Agricultural landscape generators for simulation models: A review of existing solutions and an outline of future directions

There is an increasing need for an assessment of the impacts of land use and land use change (LUCC). In this context, simulation models are valuable tools for investigating the impacts of stakeholder actions or policy decisions. Agricultural landscape generators (ALGs), which systematically and automatically generate realistic but simplified representations of land cover in agricultural landscapes, can provide the input for LUCC models. We reviewed existing ALGs in terms of their objectives, design and scope. We found eight ALGs that met our definition. They were based either on generic mathematical algorithms (pattern-based) or on representations of ecological or land use processes (process-based). Most ALGs integrate only a few landscape metrics, which limits the design of the landscape pattern and thus the range of applications. For example, only a few specific farming systems have been implemented. We conclude that existing ALGs contain useful approaches that can be used for specific purposes, but ideally generic modular ALGs are developed that can be used for a wide range of scenarios, regions and model types. We have compiled features of such generic ALGs and propose a possible software architecture. Considerable joint efforts are required to develop such generic ALGs, but the benefits in terms of a better understanding and development of more efficient agricultural policies would be high.