Comparing spatial modeling techniques for exploratory mapping : applications in wilderness campsite searches
Cross, Edward Tyson
MetadataShow full item record
Camping impacts are known to damage vegetation, impede ecological processes, and negatively affect visitor experiences in wilderness areas. In response to national mandates from the Chief of the Forest Service in 2004, wilderness managers are pursuing an effort to find and inventory 95% of the campsites in designated wilderness. In 2009 the Rocky Mountain Region (Region 2) Wilderness Program of the United States Forest Service focused its resources on carrying out the mandate. This thesis uses spatial models to predict areas likely to have camping impacts. The resulting maps can be utilized by wilderness managers in Region 2 find and inventory camping impacts in wilderness areas. Understanding the spatial distribution of campsites is an important step in the inventory, monitoring and analysis of damage related to camping activities. Recreation Habitat Suitability Indices and Maxent are two tested methods for predicting where an object can be found based on the objects relationship to environmental variables. This research employs these methods for determining the suitability of an area for camping. For this study, A Recreation Habitat Suitability Index was developed using a priori knowledge, without the use of data. A Maxent approach was also implemented using n= 1658 points for model development. Both models were tested with n=1446 points, using Area Under the Curve (AUC) and maximized Cohen's Kappa methods for validating the models. Using these test procedures, the research found that both methods performed outstanding when tested with independent data. Models were then compared, and it was found that the models predict camping impact location with remarkable similarity. The research determined that both methods work well for predicting the most important areas to prioritize campsite inventories. This thesis is divided into four chapters and a technical appendix. It begins with an introductory chapter which overviews the project and explains the goals and objectives. The second chapter is a literature review of past efforts which used similar methods. The third chapter presents the modeling research as a journal article which compares two methods for using predictive modeling to understand the spatial distribution of wilderness camping sites. Following the research chapter, a fourth, conclusions chapter, speaks to the limitations of the model, and indicates what future research efforts may focus on. Following the conclusions chapter is a technical appendix (Appendix A) which contains a technical report in the form of a desktop guide which is intended to help wilderness managers understand camping impacts and analysis techniques. The technical appendix is a summary of knowledge gained by working in the field with camping impacts. The technical report is designed to help managers understand and train wilderness staff to complete campsite inventories, start to finish. The research portion of this thesis provides wilderness managers with two tested methods for determining areas most likely to contain camping impacts in wilderness. This research offers managers an option for determining which areas of a wilderness are important to search in order to find 95% of the campsites. When combined with the technical appendix this thesis as a whole provides wilderness managers with tools for understanding, finding and inventorying camping impacts in the wilderness areas of Region 2.