to a wide body of users SIENA was built within a Geographic Information System (GIS). Generality is aspired by pooling information from a sample of real-world cities to inform the development of a typical urban area rather than building a representation of a specific city. This generalisation of information allows to explore broad trends rather than unique or outlier observations specific to one city. At the same time, SIENA strives to maximise realism by rooting its development
on structural and design rules obtained from the statistical exploration of these real-world cities. It was developed probabilistically which has the additional advantage that the uncertainties inherent in environmental data can be accounted for by allowing a range of values rather than a single value. SIENA offers a novel tool to address some of the fundamental challenges faced by environmental health studies in urban settings. Possible applications of this user-controlled urban simulation model in the context of exposure science and health risk assessment are many. They include:
the development and testing of new models such as models of pedestrian flow, urban air pollution or micro-environments
under conditions that the user can control;
exploring uncertainties in exposure assessments (e.g. data gaps or the effects of unmeasured processes such as population movements and migration) and assessing their potential influence on the outcome of the analysis;
research into spatial processes and relationships operating in urban areas relating to environmental health and in how these might be affected by different scenarios (e.g. introduction of congestion charging, change of flight path);
investigating the possible effects of such policies to show how these might play out across a heterogeneous urban population (e.g. who will gain and who will lose) allowing for the ex-ante assessment of potentially expensive and negative outcomes of policy decisions. SIENA provides the spatial platform and data infrastructure to support these and other applications. Its pre-eminent strength compared to simulations of real-world cities lies in its flexibility to modify and manipulate the properties and appearance of the urban system (e.g. to alter the spatial pattern of the urban infrastructure,
to change the demographic composition, to vary exposureresponse functions for specific health endpoints). Without any
feasibility restrictions Monte Carlo-type analyses can be conducted by repeated random sampling of parameter values of the urban properties.