Get Adobe Flash player

Scientific programme

Scientific focus

Over the past decade, a broad consensus among scientists and decision makers regarding the imperative need to 'critically examine the observational and modelling tools used for tracking the atmospheric dispersion of hazardous materials and to assess the value of dispersion forecasts for providing useful information to emergency responders and the generals public' (National Research Council of the National Academies, 2003) has formed. The major tasks in such an initiative are:

  • To review the current suite of tools and models that are used in characterizing hazard dispersion and examine how these models are applied operationally for emergency response efforts.
  • To identify deficiencies in the tools and models that limit their effectiveness and operational use in emergency situations.
  • To determine the observational and input data needed to initialize, test and use these models effectively and to identify ways to improve their accuracy.

COST Action ES1006 addressed these three tasks in a comprehensive and scientific way. Both developers and users of airborne hazards dispersion models have a mutual interest in assessing the performance and reliability of tools applied for emergency management. In order to measure the quality of model results and to improve the implementation of dedicated local-scale models, a task specific validation and application procedure was adopted. Assessing the fitness for purpose of different modelling approaches requires a structured set of local threat scenarios to be established. Test cases must clearly distinguish between building/street-scale scenarios, neighbourhood scale threats and larger threat zones in open terrain. In this context, possible sources and release situations have to be characterized and categorized considering specific model requirements. Already existing exposure indices, as defined for example by the ECETOC (European Centre for Ecotoxicology and Toxicology of Chemicals) have to be taken into account. Defining threat measures for groups of released agents (e.g. dosage-based or extreme – event - based exposure evaluation) and distinct release scenarios (instantaneous/continuous release, release after explosion/fire/leakage) will enable a qualified and precise definition of model requirements. Hence, one of the first scientific deliverables of the Action was a state-of-the-art report on a modeling oriented characterization of local-scale threat scenarios as seen by emergency management and first responders. 

Another major task of the Action was to setup a dedicated comprehensive inventory of models applicable to local-scale accidental releases. In many cases, National and/or institutional inventories have been compiled but a complete and consistent European catalogue of tools and models was not yet available at the time when the Action was still in preparation. The Action developed a flexible structured, relational model data base considering the variety of models and tools both currently listed and those that will potentially be listed the future. This enabled efficient access to the desired information such as physical background, computational demands and information on model verification or related performance measures. More specific features such as backtracking sources from existing measurement data or interfacing with common information- and emergency-management systems were documented as well.

Testing and evaluating available models by model inter-comparison as well as by comparison against test data from qualified field and laboratory experiments was another important research task that was coordinated and worked on throughout the entire Action. Assuming that a typical emergency response model has already been validated with regard to local-scale dispersion modelling, the existing model evaluation and validation strategies (e.g. Model Evaluation Group, 1994 or COST 732, 2009) were extended towards task- and application-specific measures for accidental release scenarios such as extreme value prediction, backtracking verification, source reconstruction or exposure assessment. In this regard, the Action provided qualified and quality-assured model test data from field and laboratory tests. It further classification existing test data with respect to completeness and usefulness for the aforementioned purpose. The uncertainty in the test data were assessed and to a logical degree quantified. The information gaps in required data were reported to scientists engaged in selected experiments in order to redirect funds and to influence pertinent experimental research. In view of that, COST Action ES1006 defined desirable test scenarios for which data may be collected during field and/or laboratory experiments in the future.

A key task of the Action was to identify the main gaps, deficiencies and limitations in the available knowledge and models at the time and to determine the directions for the development of the next generation of models. Future models will not be just computationally faster and provide higher resolution but will have the potential to include substantially more detailed treatment of the source term and intricate processes characterizing the very early stages of an event at distances very close to the release location. Processes like fire and explosion dynamics, source aerosolization and heavy & light gases dispersion were not yet fully understood and often poorly parameterized in present models, thus scientific input was needed for corresponding improvements in future models. A further scientific task that was addressed by the Action was the integration of airborne hazards modelling tools in existing and/or evolving information systems for urban/industrial emergency management. In this context it was important to consider not only the output results of local-scale airborne hazards modelling but also the possibility of information input, fed into dispersion modelling to improve the quality of model results, such as meteorological input data as they are available from localized chemical weather analysis and forecast systems. With the fast-changing information management systems (like GIS), particularly in urban environments, and the development of dedicated decision support systems in mind, it was of special importance to consider the definition and documentation of commonly accepted, bi-directional information interfaces. Collecting and integrating as much as possible information from outside of the Action for example by organizing workshops enabled the information input to the Action to be maximized.

Scientific work plan methods and means

WG 1 – Threats, Models and Data Requirements

WG 1 characterized/categorized existing models and typical release scenarios as well as compiled, evaluated, completed and documented existing test data and defined application – oriented data requirements for further improvements in neighbourhood scale airborne hazard modelling. Specific research questions asked and the tasks that were performed included:

  • What are the currently applied models and methodologies and what is the current state-of-the-art in operational neighbourhood-scale emergency response modelling?
  • Which model types and model systems are currently under development and favoured for future application?
  • How are the emergency response models usually operated (pre-/post-accidental, onsite/off-site, operational/ event driven, manual/automated input, etc.)?
  • How can models be efficiently classified regarding physical background, computational demands and application requirements?
  • What information is available from past events of short-term releases of hazardous agents in urban and industrial environments?
  • Which models are available, how have they been applied during actual accidents or training scenarios and how did they perform?
  • What are the critical and challenging situations identified during a post-accidental analysis of real events?
  • What are the threat scenarios and source terms specified by the different communities involved in local-scale emergency response such as civil protection, homeland security and industrial safety?
  • What data is already available for testing emergency response models?
  • Is existing data of sufficient quality?


