Visualising Changes in Agricultural Landscapes - Environmental and Agricultural Modelling - page 132

professional versions, http://www.geomantics.com/ or LandSIM3D® commercialised
by Bionatics in France http://www.bionatics.com/ ) . Many are linked to GIS, and
have been described in several literature reviews (Ervin and Hasbrouck 2001 ;
Appleton et al. 2002 ; Pettit et al. 2008) . Often specific 3D objects must be included
in the scene, such as vegetation (Auclair et al. 2001 a; b ; Muhar 2001) , and several
approaches have been described in order to integrate indicators into landscape
visualisations (Bishop et al. 2008 ; Wissen et al. 2008) . A particular emphasis has
been put on reliability and validation of landscape simulation (Sheppard 1982 ;
Daniel 1992 ; Lange 2001 ; Bishop et al. 2005) .
Several European research projects are involved in the issue of landscape
visualisation, such as VisuLands (Wissen et al. 2008) , Greenspace (Lange et al.
2008) , SENSOR (Helming et al. 2008) and BioScene (Soliva and Hunziker 2009) .
These projects however address different issues, they are in particular concerned by
urban landscapes and communication infrastructures, at country or regional scales.
Our aim here was to concentrate on detailed agricultural land-use at the scale of a
territory, it therefore appeared necessary to develop separate components, although
on a common basis.
In the present chapter, we describe a landscape visualisation freeware
component, which is part of the SEAMLESS project (Van Ittersum et al. 2008) .
This component can be launched at the end of a modelling chain destined to
simulate a particular agricultural and/or environmental policy, to allow for explora-
tion of landscape changes. Visualisation could have a significant implication for the
choice of effective land-use policy, and could be used as a basis for discussion and
negotiation within the community.
The pressures causing changes in landscape can be simulated by a bio-economic
farm model, such as described in Chapter 5 of this volume. This can then be
translated into changes in the spatial configuration of the landscape. The mapped
results (environmental data such as land cover and land use) will be specific to each
individual region, and should be available from a GIS database. They will be used
here to compute and visualise a 3D scene.
In this chapter we first describe briefly the software methodology and design
and the data processing, and in a second phase provide an example of application
based on a study of four scenarios in the Pic Saint Loup area of the French
Mediterranean region.
Software Methodology and Design
To build a system that computes a virtual landscape from accurate spatial data and
delivers a reasonable realistic representation of an existing landscape, a large
amount of data is necessary. Firstly, we need to build a real-time renderer that is
powerful enough to handle large datasets of landscape terrain. Secondly, a matching
virtual representation of an existing area has to be constructed, including vegetation
and man-made structures. We have built a system using a free graphic engine as a