Spatial modeling of riparian state dynamics in eastern Oregon, USA by using discrete event simulation

Abstract

In order to relate the effects of small- to mid-scale riparian state dynamics to land use management goals at a broader, watershed scale, I developed a general, simulation-based modeling framework for riparian landscape dynamics. Discrete event simulation (DES) was used to implement a set of models incorporating state dynamics for several potential morphological groups, along with a set of landscape events that interact with successional changes in stream vegetation and morphology. These models operate concurrently on a connected landscape, parameterized by extant spatial databases, and expert opinion. As a sample application, two landscape disturbance scenarios were simulated, representing contrasting forest management regimes for a subbasin in eastern Oregon, USA. This case study incorporates the individual dynamics of almost 8000 polygons and 10 types of events in a common framework. Both scenarios resulted in relatively homogeneous landscapes, with primarily mature multistrata forest states under historical conditions, and a mix of mature forest and pioneer forb community under current conditions. These simulation results do not appear to accurately represent current or historical landscape conditions, likely because of insufficient information for adequately calibrating model parameters and their associated uncertainty. Still, this modeling approach provides an effective method for incorporating riparian landscape dynamics in forest landscape decision support systems.