A Multi-Level Approach to Modeling Rapidly Growing Mega-Regions as a Coupled Human-Natural System

The FUTure Urban-Regional Environment Simulation (FUTURES) integrates information on nonstationary drivers of land change (per capita land area demand, site suitability, and spatial structure of conversion events) into spatial-temporal projections of changes in landscape patterns (Meentemeyer et al., 2013). One striking feature of FUTURES is its patch-growth algorithm that includes feedback effects of former development events across several temporal and spatial scales: cell-level transition events are aggregated into patches of land change and their further growth is based on empirically derived parameters controlling its size, shape, and dispersion. Here, we augment the FUTURES modeling framework by expanding its multilevel structure and its representation of human decision making. The new modeling framework is hierarchically organized as nested subsystems including the latest theory on telecouplings in coupled human-natural systems (Liu et al., 2013). Each subsystem represents a specific level of spatial scale and embraces agents that have decision making authority at a particular level. The subsystems are characterized with regard to their spatial representation and are connected via flows of information (e.g. regulations and policies) or material (e.g. population migration). To provide a modeling framework that is applicable to a wide range of settings and geographical regions and to keep it computationally manageable, we implement a 'zooming factor' that allows to enable or disable subsystems (and hence the represented processes), based on the extent of the study region. The implementation of the FUTURES modeling framework for a specific case study follows the observational modeling approach described in Grimm et al. (2005), starting from the analysis of empirical data in order to capture the processes relevant for specific scales and to allow a rigorous calibration and validation of the model application. In this paper, we give an introduction to the basic concept of our modeling approach and describe its strengths and weaknesses. We furthermore use empirical data for the states of North and South Carolina to demonstrate how the modeling framework can be applied to a large, heterogeneous study system with diverse decision-making agents. Grimm et al. (2005) Pattern-Oriented Modeling of Agent-Based Complex Systems: Lessons from Ecology. Science 310, 987-991. Liu et al. (2013) Framing Sustainability in a Telecoupled World. Ecology and Society 18(2), 26. Meentemeyer et al. (2013) FUTURES: Multilevel Simulations of Merging Urban-Rural Landscape Structure Using a Stochastic Patch-Growing Algorithm. Annals of the Association of American Geographers 103(4), 785-807.