Landscape structure results from complex interactions between geomorphology, climate, disturbance and succession. SELES is a tool for building spatially explicit simulations to model the role of disturbance in creating and maintaining landscape structure. Models built with SELES are raster-based, semi-Markov, whole landscape models (sensu Baker 1989) which use probabilistic disturbance spread. Each raster cell represents a sub-model, with landscape patterns emerging from the accumulation of cell behaviors. The modeler deenes landscape events (such as disturbance and succession), which determine the landscape dynamics. Each event is characterized by a set of expressions that govern how often the event occurs, how the cell state is aaected by the event, and how the event spreads from one cell to another. These expressions are dependent on one or more cell state variables (vegetation cover, soil type, elevation, slope, aspect, etc.). SELES models can incorporate these state raster layers from a variety of sources and uses PAMAP GIS directly as a spatial database. Starting with a simple model that operates on a neutral landscape (sensu Gardner et al. 1987), the modeler can add levels of complexity incrementally and incorporate parameters from real landscapes to reene hypotheses about landscape dynamics. We describe the event-based simulation engine and how to use it to specify landscape events. We then show how SELES is applied to model the consequences of re suppression on forested landscape dynamics.