A simulation model for wilderness use in Yosemite National Park

In the United States, legislation requires management of wilderness land to maintain its wild character. As the popularity of outdoor recreation grows, managing wilderness becomes not only more necessary, but also more difficult and complex. To help make importantdecisions, managers have been increasingly turning to computer simulation models for analysis and testing. This paper details the creation of a wilderness use simulation model for Yosemite National Park and demonstrates its utility. The model was implementedwith ExtendSim OR simulation software and stochastically simulates trips takenby parties through Yosemite’s 53 wilderness zones throughout the season. We parameterizedthe model using the database of all backcountry permits issued in 2010. Temporal distribution of trip start dates was the only deterministic element of the model. Party size and starting trailhead were determined randomly for each party based on empirical distributions taken from the database. Instead of selecting trip itineraries from a fixed set, we used a matrix of state transition probabilities to randomly simulate each trip, both spatially and temporally. The model also included algorithms to simulate the tendency for partiesto deviate spatially and temporally from their intended itinerary; these were parameterizedwith data collected from a random sample of park visitors. The model tracked backcountry use from trips the originated both from within and outside of the park. Validation statistics showed that the model successfully captured observed party and trip characteristics. We demonstrate the utility of the model by simulating two hypothetical scenarios in which visitationto high-use zones was reduced by lowering the number of parties allowed to start at popular trailheads and redistributing them to less popular trailheads. We also simulate a hypothetical steady-state scenario in which all trailheads are filled to quota on each day.