4th International Conference on Integrating GIS and Environmental Modeling (GIS/EM4):
Problems, Prospects and Research Needs. Banff, Alberta, Canada, September 2 - 8, 2000.

A Review and Assessment of Land Use Change Models

dynamics of space, time, and human choice

GIS/EM4 No. 20

Chetan Agarwal
Glen L. Green
J. Morgan Grove
Tom Evans
Charles Schweik

Abstract

Human use and management of terrestrial resources significantly alter vegetation structure, function, extent, distribution, and species composition and change nutrient and hydrologic inputs. These vegetation, nutrient, and hydrologic variables are frequently critical inputs to most biogeochemistry models of global change. Recently, the U.S. Forest Service Southern and Northern Global Change Programs determined the need for a review and assessment of existing land use change models that include social drivers. We review and classify existing models using a classification scheme that arrays models along three-axes: space, time, and human decision-making. We use this classification scheme in part to present our findings and to assess future directions for incorporating social drivers in land use change models.

Keywords

Land use, global change, human decision-making, human-environmental models, modeling, GIS

Full Report

The full 84-page report is currently being published jointly by the USDA Forest Service and the Center for the Study of Institutions, Population, and Environmental Change (CIPEC) at Indiana University. Copies will be available through CIPEC or the USDA Forest Service Northeastern Forest Research Station in Burlington, Vermont.

Introduction

Land use change is an undeniable and significant global, ecological trend. As Vitousek (1994) notes, "Three of the well-documented global changes are increasing concentrations of carbon dioxide in the atmosphere; alterations in the biochemistry of the global nitrogen cycle; and on-going land use/land cover change." In the case of the United States for instance, 121,000 km² of non-federal lands were developed over the 15-year period between 1982 and 1997 (NRCS/USDA 1999). On a global basis and a longer time frame, Ramankutty and Foley (1999) note that nearly 1.2 million km² of forests and woodlands and 5.6 million km² of grasslands and pastures have been converted to other uses during the last three centuries. During this same time, cropland areas increased by 12 million km². Currently, human transformations of the Earth's land surface are significant with 10-15% in row crop agriculture or urban-industrial areas and 6-8% in pastureland (Vitousek et al. 1997). These changes in land use have important implications for changes in the Earth's climate and, in return, implications for subsequent land use change.

This paper is structured in the following way: In the Methods section, we develop an approach for comparing different models of land use change. In particular, we propose that models of land use change can be compared in terms of scale and complexity. Subsequently, we describe the methods we used for identifying the models we reviewed. In the Discussion section, we present our findings in terms of temporal, spatial, and human decision-making complexity; current land use models and social drivers; model types, systems, and modularity; and future directions in land use modeling. Finally, we conclude with some thoughts regarding land use modeling and policy.

Methods

Models can be categorized in several ways. We could focus on the subject matter of models, on modeling technique or methods (from simple regression to advanced dynamic programming) or model use. Alternatively, a review of models could focus on techniques in conjunction with assessments of model performance for particular criteria, such as scale (see, for example, the review of deforestation models by Lambin 1994).

We developed an alternative analytical framework. As Veldkamp and Fresco (1996a) note, land use "is determined by the interaction in space and time of biophysical factors (constraints) such as soils, climate, topography, etc. and human factors like population, technology, economic conditions etc." In this review, we utilize all of the factors that Veldkamp and Fresco identify in the construction of a new analytical framework for categorizing and summarizing models of land use change.

Framework for Reviewing Human-Environmental Models

We propose a framework based on three critical dimensions for categorizing and summarizing models of human-environmental dynamics. Time and space are the first two dimensions and provide a common context in which all biophysical and human processes operate. In other words, models of biophysical and/or human processes operate in a temporal context, a spatial context, or both. When models incorporate human processes, a third dimension—what we refer to as the human decision-making dimension—becomes important as well (Figure 1).

Figure 1. The Three Dimensions of Human-Environmental Models.

