Computational System

The HEIMTSA platform is the computational environment that brings together the different tools available from the HEIMTSA consortium addressing all the aspects of the full chain for health impact assessment of Community policies. The platform allows the execution of complex calculations serving the needs of integrated health impact assessment of environmentrelated stressors and policies.

The platform consists of an integrated set of stand-alone (‘loosely-coupled’) modules, designed to link together seamlessly, in order to demonstrate the impact of a range of policies on health. This system covers emissions of all relevant substances into air, water and soil, the transmission and chemical transformation of these substances across all environmental media, and includes indoor air quality and road traffic noise as key stressors as well. The platform closely links models and datasets for an integrated, cross-media impact assessment, implementing the assessment chain for a first time covering the full impact pathway. This closely-coupled modelling system allows for a full exploration of trade-offs and synergies of individual policy options and provides at the same time flexibility for the evaluation of different variations of policy options The models are operated as a distributed system, run and maintained by the model developers. The individual modules are linked via fast internet based data exchange protocols (e.g. implementing GRID FTP).

The main characteristic of the architecture is its decentralized nature.  While partner institutions run their own models on their own servers, the central system activates the models and stores the data from these models.  In addition, the central system can use these data as input to the next model of the running chain.

The main advantage of this architecture is its flexibility.  Indeed, the participating institutions can update regularly their models.  Furthermore, the spatial scale can vary from the European scale to the city scale according to specific needs.

The platform is a typical three tier architecture composed by:

  • Presentation tier: this is the user interface typically accessible using a web browser.
  • Application tier: this layer controls functionalities of the application. It performs data processing and interfaces the database for importing/exporting data.
  • Data tier: in this particular case the database is a geo-referenced database where data are stored and all types of the underlying geo-referenced information are managed.

Expected users of the system can be divided in three main categories:

  • Normal users: they can execute full chains;
  • Model owners: they can execute full chains and they are responsible of one or more models;
  • Administrators of the system: they can do what other users can. Additionally they administer the system managing the components mentioned above and users. They are also responsible for approving or not executions.

Health Impact Assessment Chains can be executed starting from the very first step (the definition of a policy scenario) or from an intermediate position (e.g. starting from a change in ambient air levels). Users can ask for a new emission scenario directly from the interface: a form requiring all the necessary information must be compiled and submitted to scenarios developers. Once the new emission scenario is ready, it will be created directly into the system.

Results are reported and presented in innovative ways, through the visualization module, including the generation of maps that describe the spatial distribution of health impacts and the presentation of the distribution of impacts on different groups of the population. This module, accessible to the user from the GUI of HEIMTSA platform, allows the  user to process spatial data stored in the HEIMTSA Db coming from each step of the fullchain and to visualize the results using Web application client with geo-processing services. This module performs basic interactive functionalities for visualizing geospatial data as well as for carrying out deeper analysis of the data through specific services. After the registration procedure every user accessing the platform can launch an execution but before it starts one of the platform administrators has approve it. For each step already executed of the full-chain the user can download the output files generated and, when available, invoke the visualization module to see on a GIS environment the results obtained. All the information about an execution is stored into the database of the platform and all the input/output files of the models are stored into the server file system.

The Computational System integrates a tool for qualitative uncertainty analysis. For each component stored into the system a qualitative uncertainty analysis can be made and managed by the Computational System. The uncertainty analysis for executions is composed by the analysis directly inserted by the user plus the analysis of the components grouped under the execution (i.e. chains, models and scenarios).