The Graz Lagrangian Model - GRAL - was initially developed in 1999, and has been used extensively in regulatory assessments and scientific studies.
The initial driver for the development of GRAL was the need for a model that could deal with the frequent low-wind-speed conditions (< 1.5 m s-1 for up to 90 per cent of the time) in the inner-Alpine basins of Austria. Another important feature of GRAL is the ability to deal with the dispersion of pollutants emitted from road tunnel portals. A series of national research projects resulted in a new physical modelling approach for use in GRAL, and this approach is still unique among dispersion models.
Over the years the capabilities of GRAL have been extended, and the current version of the model can simulate the following:
- Dispersion of chemically non-reactive pollutants.
- Computation of so-called odour-hours based on a recently developed concentration-variance model.
- Dry and wet (only in transient mode) deposition and sedimentation.
- Dispersion from road tunnel portals. GRAL fulfils the requirements of the Technical Guideline RVS 04.02.12 in Austria.
- Dispersion over the full range of wind speeds without any lower threshold, and for all stability conditions.
- Dispersion in built-up areas, including building downwash effects.
- Dispersion of stack emissions, taking into account temperature and exit velocity.
- Dispersion in complex terrain, allowing for the effects of buildings.
- Decay rates (e.g. bacteria die off, radioactive decay)
- Flow and dispersion within vegation layers
- The model can handle steady-state (standard mode) as well as transient simulations (e.g. puff releases)
The effect of buildings and vegetation on dispersion is taken into account using a micro-scale flow-field model. This is fully integrated into the GRAL code, and is therefore not visible to the user; it is automatically launched whenever buildings or vegetation layers are added to the model domain. In the case of complex terrain, GRAL can be coupled with the prognostic, meso-scale wind field model GRAMM (‘Graz Mesoscale Model’). Both GRAL and GRAMM are parallelised and can be run on both Windows and Linux operating systems. The models can be operated through a graphical user interface (GUI) which has been thoroughly tested for Windows operating systems. Since 2017 a LINUX version for the GUI is available, though, it is not as intensively tested as the Windows version.
There is no limit to the number of separate emission sources that can be included in a GRAL simulation. The lower bound for the horizontal grid size is 2 m, and there is no upper bound. The scale of application ranges from individual streets (e.g. street canyons) to urban agglomerations that are several tens of kilometres across. At all scales the effects of buildings and/or topography (e.g. cold air drainage flows) on dispersion are taken into account.
It should be noted that it is not possible to include chemical reactions. The conversion of NO to NO2 for example, can be taken into account by applying empirical relationships between NOx and NO2 in postprocessing mode.
Quality assurance is central to the ongoing development of GRAL, based on these fundamentals:
- Regular reports detailing the model physics, and the publication of results in international peer-reviewed scientific journals.
- Comprehensive documentation of the software, with version control.
- A handbook for the GUI that includes hints and recommendations for good practice.
- Validation of every update using 30 different data sets (field experiments, wind tunnel experiments, air quality measurements), as published in the GRAL documentation.
Tutorial videos about the GRAL model are also available: https://www.youtube.com/watch?v=vfEVl-j4P5s