Land Change and the Human Element, challenges in the creation of Land Change Models (LCMs)

Humans are like no other force in reshaping the Earth’s surface.  People have impacted climatic and eco-systems through changes in the use of land.  Many of the environmental impacts are unattended, but it is possible to describe the Earth’s surface (land-cover) and project the future land change by using land-change models (LCMs) (NRC, 2013).  Developments of LCMs are critical to evaluating policy and mitigating the impacts of human development. The challenge in developing LCMs is the human element.  LCMs have to model human behavior and then translate individual decisions into collective impacts on the environment.  Uncertainties in environmental models coupled with the challenges in modeling human behavior make LCMs a particular challenge. Land-use change is a complex process that involves many competing actors and factors, from different social and spatial levels (Valbuena et al., 2010).

What LCMs do is generalize the complex interactions of humans with the environment into particular sets of rules, i.e. if one action occurs then a resulting action will take place.  When a system is created into general rules, then it can be parameterized and simulated to either describe current or past land change, or project future land change. For example in a study of the response of landcover changes to road paving in the Amazon,  Soares-Filho et al. (2004) were able to utilize LCMs to integrate our knowledge of Amazon land-use dynamics, and consequently test a large number of hypotheses concerning its landscape evolution.  They were able to explore the causes of deforestation and evaluate factors, and were able to determine that Brazilian Amazonia land cover can be partially explained in relationship to social and economic opportunities in other regions of Brazil (Soares-Filho et al., 2004).  This was accomplished through scenario-generating models coupled to a landscape dynamics simulator. The outputs were transition probability maps that they analyzed to come to their conclusions, which gives insight for development policy. They found that governance or the enforcement of environmental regulations did mitigate some land cover change.

Many frameworks exist in models including, Equation-Based, System, Expert, Evolutionary, and Cell based models (Parker et. al, 2003), but there is a lot of promise in Agent-based models (ABMs).  ABMs analyze and simulate land-use/cover change as the result of individual decisions (Valbuena et al., 2010). It adds a focus on the human element of land change, simulating the behaviors directly. Challenge in ABM is including the diversity of decision-making actors (Valbuena et al., 2010).  Questionnaires are the typical tool of use in ABMs, where surveys are used to create the rules needed to guide the model.  So in general ABMs have been limited to local problems, but regional simulations are beginning to be experimented with.  One approach is through agent topology, or generalizing a larger population to a few typical agents to simulate.  Once rules are developed and choice models created, then random probabilities can be used to simulate what possible outcomes can take place. 

ABMs are powerful at simulating reality, and LCMs in general give a lot of insight into how the world works.  It must be always taken into account when working with LCMs, that one is working with a simulation.  It is a model, and explaining the model does not always make sense when translating finding to reality.  Simulations have to rely on simplification of reality.  General rules replace complex interaction.  Rational thought is assumed over emotional decision-making.  So it is imperative to always bring your model back to reality when creating them, constantly reference real life, and ground truthing results.

Works Cited

National Research Council. Advancing Land Change Modeling: Opportunities and Research Requirements. Washington, DC: The National Academies Press, 2013.

Parker, D. C., Manson, S. M., Janssen, M. A., Hoffmann, M. J., & Deadman, P. (2003). Multi-agent systems for the simulation of land-use and land-cover change: a review. Annals of the Association of American Geographers, 93(2), 314-337.

Soares‐Filho, B., Alencar, A., Nepstad, D., Cerqueira, G., Diaz, V., del Carmen, M., ... & Voll, E. (2004). Simulating the response of land‐cover changes to road paving and governance along a major Amazon highway: the Santarém–Cuiabá corridor. Global Change Biology, 10(5), 745-764.

Valbuena, D., Verburg, P. H., Bregt, A. K., & Ligtenberg, A. (2010). An agent-based approach to model land-use change at a regional scale. Landscape Ecology, 25(2), 185-199.