In the wide field of modelling, a number of land use models exist that provide consistent time series and gridded data of historic land use and/or possible pathways of future land use following projected human demands. To fulfil these demands, land use models face a number of complex challenges, such as using incomplete global datasets as the basis, deciding on a way and degree of complexity to translate socio-economic drivers or policy directions into absolute land use changes, etc.
In this study we make use of one recently developed land use model (CLUMondo) and three frequently used land use and integrated assessment models (IMAGE, MAgPIE, LUH) that provide data for in total 15 different futures of global land use. Land use scenarios follow projected developments of future human demands under a business-as-usual scenario, provide alternative pathways under different climatic developments (RCPs), or have additional demands superimposed to business-as-usual (e.g. land-based climate mitigation with BECCS or afforestation, biodiversity conservation). In a first step, we evaluate differing strategies to select multiple possible futur ... mehre human demands and to translate them into land use maps and changes over time. In a second step, we quantify and compare ecosystem service indicators (C storage, NPP, evapotranspiration, N leaching, crop yields, water supply, BVOCS) for the different land use options until 2040 that were simulated with the LPJ-GUESS global vegetation model.
The analysis reveals how well demands (e.g. a carbon dioxide removal target) that were initially implemented by the land use models were met under a potentially more realistic, process-based modelling approach and what the effects of these few implemented demands are on a suite of ecosystem services. This provides a valuable insight to the question whether land use models capture everything that is needed to set up realistic projections of future land use based on projected human demands.