Keywords

land use modelling, bayesian networks, decision modelling, validation, subjective validation

Start Date

1-7-2012 12:00 AM

Description

Taking into account land use models in spatial planning processes would be a valuable support for facilitating the development and the acceptance of innovative land use management strategies. But, the utilization of such models in spatial planning processes is rarely done. One condition for implementation is trust and credibility in the model, which can be increased through validation. We argue that subjective validation can be a major contribution to efforts advancing the implementation of models. The presented land use model applies Bayesian networks (BN) and incorporates secondary data as well as local actors’ characteristics. Stakeholders from the study area and experts were involved in designing and updating the network in workshops and questionnaires. The subjective validation procedure is based on the concepts of conceptual and operational validation and these are applied in a workshop format. Results of the subjective validation process show that experts have understood mechanisms of the BN and their inputs can be effectively used for model improvement. The parameterization of the BN was adapted and the understanding of the processes improved. Comparing the results of the subjective validation with the outcome of an objective validation approach reveals surpluses concerning credibility of the model and mutual learning processes.

Share

COinS
 
Jul 1st, 12:00 AM

Participatory Land Use Modeling with Bayesian Networks: a Focus on Subjective Validation

Taking into account land use models in spatial planning processes would be a valuable support for facilitating the development and the acceptance of innovative land use management strategies. But, the utilization of such models in spatial planning processes is rarely done. One condition for implementation is trust and credibility in the model, which can be increased through validation. We argue that subjective validation can be a major contribution to efforts advancing the implementation of models. The presented land use model applies Bayesian networks (BN) and incorporates secondary data as well as local actors’ characteristics. Stakeholders from the study area and experts were involved in designing and updating the network in workshops and questionnaires. The subjective validation procedure is based on the concepts of conceptual and operational validation and these are applied in a workshop format. Results of the subjective validation process show that experts have understood mechanisms of the BN and their inputs can be effectively used for model improvement. The parameterization of the BN was adapted and the understanding of the processes improved. Comparing the results of the subjective validation with the outcome of an objective validation approach reveals surpluses concerning credibility of the model and mutual learning processes.