Objective 1 Model Rangeland Health at Local and International Sites: Rangeland health is a term that has recently received much emphasis (O'Brien et al 2003, Pellant et al 2005, Pyke et al 2002, National Research Council 1994, Wilcox and Thurow 2006). However, there are many definitions of rangeland health, none of which are universally accepted. At the annual "Geo-spatial and Rangeland Sciences" conference (part of the GIS TReC's public outreach effort) in 2004, attendees participated in an exercise to define rangeland health. While all attendees had valid credentials to not only understand rangeland health but also define rangeland health, there was no clear consensus of the definition. For this reason, we have developed a working definition of healthy rangelands that will be used for this study: "Healthy rangelands exhibit effective water cycles as indicated by minimal bareground exposure and minimal evidence of soil erosion. In addition, the vegetation present will be a mixture of grasses, forbs, and shrubs that act synergistically to provide quality forage and habitat for wildlife and grazing animals. Litter will be present and biodegrading. Very little if any litter will be decomposing through oxidation". Using this definition as our guide, we have developed and refined rangeland health models that describe range health for various areas in southeastern Idaho. The models were produced using satellite imagery (Landsat, SPOT, and Quickbird). In addition, a broad-scale model is being developed using MODIS NPP and LAI data (spatial resolution of 250m). The models are the result of three component models describing bareground exposure, vegetation composition (presence of shrubs and/or grass vegetation components), and litter. A goal of past research was to develop a technique for rangeland health modeling that could be applied elsewhere in the United States or other parts of the world (Gokhale and Weber 2006). Our study is intended to seek global exposure and acceptance for the rangeland health modeling process developed by ISU's GIS TReC. In addition, it is a goal of this proposal to test the applicability of our rangeland health models in other parts of the world.

Objective 2 Compare Rangeland Health Between Local and International Sites: We propose to compare overall rangeland health index values (the output from our rangeland health model) (generalized to 1 ha MMU) within study site pairs (i.e., O'Neal, USA versus the Ebro Valley, Aragon, Spain (between study sites) (Figure 5). In addition bare ground exposure models (a primary indicator of rangeland health, desertification, and a driver of effective hydrologic cycling) (Savory 1999, Thurow and Hester 1997, McGinty et al 1995, Rauzi et al 1968) will be compared between study site pairs. This comparison will be accomplished by first calculating/measuring two response variables at each site: 1) rangeland health indices (generalized to 1 ha MMU) and 2) bare ground exposure (also generalized to 1 ha MMU). Next, a suite of intrinsic (elevation, slope, aspect, soil type, soil erodability) and extrinsic (rainfall, temperature, growing degree days, grazing animal, grazing system, stocking rate and stocking density [the latter two, while similar, are used to determine the importance of spatial versus spatio-temporal livestock distribution on the response variables]) factors will be assessed. Multiple regression will be used to determine if observed differences in the response variables are most attributable to intrinsic factors or extrinsic factors. Further, we seek to identify the individual factors that best explain variations in the response variables (rangeland health index and bare ground exposure). We anticipate that a portion of the variation observed in rangeland health and bare ground will be attributable to variations within the intrinsic factors. However, we also anticipate that the extrinsic factors explored in this study will play a key role in completing our understanding of rangeland health and rangeland ecosystem function. Moreover, this study seeks to determine how effective the geospatial technologies are at capturing extrinsic factor effects as explored in this study; especially management decisions like applied grazing treatments. It is of interest to note that all sites compared are being grazed, however the grazing animal, stocking rate, and spatial and temporal distribution of grazing varies at each site. Spatial distribution has been identified as an important factor influencing vegetation response to grazing (Adler et al 2001, Savory 1999). These specific grazing factors encapsulate what we are collectively referring to as the effect of the decision-making process.

Objective 3 Test Scalability of Composite Rangeland Health Models from SPOT to MODIS: Throughout this study we will develop and refine a composite rangeland health model for each study site using fine resolution SPOT/Aster imagery. It is of interest, from both an academic as well as applied point of view, to determine if the composite rangeland health indices are scalable from fine spatial resolution platforms to more coarse resolution platforms like MODIS. If so, MODIS imagery would most likely offer a cost-effective and end-user friendly solution to range managers, range scientists, ranchers, and land stewards. The MODIS models will be based upon the same component models (bareground exposure, vegetation composition, and litter) that were used with SPOT imagery or with MODIS products that are surrogates for the component models. MODIS surrogates for bareground exposure, vegetation composition, and litter will be explored carefully using cross-tabulation and correlation tests. For example, we anticipate an inverse relationship may exist between the SPOT-based bareground exposure component model and MODIS-based leaf area index (LAI). If proven true, an inverse LAI would be used within the rangeland health model. Similar explorations will be made throughout the modeling process.

Objective 4 Planned Public Outreach Activities: The results from this study will be freely distributed to range managers, range scientists, ranchers, and land stewards. Several venues of public outreach will be implemented throughout the course of this study. Each is intended to help educate the end-user and general public regarding the benefits of the geotechnologies and their application for real-world problems/issues. These venues are 1) an annual "Geo-spatial and Range Sciences" conference, 2) Geotechnologies workshops, 3) publication of "Integration: the newsletter of Geo-spatial and Range Sciences, 4) a Rangeland health comparison website, and 5) the GLOBE forum.