Using Agent-Based Modelling With Inundation Predictions to Enhance the Resilience of Small Businesses (PAP014521)

Código

PAP014521

Autores

Nigel Wright, Sangaralingam Ahilan, Graham Coates, Andrew Sleigh

Tema

Urban Floods

Resumo

Flooding affects different individuals, communities and organisation in different ways. All of these can suffer either directly or indirectly due to the impact on small businesses. These businesses are often vital to a community's well-being and economic success, but are also vulnerable due to the lack of time and resources to make themselves resilient to flooding. The UK's Engineering and Physical Sciences Research Council is funding a project on innovative solutions to flood risk and in this research is being carried out to understand and improve business response to and preparedness for flood events in urban areas. To achieve this aim, agent based modelling and inundation simulation are being combined to study organisational business continuity over time periods beyond the immediate flood impact (i.e. days rather than hours or minutes). This paper reports on the development of the virtual geographic environment (VGE) component of an agent based model along with how this has been combined with inundation modelling to identify businesses affected by flooding in urban areas. The VGE has been developed, using OS MasterMap©, to identify and classify the UK businesses affected by actual and hypothetical flood events. Inundation modelling based on 1D and 2D approaches combined then allows the identification of all flood affected organisations within the geographical area of interest. In the modelling careful attention has been paid to 1D-2D linking and correct representation of buildings in order to allow for the identification of individual properties at risk. Coupling the VGE with inundation predication provides credibility in modelling flood events in urban areas, even those not yet exposed to flooding. A case study of the Lower Don Valley in Sheffield, which was severely affected by flooding in June 2007, is presented leading to the identification and classification of businesses impacted. Further work will focus on the development of intelligent agents to model and simulate businesses during floods and in the aftermath of these events such that changes in their actions and behaviours can be investigated leading to improved operational response in the future.