In reviewing and comparing human-environmental models along these dimensions, there are two distinct and important attributes that must be considered: model scale and model complexity. We begin with a discussion of scale since it is a concept that readers will probably find most familiar.

Model Scale

"Real world" processes operate at different scales (Allen and Hoekstra 1992, Ehleringer and Field 1993). When we discuss the temporal scale of models, we usually talk in terms of "time step" and "duration." Time step is the smallest unit of analysis for change to occur for a specific process in a model. Duration refers to the length of time that the model is applied. When we discuss the spatial scale of models, we talk in terms of "resolution" and "extent." Resolution refers to the smallest geographic unit of analysis for the model such as the size of a cell in a raster system. (Note that each cell area is typically treated as constant, while a vector representation would typically have polygons of varying size). Extent describes the total geographic area to which the model is applied.

Most readers will find this discussion to be familiar. But how do we discuss human decision-making in terms of scale? To date, the social sciences have not yet described human decision-making in paired terms that are as concise and widely accepted for modeling as time step and duration, and resolution and extent. Like time and space, however, we propose that an analogous approach can be used to articulate scales of human decision-making in terms of two components: "agent" and "domain." "Agent" refers to the human actor(s) in the model who are making decisions. The individual is the most familiar human decision-making agent. But there are many human models that capture decision-making processes at higher levels of social organization; such as household, neighborhood, county, state or province, or nation. While the agent captures the concept of who makes decisions, the "domain" describes the specific institutional and geographic context in which the agent acts. Representation of the domain can be articulated in a geographically explicit model through the use of boundary maps or GIS layers

Model Complexity

The second important and distinct attribute of human-environmental models is the approach used to address the complexity of time, space, and human decision-making found in "real world" situations. We propose that the temporal, spatial, or human decision-making (HDM) complexity of any model can be represented with an index, where a low score signifies only simple components and a high score signifies more complex behaviors and interactions. Consider an index for temporal complexity of models: A model that is low in temporal complexity is a model that has one, or possibly a few, time steps and a short duration. A model with a middle to high score for temporal complexity is one that has many time steps and a longer duration. Models with a high score for temporal complexity are ones that have a large number of time steps, a long duration, and the capacity to handle time lags or feedback responses among variables, or have different time steps for different sub-models.

An index of spatial complexity would represent the "spatial explicitness" of a model. There are two general types of spatially explicit models: spatially representative or spatially interactive. A model that is spatially representative can incorporate, produce, or display data in at least two and sometimes three spatial dimensions—northing, easting, and elevation—but can not model topological relationships and interactions among geographic features (cells, points, lines, or polygons). In these cases, the value of each cell may change or remain the same from one point in time to another, but the logic that makes the change is not dependent on cells neighboring it. In contrast, a spatially interactive model is one that explicitly defines spatial relationships and their interactions (e.g., among neighboring units) over time. A model with a low score for spatial complexity would be one with little or no capacity to represent data spatially, a model with a medium score for spatial complexity would be able to fully represent data spatially, and a model with a high score would be spatially interactive in two or three dimensions.

What might we use to characterize an index for model complexity of human decision-making? We use the term HDM complexity to describe the capacity of a human-environmental model to handle human decision-making processes. In Table 1, we present a classification scheme for estimating HDM complexity using an index from one to six. A model with a low score for human decision-making complexity (1) is a model that does not include any human decision-making. In contrast, a model with a high score (5 or 6) is a model that includes one or more types of actors explicitly or can handle multiple agents interacting across domains.

Applications of the Framework

The three dimensions of land use change models—space, time and human decision-making—and two distinct attributes for each dimension—scale and complexity—provide the foundation for comparing and reviewing land use change models. Figure 2 is an example of the framework with the three dimensions represented and a few models that were included in this study. Various modeling approaches might vary in their placement along these three dimensions of complexity since the location of a land use change model reflects its technical sophistication and application.

