Keywords
bayesian networks, decision support, integrated catchment management, weir, modelling, interactive 3d landscape, design process
Start Date
1-7-2010 12:00 AM
Abstract
Sustainable integrated catchment management is a complex task that often involves finding compromise between the views of multiple stakeholders that may not easily be brought to consensus. Interactive three dimensional (3D) visualisations of landscapes can facilitate discussion, but these can only partially inform decision makers as many important aspects of interventions are abstract. Such abstract criteria however can be modelled using a Bayesian Network (BN), combining expert opinions, empirical evidence and data derived from existing models. Thus, it is hypothesised that by combining an interactive 3D landscape design software with a BN, a decision support tool is created that brings together the complementary strengths of both techniques. To test this idea, such a tool has been developed for a river management problem in Sheffield, UK. Impounding the River Don that flows through the city are many weirs which form significant landmarks of the urban riverscape and determine the recreational quality for canoeists. Consequently, management decisions must account for the effect of weir modification not only on the visual aesthetic, but also on the abstract recreational quality of the river for canoeists. To support this problem, an interactive tool has been developed that integrates 3D design and visualisation of weir modification options with indicators of the utility of the river from the perspective of the canoeists, modelled using a BN. It is demonstrated that it is possible to feed back the indicator assessment of a design alongside the visualisation or as part the visualisation.
An interactive visual decision support tool for sustainable urban river corridor management
Sustainable integrated catchment management is a complex task that often involves finding compromise between the views of multiple stakeholders that may not easily be brought to consensus. Interactive three dimensional (3D) visualisations of landscapes can facilitate discussion, but these can only partially inform decision makers as many important aspects of interventions are abstract. Such abstract criteria however can be modelled using a Bayesian Network (BN), combining expert opinions, empirical evidence and data derived from existing models. Thus, it is hypothesised that by combining an interactive 3D landscape design software with a BN, a decision support tool is created that brings together the complementary strengths of both techniques. To test this idea, such a tool has been developed for a river management problem in Sheffield, UK. Impounding the River Don that flows through the city are many weirs which form significant landmarks of the urban riverscape and determine the recreational quality for canoeists. Consequently, management decisions must account for the effect of weir modification not only on the visual aesthetic, but also on the abstract recreational quality of the river for canoeists. To support this problem, an interactive tool has been developed that integrates 3D design and visualisation of weir modification options with indicators of the utility of the river from the perspective of the canoeists, modelled using a BN. It is demonstrated that it is possible to feed back the indicator assessment of a design alongside the visualisation or as part the visualisation.
