Modelling future landscapes is conducted as an approximation of reality to anticipate possible future realities for the purpose of policy advise. In this modelling approach we aim to show two future realities within the basic scenario combinations:
This is increasingly important with short- to long-term policy goals like the Agenda 2030, the Bonn Challenge 2030 and the Paris Agreement which are challenged through climate change and population development over the course of the century.
In the context of forests many questions arise regarding future trajectories, where we will focus on the following:
We are showing different possible landscape configuration trajectories using a dynamic land use simulation tool on a yearly basis (i.e. a land use map is produced for every year based on the land use configuration of the prior timestep). Scenario simulations will be conducted using an updated version of the dynamic and probabilistic PCRaster Land Use Change (PLUC) model (Verstegen et al. 2012) called “LAFORET-PLUC-BE” [Landscape Forestry in the tropics – PCRaster Land Use Change – Biogeographic & Economic model].
LAFORET-PLUC-BE (LPB) will include additionally ecological (forest carbon storage, forest fragmentation) and economic impact assessment (opportunity costs). Model simulations will focus on smallholders’ land use dynamics using a spatial scale of 1 hectare.
The 36 LaForeT landscape tiles are subsumed into eight modelling regions following geographical and administrative system boundaries. Due to the cross-country approach, we rely mostly on global datasets for comparability, but rely on empirical data conducted by the different LaForeT work packages to specify regional conditions of e.g. crop area demands and crop yields among other. The LBP modelling approach is realized in three model stages, each one aiming to compare selected regions and landscapes in a cross-country fashion.
The outputs consist of two different datasets. The first one is the direct model output of the LBP model which simulates the most probable landscape configuration based on the given assumptions and approximations in the chosen scenario combination setting. Additionally, a second dataset is derived representing a possible landscape configuration under a restoration paradigm for each timestep within the prior calculated most probable landscape configuration.
The combination of both outputs enables us to derive for the simulated scenario combinations: (i) a picture of what might happen to a particular landscape and region if no measures are implemented to adapt to changing conditions, as well as possible indications for landscape management how forest landscapes could be restored and what the potential costs of such measures might entail to compensate the replaced or adapted local livelihood options. These findings shall enable policy makers to establish policies for combined top-down and bottom-up measures on the basis of short- to long-term information. Further on the approach could be used as tool box addressing similar problems in other regions of the world.