Keywords

BBN Toolbox; decision support system; risk; uncertainty; GIS

Location

Session D1: GIS and Environmental Modeling for Decision Support

Start Date

16-6-2014 2:00 PM

End Date

16-6-2014 3:20 PM

Abstract

Ecosystem services are gaining more and more attention in decision support applications. Nevertheless, modelling and mapping ecosystem services to support landscape planning decisions remains challenging. Recently, Bayesian belief networks (BBNs), a probabilistic modelling technique, has been introduced in ecosystem service modelling. Major advantages of this modelling approach include high model transparency which enables stakeholder involvement in model development and evaluation, the ability to incorporate expert knowledge on top of data and the possibility to take into account uncertainties. To combine the advantages of BBNs and spatially explicit modelling in the context of ecosystem service modelling, we developed a Quantum GIS plug-in. The plug-in enables pixel-based application of BBN models to map ecosystem service delivery and associated uncertain ties. The obtained probabilistic maps can be used for stakeholder involvement, decision support and probabilistic, regional ecosystem service accounting.

 
Jun 16th, 2:00 PM Jun 16th, 3:20 PM

Probabilistic Mapping With Bayesian Belief Networks: An Application On Ecosystem Service Delivery In Flanders, Belgium

Session D1: GIS and Environmental Modeling for Decision Support

Ecosystem services are gaining more and more attention in decision support applications. Nevertheless, modelling and mapping ecosystem services to support landscape planning decisions remains challenging. Recently, Bayesian belief networks (BBNs), a probabilistic modelling technique, has been introduced in ecosystem service modelling. Major advantages of this modelling approach include high model transparency which enables stakeholder involvement in model development and evaluation, the ability to incorporate expert knowledge on top of data and the possibility to take into account uncertainties. To combine the advantages of BBNs and spatially explicit modelling in the context of ecosystem service modelling, we developed a Quantum GIS plug-in. The plug-in enables pixel-based application of BBN models to map ecosystem service delivery and associated uncertain ties. The obtained probabilistic maps can be used for stakeholder involvement, decision support and probabilistic, regional ecosystem service accounting.