Figure 2. A Framework for Reviewing Land Use Change Models.

The analysis that follows articulates existing land use models on each modeling dimension. Models are given a rank in the human decision-making dimension, and their ability in the spatial and temporal dimensions are discussed as well.

Models Surveyed

1. General Ecosystem model (GEM) (Fitz et al. 1996)

2. Patuxent Landscape Model (PLM) (Voinov et al. 1999)

3. CLUE Model (Conversion of Land Use and Its Effects) (Veldkamp and Fresco 1996a)

4. CLUE_CR (Conversion of Land Use and its Effects - Costa Rica) (Veldkamp and Fresco 1996b)

5. Area Base Model (Hardie and Parks 1997)

6. Univariate Spatial Model (Mertens and Lambin 1997)

7. Econometric (Multinomial Logit) Model (Chomitz and Gray 1996)

8. Spatial Dynamic Model (Gilruth et al. 1995)

9. Spatial Markov Model (Wood et al. 1997)

10. CUF (California Urban Futures) (Landis 1995, Landis and Zhang 1998)

11. LUCAS (Land Use Change Analysis System) (Berry et al. 1996)

12. Simple Log Weights (Wear et al. 1998)

13. Logit Model (Wear et al. 1999)

14. Dynamic Model (Swallow et al. 1997)

15. NELUP, Natural Environment Research Council (NERC)-Economic and Social Research Council (ESRC): NERC/ESRC Land Use Programme (NELUP) (O'Callaghan 1995)

16. NELUP - Extension, (O'Callaghan et al. 1995)

17. FASOM (Forest and Agriculture Sector Optimization Model) (Adams et al. 1996)

18. CURBA (California Urban and Biodiversity Analysis Model) (Landis et al. 1998)

19. Cellular Automata Model (Clarke et al. 1998, Kirtland et al. 2000)

Discussion

Temporal Complexity

A graphical representation of the temporal time step and duration and the spatial resolution and extent of the models (Figure 3 Space-Time diagram) facilitates several observations. First, many models with separate ecological modules operate at fine time steps, of a day or a month (exceptions include certain climate-focused models). This fine temporal resolution allows models to more accurately represent rapid ecological changes in certain biophysical spheres — e.g., hydrology. Second, models with multiple time steps, e.g. #s 1, 2, 3, 4, 15, 16, can span both fine and coarse time steps and reflect the temporal complexity of different socio-economic and biophysical sectors more effectively. Third, some of the more complex models also incorporate time lags. For example, the CLUE (#3) and CLUE_CR (#4) models take into account the time taken for different crops and other land uses to provide economic returns as well as provide a two-year buffer against food shortages by carrying over yield surpluses from previous years.

Figure 3. Spatial and Temporal Plot of Models.

Spatial Complexity

More than half of the models provide for spatial interaction and demonstrate the advantages of spatially explicit models that move beyond simple spatial representation. These models include the impact of variations across space and time of different biophysical and socio-economic factors on land use change. Eleven of nineteen models are raster based, four are vector based, and four are classified as neither (Figure 4). This may change, if, for example, model #14 (Swallow et al. 1997) goes beyond the conceptual stage. The mechanistic vector models (#10 and 18) are focused at a city and county level and provide the finest spatial resolution. Their extent is limited by availability of data. Most of the raster models have a spatial resolution range in the range of 80m-1km, broadly mirroring the scale of common remote-sensing data (e.g. Landsat MSS and AVHRR). The latter resolution is also an intermediate one that has the advantage of coverage of larger extents suitable for regional- and global-scale modeling.

The model with the largest extent was the continental scale FASOM model (#17), with a 100-year time horizon, a good example of a dynamic mathematical programming model that predicts allocation of land between agriculture and forestry, and is spatially representative.

Figure 4. Raster and Vector Distribution of Models.

