Combined impacts of food demand and climate change on land and water use: an integrated environmental-economic modelling approach

In the coming decades, world agricultural production systems will face serious challenges. Population growth, economic development, and lifestyle changes towards rich-country dietary patterns will lead to considerable increases in global food demand. Moreover, rising demands for renewable, biomass-based energy products as well as land to be set aside for nature conservation and biodiversity protection may restrict the area available for food production. On the biophysical side, global climate change will affect production conditions, for better or worse depending on regional characteristics. Moreover, the future path of long-term productivity growth in agriculture to adapt to these changes is subject to considerable uncertainty. In order to simulate these combined effects consistently and in a spatially explicit way, we have linked the Lund-Potsdam-Jena Dynamic Global Vegetation Model (LPJ) with a "Management model of Agricultural Production and its Impact on the Environment" (MAgPIE). LPJ simulates the major biogeochemical features of plant growth and represents the global biosphere with a spatial resolution of 0.5 degree, which is equivalent to a grid size of about 50 by 50 km. LPJ provides potential yields for 13 major crop types, depending on soil conditions, climate and precipitation. MAgPIE is designed as a programming model covering the most important agricultural crop and livestock production types worldwide. It takes regional economic conditions (eg food demand, cost structures) as well as LPJderived biophysical constraints (eg potential crop yields, water availability) into account and derives specific land use patterns for each grid cell. These land use patterns are fed back into LPJ in order to derive biophysical impacts of different land use scenarios. One major objective of this modelling exercise is to assess under which economic and climate conditions regional water constraints may become binding. A prototype of the coupled modelling system has already been applied to a small sample region. In this paper we also describe a global version of the land use model, covering 10 economic regions and international trade. Input data for regional cost structures and net-trade flows are taken from FAO statistics and the GTAP database.