Introduction

1 Introduction

Land-use change is considered as one of the most important processes when attempting to understand and model global change (Foley et al., 2005 ). Conversion of natural land worldwide has been identified as a key driver for the release and uptake of carbon dioxide from the biosphere (e.g. Achard et al., 2004 ), for changing of green water fluxes (Rockström et al., 1999 ) and for the decrease of biological diversity (Sala et al., 2000 ). During the past 10 years substantial research has been conducted on land-use processes (Lambin and Geist, 2006 ) and on how they affect human societies, so that there is a much clearer understanding of these processes and a much better appreciation of their complexity.

As there is currently no single theory to describe all the complexities involved (Veldkamp et al., 2001 ), a variety of simulation models have been developed during the past years to simulate changing land-use and land-cover pattern. The problem domain of these models ranges from urbanization (e.g. Couclelis, 1997 ; Clarke et al., 1997 ) over management of coastal zones (e.g. de Kock et al., 2001 ) to deforestation (e.g. Pontius Jr et al., 2001 ). Existing reviews of land-use models develop classification schemes that are either focused on model application or on methodological issues. Examples of the first category are Lambin et al. (2000 ), who evaluate models in respect of their ability to simulate agricultural intensification, and Heistermann et al. (2006 ) who concentrate on global and continental scale applications. Both reviews categorize models according to the underlying model principles (i.e. empirical-statistical, economic and process-based). In both reviews, the term “integration” refers to the combination of different model approaches. Other reviews of this category concentrate on economic models of deforestation (Angelsen and Kaimowitz, 1999 ), on agricultural trade (van Tongeren et al., 2001 ; Balkhausen and Banse, 2004 ) and on model approaches for analysing of impact of multifunctional agriculture (Rossing et al., 2007 ). Examples of the second group of reviews (based on methodological issues) are Verburg et al. (2004 ) who discuss model concepts like driving forces, spatial interaction or level of integration as well as Verburg et al. (2006a ) and Agarwal et al. (2002 ) who use model characteristics like spatial versus non spatial or dynamic versus static as discriminators. Moreover, Parker et al. (2002 ) concentrate on multi-agent systems.

In this paper we develop a framework for model review, taking a strict system oriented perspective, which is based on the conceptual model of land systems defined within the Science plan of the Global Land Project (GLP, 2005 ). The key elements of these systems serve as major categories to analyse the structure and functionality of state-of-the-art land system models for the regional to global scale. Section 2 introduces the conceptual model of land systems. In Section 3, first the review framework and the criteria that were used to select the different models are described. Then, each model is analysed in detail, according to the defined review categories. The discussion synthesizes the findings of the model review and further addresses questions of model application and adaptation to other geographic regions. Finally, further research needs are identified.