Human Decision-Making Complexity

Figure 5 adds the level of precision of human decision-making to the graphical representation of temporal and spatial scales. Models at level 3 of Table 1 (7 of the 19 models) include significant levels of human decision-making beyond demographic drivers, but are defined by the lack of feedback; thus, the CUF model allocates land based on cost, but does not factor in feedback on prices. At level 4 of human decision-making in Table 1, models incorporate feedback but mostly do not model a particular kind of actor overtly. Thus the PLM and CLUE/CLUE_CR models (#s 2, 3, and 4) have well-developed ecological sectors and extensive human decision-making elements as well as feedback among sectors, but do not explicitly model different types of actors. Model #8 (a land-use model and shifting cultivation decisions) is ranked at level 4 based on its overall complexity in portraying human decision-making. Models at human decision-making levels 5 (#s 14 and 16) and 6 (#15) explicitly model one or more kinds of actors. Model # 14 simulates harvest decisions and includes both economic and non-economic criteria (e.g. habitat for wildlife). The NELUP model extension (#16) is a farm-level model that includes the impact of farming decisions on changes in intensity of land use and in land cover. The general NELUP model has ecological and economic components, farming decisions, and can serve as a decision support tool by providing feedback on the impact of collective-level policies (e.g. support prices, or conservation programs). These characteristics position the NELUP model among the most detailed in terms of model specification in a variety of sectors affecting land use change. However, it should be noted that a highly detailed model is not necessarily more suitable than a model with less specificity. The utility of land use change models can be measured primarily by its ability to demonstrate emergent patterns in land use change processes and secondarily as a predictive tool.

Figure 5. Plot of Spatial and Temporal Scales of Modes, with Human Decision-Making Complexity.

Current Land Use Models and Social Drivers

A general consensus has emerged regarding social drivers of global change, particularly as it relates to land use change. Both the National Research Council's (NRC) report, Global Environmental Change: Understanding Human Dimensions (1992:2-3), and the Long-Term Ecological Research (LTER) Network report (Redman et al. 2000), Toward a Unified Understanding of Human Ecosystems: Integrating Social Science into Long-Term Ecological Research, have identified a list of core social science areas that need to be studied in order to understand variations in land use. This list includes:

1. demography;

2. technology;

3. economy;

4. political and social institutions;

5. culturally determined attitudes, beliefs, and behavior; and

6. information and its flow.

Table 2 summarizes all the land use models that have been reviewed in this paper in the context of the social drivers identified by both the NRC and LTER reports. While some aspects of social drivers are clearly included—demography (population size, density, growth), markets (land production profits and rent), institutions (zoning, tenure), and technology (types of and access to transportation)—there was no clear and systematic consideration of each type of driver and the relationships among them. This is not to say that all sets of drivers are equally important over time, space, and at different scales. We propose that there is a need for land use models to be able to include the relative effects of different social drivers on land use change in the context of space, time, and human decision-making. This is particularly crucial for assessing alternative future scenarios and relative impacts of different policy choices. Thus, it is crucial for developers of land use models to discuss and adopt a more comprehensive and systematic approach to including social drivers of land use change within the context of the NRC and LTER reports and existing social science efforts.

Model Types

The models reviewed covered a range of modeling methods. THE CUF and CURBA models (#s 10 and 18) both employed a mechanistic GIS simulation, combining layers of information with growth projections. Both were noted for their detailed vector resolution. A range of statistical/econometric models (#s 5, 6, 7, 9, 12, 13) applied either raster or vector approaches, though at least two used neither, using aggregated county-level data, mostly without spatial explicitness. Dynamic systems models include the GEM and its application, the PLM (#s 1 and 2). The NELUP model (#15) also utilized a general systems framework. Additionally, several other models utilized dynamic approaches (#s 3, 4, 8, 11, 14 and 17). One model (#19) applied a cellular automata approach to analyze urban expansion.