Based on previous and ongoing research work, a state-of-the-art report on applied local-scale accidental release modelling was compiled which included the following:  

  • A structured catalogue of threat scenarios and source terms of concern.
  • A collection of complete data sets qualified for testing emergency response models.
  • A glossary of terms to be used throughout model development, evaluation and application.

WG 2 - Test, Evaluation and Further Development

WG2 defined (partially blind) test scenarios and tested and assessed different modelling approaches, In addition it developed scientific strategies for improving the implementation of corresponding tools. Simultaneously working on the testing of tools and models as well as on the improvement of their reliable implementation  increased the efficiency of the Action. WG2 comprised of model developers and model users in order to facilitate a direct information exchange. Within the frame of WG2 the following specific issues were addressed:

  • Quantifying and documenting the performance of modelling approaches with regard to efficiency (speed and computational demands) and reliability (quality of application oriented model results) based on qualified test schemes that will be developed.
  • Quantifying the scatter of results inherent in model results when different models and tools are applied to exactly the same threat scenario.
  • Quantifying the effect of uncertainties in input data (meteorology, release conditions, source term, etc.) on relevant model results such as cloud travel time and location, displacement of the maximum concentrations, dosage or the persistence of hazardous materials in built environments.
  • Developing and testing strategies for defining 'worst case' conditions to be applied when models are operated without detailed knowledge about the release and boundary conditions of an accident/release as usual for emergency response planning.
  • Demonstrating and evaluating the potential of integrated information systems as source of improved input data for simulations.
  • Developing and testing a structured application scheme for providing instant guidance in the selection of optimum simulation strategies for given threat scenario.
  • Developing model/tool-specific user training and guidance documents in the form of a dedicated best practice guideline


WG2 generated the biggest scientific added value of the Action. In this context, the Action’s scientific interest was not to rank or pillory individual modelling concepts but to facilitate open discussions on specific reasons for diverging model results and possible ways for improving modelling quality. The outcomes from WG2 were the following:

  • A critical review of the application-oriented model quality assurance procedures applied.
  • The strengths and weaknesses of particular modelling concepts in order to stimulate further improvement of model quality were identified, quantifed and documented.
  • The requirements for future model developments were identified and recommendations on necessary improvements and extensions were outlined and published.
  • The immediate dissemination of the scientific achievements by annual topic - oriented open workshops and symposia documenting the progress of the Action on a regular basis for a wider scientific- and user-community were facilitated.

The proceedings of the symposia were published as COST documents. The workshops/symposia on local-scale airborne hazards modelling are intended to be continued beyond the duration of the Action. A first version of a best practice manual for the application of neighbourhood-scale airborne hazards models was compiled and released in order to immediately improve the quality of model results.

WG3 - Applicability, Implementation and Practical Guidance

WG3 focused on the practical constraints in the use of local-scale emergency response models. The specific needs of first responders and authorities in charge of neighbourhood - scale emergency response management were taken into account in order to successfully implement the scientific improvements gained by the Action. WG3 worked closely together with the Panel of External Experts to ensure a bilateral feedback between the scientific and emergency management communities. The main tasks covered in WG3 were the following:

  • To collect the requests and demands of the emergency-response experts for improving the practical applicability of the modelling systems.
  • To provide guidance regarding the suitability of different types of model and methodologies for specific problems and at different stages of an incident.
  • To give recommendations on the proper use of different models and methodologies and the reliable assimilation of measured data in the context of emergency response.
  • To identify, characterize, visualize and quantify the uncertainties of emergency response modelling, enabling decision makers to interpret model results properly.
  • To support spokespersons, health authorities, social workers and others by providing reliable, proficient information regarding the consequences of the incident.


End users manuals as well as guidance and training documents were produced. The Action’s intentionwas not only to evolve scientific contents, but also to provide immediate practical tools. This was emphasized by delivering and publishing:

  • A documentation of recommended application - oriented procedures for the use of local scale dispersion models in the context of accidental releases and emergency management.
  • Information on what type of model(s) or approach (es) to be used for which type of release/threat scenario.
  • Quantitative information on the expected reliability / uncertainty of model results with reference to the quality and quantity of the input data available in a given scenario.
  • Practical guidance for the optimum use of models or modelling approaches and the meteorological input data required to improve the quality of model results.

In the last year of the Action, a summarizing report was brought into a final form, peer-reviewed and published. It was the task of the members of the Management Committee (MC) throughout the entire Action to actively promote the tools, strategies and standards developed by the Action not only in their member states or national bodies but also at the level of the European Commission and the European Security Research and Advisory Forum (ESRIF). The commonly accepted quality measures and application recommendations for specific models and modelling strategies put pressure on model developers to carry out and document a commonly accepted quality assurance procedure and to improve modeling capabilities not solely driven by commercial interests. Similarly to previous COST activities, the Action improved the quality of model results and in addition developed the 'culture' of using such tools for example by a scientifically justified selection of proper tools for a specific release scenario.