Transactions in GIS, 2003, 7(3): 325–343 © Blackwell Publishing Ltd. 2003. 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA. Blackwell Publishing Ltd Oxford, UK TGIS Transactions in GIS 1361-1682 © Blackwell Publishing Ltd 2003 June 2003 7 3 1 000 Research Article Spatial Simulation and Fuzzy Threshold Analyses B Güneralp, G Mendoza, G Gertner and A Anderson Spatial Simulation and Fuzzy Threshold Analyses for Allocating Restoration Areas Burak Güneralp Department of Natural Resources and Environmental Sciences University of Illinois George Gertner Department of Natural Resources and Environmental Sciences University of Illinois Gil Mendoza Department of Natural Resources and Environmental Sciences University of Illinois Alan Anderson US Army Corps of Engineers Research Laboratory Champaign, Illinois Abstract This paper presents a methodology for the evaluation of land condition and for the allocation of areas requiring restoration. It is based on spatial simulation analysis and fuzzy logic. The method is demonstrated in a restoration allocation problem within a military training area in Texas. Fuzzy logic is integrated with spatial analysis through Geographic Information Systems (GIS) to make land condition assessment geographically specific. Two sources of uncertainty in Land Condition Analysis are considered in this paper. First is the uncertainty due to incomplete information on land condition. Second is the uncertainty emanating from identifying the condition of a particular parcel of land. The first is addressed by using sequential Gaussian simulation, a geostatistical tool. Erosion status is selected as the land condition factor, and uncertainty associated with it is considered in this study. Land allocation is based on fuzzy logic to reflect the continuous transition between different land conditions and the minimization of loss that is expected to occur in the case of mis- allocation. Various forms of loss functions are used for allocating areas in need of restoration. An important result of the study is a map showing the areas allocated for restoration. The proposed method is compared to two alternative methods with varying degrees of determinism and uncertainty. The incorporation of uncertainty led to better allocation strategies and results that are more realistic. Address for correspondence: Gil Mendoza, Department of Natural Resources and Environmental Sciences, University of Illinois at Urbana-Champaign, W-503 Turner Hall, 1102 South Goodwin Avenue, Urbana, IL 61801. E-mail: gamendoz@uiuc.edu