Systems Approach

Many land use change models focus on specific processes affected by a defined set of variables. An alternative approach is to examine land use change as one component of a socio-ecological system. In developing this systems approach, one difficulty lies in deciding where to add model complexity. Researchers from the social sciences will tend to add complexity on the social side while generalizing components on the biophysical side. Researchers in the natural sciences will do the opposite. A multi-disciplinary team must struggle to find a compromise in complexity, making the model complex enough to operate properly and produce reasonable behavior without over-complicating parts of the model. Another struggle is how to incorporate scale issues into this systems approach. A number of researchers have developed systems models that have provided great insight into highly complex systems (Costanza et al. 1993, Voinov et al. 1999). Many of these models operate at a set spatial scale, but there may be important processes or relationships that are not evident at that spatial scale.

Once a systems model has been constructed, what-if scenarios can be explored more easily than with other modeling approaches that are not systems oriented. In particular, a systems approach can examine which feedbacks exist in a socio-ecological system such as the impact of increases or decreases in agricultural productivity on the local-market prices of those agricultural goods. This scenario testing ability has proved valuable both to researchers and to policy experts in elucidating important relationships to a variety of different systems.

Modularity of Models

The multidisciplinary nature of land use change modeling is paralleled by modularity in the models themselves. Of the 19 models evaluated for this study, all but four were characterized by modular components. In general, modularity may facilitate land use change analysis by assigning a particular disciplinary aspect of the model to separate modules. We found that the majority of the modular models tended to consider multiple disciplines. This was true for those models with explicit biophysical and social components, for example models 1, 2, 3, 4 and 15. This also held true for the largely biophysical models as there were multiple processes to capture in the model.

The complexity of a model is also related to model modularity. Complex models typically involve the interaction of multiple parameters and their creation and validation can be facilitated by utilizing multiple modular components; for example, modularity allows different processes to run at different time steps, different actors can be modeled simultaneously in different modules, and differences in their decision-making horizons can be incorporated in the time-step of different modules. In particular, as we proposed previously, there is a need for modular approach to land use models that includes the relative effects of different social drivers—such as demography; technology; economy; political and social institutions; culturally determined attitudes, beliefs, and behavior; and information and its—flow on land use change in the context of space, time, and scale.

Future Directions in Land Use Modeling

Many of the models reviewed in this report have been under development for long periods of time. These models have evolved over these long terms of development often with a change in focus or expansion into new substantive areas important to the system being modeled but not originally included in the early versions of the model. For example, the Patuxent Landscape Model (PLM) was originally developed as an ecologically based model of the Patuxent watershed in the Eastern United States. Subsequent functionality has been added to the PLM model to incorporate various social-based inputs including population growth, agricultural policy and land use management. This new functionality has expanded the domain of the model but the social-based inputs are not necessarily accommodated by the modeling framework developed for the ecological systems. This is not to detract from the considerable accomplishments of the PLM model and the SME framework in which the PLM model has been implemented. However, developing models in this fashion might lead to early design decisions, which obstruct the performance of future model components added to the base model.

A second important issue in the landuse modeling community is the duplication of effort and sharing of models. It is the view of the authors that several models addressing similar systems are often developed independently. This has the advantage of demonstrating unique approaches to the same research questions. Behind this issue is the considerable documentation necessary to allow model developers to understand each other's code. Until now, most models have developed in isolation. This is related to the fact that modelers have been funded through grants or focused funds from a particular organization with an interest in land use modeling. Even in the context of large interdisciplinary research centers, efforts have been constrained by funds, staff, and expertise. In contrast to traditional approaches to model development, recent advances in Internet technology have created new types of opportunities for collaboration in the development of land use modeling. Already, "open source" programming efforts have been used to solve complex computing problems (see for example, Kiernan 1999, Learmonth 1997, McHugh 1998, and http://www.opensource.org). For further discussion of open source approaches to land use modeling, see Schweik and Grove (In Press) in these conference proceedings

Conclusion

In this paper we developed an approach for comparing different models of land use change. In particular, we proposed that models of land use change can be compared in terms of scale and complexity. Subsequently, we presented our findings in terms of temporal, spatial, and human decision-making complexity; current land use models and social drivers; model types, systems, and modularity; and future directions in land use modeling. We would like to conclude with some thoughts about land use models and policy.

