Carbon, vegetation, agriculture and water

Key publications

Bondeau et al., 2007 Gerten et al., 2004 Sitch et al., 2003

Description of Carbon, vegetation, agriculture and water

Description of Carbon, vegetation, agriculture and water

LPJmL is the carbon, vegetation, agricultural and hydrology model in IMAGE 3.2 and consists of the three components: Carbon cycle and natural vegetation, Crops and grass, Water.

Within the Earth system, the terrestrial biosphere is the component that bears the most visible impact of human activity. Large proportions of the land surface and the terrestrial vegetation have been converted for human use, for instance, to cropland and urban areas.

Agriculture, terrestrial carbon, water and nutrient cycles were separate modules in previous versions of IMAGE and thus interactions were not adequately covered. IMAGE 3.2 covers natural and agricultural terrestrial ecosystems, and associated carbon and water dynamics via the link with the dynamic global vegetation, agriculture and water balance model LPJmL (Lund-Potsdam-Jena model with managed Land; Sitch et al., 2003; Gerten et al., 2004; Bondeau et al., 2007; Schaphoff et al., 2018a; Schaphoff et al., 2018b). This enables more detailed and process-based representation of the interacting dynamics in vegetation, carbon and agricultural production, and extends the model scope to terrestrial freshwater dynamics.

LPJmL is one of the most extensively evaluated dynamic global vegetation models (DGVM) and is widely applied either stand alone or linked to other models. To show the complex dynamics in the terrestrial biosphere and to reflect the historical IMAGE modules, LPJmL is described in three components: carbon cycle and vegetation; agricultural land use; and terrestrial freshwater flows.

IMAGE 3.2 and LPJmL are linked through an interface that enables close and consistent interaction between the two models in annual time steps (Müller et al., 2016a). An even more direct link to simulate detailed land-atmosphere interaction would require higher temporal resolutions also in other IMAGE components (e.g., the climate model), which is not necessarily congruent with the philosophy of an integrated assessment model. Incorporating nutrient cycles and improving representations of grassland management in LPJmL will require further adjustments to other IMAGE 3.2 components, and will increase consistency.

The dynamic coupling between IMAGE and LPJmL makes it the standard approach to always take impacts of a changing climate into account: most importantly the effects on crop yields, natural vegetation and water dynamics from changes in temperature, precipitation and CO₂ concentrations. The CO₂ fertilisation effect on crop yields in LPJml is found to be relatively optimistic compared to other crop models, partly because other processes negatively affect yields such as nutrient limitations, and the effects of drought and extreme weather events are not accounted for (Toreti et al., 2020). Therefore, IMAGE assumes a 50% efficacy of the default CO₂ fertilisation in LPJmL.