Increasingly, the policy community is interested in land use models that are relevant to their needs. This does not mean that land use models have to be "answer machines." Rather, we expect that land use change models will be good enough to be taken seriously in the policy process. King and Kraemer (1993:356) list three roles a model must play in a policy context: A model should clarify the issues in the debate; it must be able to enforce a discipline of analysis and discourse among stakeholders; and it must provide an interesting form of "advice", primarily in the form of what not to do-since no politician in his or her right mind will ever simply do what a model suggests. Further, the necessary properties for a good policy model have been known since Lee (1973) wrote his "Requiem for large-scale models": 1) transparency, 2) robustness, 3) reasonable data needs, 4) appropriate spatio-temporal resolution, and 5) inclusion of enough key policy variables to allow for likely and significant policy questions to be explored.

To answer policy questions, policy makers will have to begin to identify the key variables and sectors that interest them, their scales of analysis, and the scenarios they anticipate. At the same time, land use modelers should begin discussion(s) with policy makers to understand their needs. Given policy makers' needs, land use modelers will have to translate those needs with particular attention to implicit and explicit temporal, spatial, and human decision-making scale and complexity and the interactions among scale and complexity. Further, land use modelers will need to consider the relative significance of different drivers—demography; technology; economy; political and social institutions; culturally determined attitudes, beliefs, and behavior; and information and its flow on— land use change within the context of policy makers' needs. Finally, there is the need to provide a framework for collaboration and model development. We propose an open source approach. Perhaps there are others? Regardless, we believe the issue of land use change is an important and complex environmental issue that needs "many eyeballs" working together.

Acknowledgements

We would like to gratefully thank the US. Forest Service's Southern and Northern Change Programs and Northeastern Research Station who funded this project through Cooperative Agreement 23-99-0074 and contributions of staff.

We deeply appreciate the guidance provided by Elinor Ostrom throughout the generation of this report as well as some important substantive ideas, particularly in the section on measuring complexity along the human decision-making dimension. We would like to thank John Branigin, Laura Wisen, and Charlotte Hess at the Workshop in Political Theory and Policy Analysis library, Indiana University, who conducted the literature search for the project. The Indiana University Library provided all of the databases, except for the web search engines. We also gratefully acknowledge institutional support from the Center for the Study of Institutions, Population, and Environmental Change (National Science Foundation grant; SBR 9521918).

References Used

Adams, D.M., Alig, R.J., Callaway, J M, McCarl, B.A, Winnett S.M. 1996. The Forest and Agricultural Sector Optimization Model: (FASOM): Model Structure and Policy Applications, Research Paper: PNW-RP-495. Portland OR; USDA, Pacific Northwest Research Station. 60p.

Allen, T.F.H. and T.W. Hoekstra. 1992. Toward a Unified Ecology. New York: Columbia University Press.

Berry, M.W., Brett, C.H., MacIntyre R.L., Flamm, R.O. 1996. "LUCAS: A System for Modeling Land-Use Change." IEEE Computational Science and Engineering 3(1):24-35.

Chomitz, K.M., Gray, D.A. 1996. "Roads, Land Use, and Deforestation: A Spatial Model Applied to Belize." World Bank Economic Review 10(3): 487-512.

Clarke, K.C., Hoppen, S., Gaydos, L.J. 1998. Methods and Techniques for Rigorous Calibration of a Cellular Automation Model of Urban Growth. <http://geo.arc.nasa.gov/usgs/clarke/calib.paper.html> Accessed 2000 June 20

Costanza R, et al. 1993. "Modeling Complex Ecological Economic Systems." Bioscience 43(8):545-555.

Ehleringer, J.R. and Field, C.B. 1993. Scaling Physiological Processes: Leaf to Globe. San Diego: Academic Press.

Fitz, H.C., DeBellevue, E.B., Costanza R., Boumans R., Maxwell T., Wainger L., Sklar F.H. 1996. "Development of a General Ecosystem Model for a Range of Scales and Ecosystems." Ecological Modeling (88):263-295.

Gilruth, P.T., Marsh, S.E., Itami, R. 1995. "A Dynamic Spatial Model of Shifting Cultivation in the Highlands of Guinea, West Africa." Ecological Modeling (79):179-197.

Hardie, I.W., Parks, P.J. 1997. "Land Use with Heterogeneous Land Quality: An Application of an Area Base Model." American Journal of Agricultural Economics 79(2):299-310.

Kiernan, V. 1999. "The ‘Open Source Movement’ Turns Its Eye to Science." The Chronicle of Higher Education, (November)5:A51.

King, J.L., Kraemer, K.L. 1993. "Models, Facts, and the Policy Process: the political ecology of estimated truth." In Environmental Modeling with GIS. Eds. M.F. Goodchild, B.O. Parks, and L.T. Steyaert, 353-60. New York: Oxford University Press.

Kirtland, D., Gaydos, L., Clarke, K., DeCola, L., Acevedo, W., Bell, C. 2000 An Analysis of Human Induced Land Transformations in the San Francisco Bay/Sacramento Area. <http://edcwww2.cr.usgs.gov/umap/pubs/WRR_paper.html>Accessed 2000 June 26

Lambin, E.F. 1994. Modeling Deforestation Processes: A Review. TREES Series B. European Commission. Research Report No 1, EUR 15744. p 1-108.

Landis, J.D. 1995. "Imagining Land Use Futures: Applying the California Urban Futures Model." APA Journal 61(4):438-457.

Landis J.D., Monzon J.P., Reilly M., Cogan C., 1998. "Development and Pilot Application of the California Urban and Biodiversity Analysis (CURBA) Model." 1998 ESRI International User Conference, October 7-9, 1998.

Landis, J.D., Zhang, M. 1998. "The Second Generation of the California Urban Futures Model: Part 2: Specification and Calibration Results of the Land-Use Change Submodel." Environment and Planning B 25(6):795-824.

Learmonth, M. 1997. Giving It All Away 15 Feb 2000, < http://www.metroactive.com/papers/metro/05.08.97/cover/linus-9719.html>

Lee, D.B., Jr. 1973. "Requiem for Large-Scale Models." AIP Journal (May):163-77.

McHugh, J. 1998. "For the Love of Hacking." Forbes, 20 Mar 2000, <http://www.forbes.com/forbes/98/0810/6203094a.htm>

Mertens, B., Lambin, E.F. 1997. Spatial Modelling of Deforestation in Southern Cameroon. Applied Geography 17(2): 143-162.

Moxey A.P., White B., O'Callaghan J.R. 1995. "The Economic Component of NELUP." Journal of Environmental Planning and Management 38(1):21-33.

National Research Council. 1992. Global Environmental Change: Understanding the Human Dimensions. Washington, DC.: National Resource Council, National Academy Press.

Natural Resources Conservation Service, United States Department of Agriculture. 1999. Summary Report 1997 Natural Resources Inventory. Iowa State University Statistical Laboratory (http://www.nhq.nrcs.usda.gov/NRI)

Oglethorpe, D. R., O'Callaghan, J. R. 1995. "Farm-level Economic Modelling within a River Catchment Decision Support System." Journal of Environmental Planning and Management 38(1):93-106.

O'Callaghan, J.R. 1995. "NELUP: An Introduction." Journal of Environmental Planning and Management 38(1):5-20.

Opensource.org, 2000. Frequently Asked Questions About Open Source, 15 Feb 2000, <http://www.opensource.org/faq.html>

Ramanukutty, N., Foley, J.A. 1999. "Estimating Historical Changes in Global Land Cover: Croplands from 1700 to 1992." Global Biogeochemical Cycles.

Redman, C., Grove, J.M, and Kuby, L. 2000. Toward a Unified Understanding of Human Ecosystems: Integrating Social Sciences into Long-Term Ecological Research. White Paper of the Social Science Committee of the LTER Network. http://www.lternet.edu/research/pubs/informal/socsciwhtppr.htm

Schweik, C.M., Grove, J.M. In Press. "Fostering Open-Research Via a Web System." Public Administration and Management: An Interactive Journal (http://www.pamij.com).

Swallow, S.K. Talukdar, P., Wear, D.N. 1997. "Spatial and Temporal specialization in forest ecosystem management under sole ownership." American Journal of Agricultural Economics 79: 311-326

Veldkamp, A., Fresco, L.O. 1996a. "CLUE: A Conceptual Model to Study the Conversion of Land Use and Its Effects." Ecological Modelling 85:253-270.

Veldkamp, A., Fresco, L.O. 1996b. "CLUE-CR: An Integrated Multi-Scale Model to Simulate Land Use Change Scenarios in Costa Rica." Ecological Modelling 91:231-248.

Vitousek, P. M. 1994. Beyond Global Warming: Ecology and Global Change. Ecology, 75:1861-1876.

Vitousek, P.M., Mooney, H.A., Lubchenco, J., Melillo, J.M. 1997. "Human Domination of Earth's Ecosystems." Science 277(15 July 1997):494-499.

Voinov, A., Costanza, R., Wainger, L., Boumans, R., Villa, F., Maxwell T., Voinov, H. 1999. "Patuxent Landscape Model: Integrated Ecological Economic Modeling of a Watershed." Environmental Modelling and Software 14: 473-491.

Wear D.N, Bolstad P. 1998. Land-use changes in Southern Appalachian landscapes: Spatial analysis and forecast evaluation. Volume 1(6), Ecosystems. p 575-594

Wear D.N., Apt, R., Mangold, R. 1998. "People, space, time: factors that will govern forest sustainability." Transactions of the 63rd North American Wildlife and Natural Resources conference, 1998, March 20-25, Orlando, Fl. Washington, DC: Wildlife Management Institute 348-361.

Wear, D.N., Liu, R., Foreman, J.M, Sheffield, R.M. 1999. "The effects of population growth on timber management and inventories in Virginia." Forest Ecology and Management 118:107-115.

Wood, E.C., Lewis, J.E., Tappan G.G., Lietzow, R.W. 1997. "The Development of a Land Cover Change Model for Southern Senegal." Presented at Land Use Modeling Workshop, EROS Data Center, Sioux Falls, South Dakota, June 5-6, 1997. <http://www.ncgia.ucsb.edu/conf/landuse97/papers/wood_eric/pecdoc.html>

Authors

Chetan Agarwal, M.P.A., Research Assistant
Center for the Study of Institutions, Population, and Environmental Change, 408 N. Indiana Ave., Indiana University, Bloomington, IN 47408, Email: cagarwal@indiana.edu

Glen Green, Ph.D., Post Doctoral Fellow
Center for the Study of Institutions, Population, and Environmental Change, 408 N. Indiana Ave.
Indiana University, Bloomington, IN 47408, Email: glgreen@indiana.edu

J. Morgan Grove, Ph.D., Research Scientist
USDA Forest Service, Northeastern Research Station, 705 Spear Street, South Burlington, Vermont 05403, Email: jmgrove@att.net

Tom Evans, Ph.D., Assistant Professor
Department of Geography, Student Building 120, Indiana University, Bloomington, IN 47405 USA,
Email: evans@indiana.edu

Charlie Schweik, Ph.D., Assistant Research Professor
Center for Public Policy and Administration, 416 Thompson Hall, University of Massachusetts, Amherst, MA 01003,
Email: cschweik@pubpol.umass.edu