CROSS Harmonization & HPC modelization of FOREST Datasets


The pilot FRAME – Forest fiRes Advanced ModElization – will focus on the realistic simulation of forest fires behaviour and spreading according to high spatial and time resolution data available or enhanced from public open data infrastructures. The attained System will consider both fire spread parameters and performance of firefighting measures, as modelled in previous experiences under AF3 project. This approach involves the need of high calculation capacities as derived from:

  • Needs of high time (sub-second) and spatial (sub-metric) resolution.
  • Big loads of data exchange between fire, scenario (high-resolution digital terrain model, infrastructures), fuel elements (amount, moisture content, distribution, structure, type), environment weather conditions (wind, humidity) and interactions with agents modifying any of the factors affecting propagation (suffocation, fuel depletion, water dropping, burning), at each iteration.

Under the use-case, a methodology will be developed and demonstrated for the use of interoperable forestry-oriented datasets for feeding efficient fire behaviour (and linked) models at different work-scale levels, useful for training Crisis Management Mechanisms in the Iberian Peninsula. Some of the specific data sources to be used within FRAME are shared datasets as Spanish IDE, Spanish Forestry Map, Portuguese IDE SNIG2020, SIGN II or OTALEX.

Selected datasets will be implemented together with a collection of models driving the simulations. Digital Terrain Models developed for modulation of scenarios will benefit of Public Spatial Data infrastructures of Spain and Portugal (IDEE, SNIG), and the usefulness of trans border initiatives will also be explored (OTALEX, SIGN II). Mechanisms for scaling and harmonizing datasets available from them will help generating crisis scenarios overcoming the limitations of the regional and national boundaries. Common protocols (Incident Command System applications) and European management systems (CECIS and derived UE structures of experts, modules and other capacities for international collaboration) will work as reference to set the goals and guidelines for harmonization of operational particularities.

Datasets harmonization is needed for feeding fire behaviour and related models in the most suitable work scale regardless data source (different data sources, different paradigms: weather, terrain, fuel). Therefore, an identification and classification of public datasets and data sources will be completed for the Iberian Peninsula, and sorted and adjusted concerning their utility as inputs for each model. A preliminary classification and sorting of the models involved would provide the system requirements concerning datasets. Those datasets will provided required input information as described above of scenario (digital terrain model and infrastructures), fuel elements (amount, moisture content, distribution, structure, type) and weather conditions (wind, humidity).

The propagation model used in ERVIN firefighting virtual training environment, developed by TRAGSA, will be used as the base model for FRAME simulations, automatizing their feeding from open data sources mentioned before.

ERVIN firefighting virtual training environment, developed by TRAGSA
ERVIN firefighting virtual training environment, developed by TRAGSA
ERVIN firefighting virtual training environment, developed by TRAGSA

Supercomputing (HPC) will allow segregating heavy calculations and speeding up the performance of simultaneous simulations, that will be provided to the Graphical User Interface outputs based on EMERCARTO emergency management system, developed by Tragsa as well.

Imagen Emercarto
EMERCARTO GUI screenshot

The expected output of FRAME pilot is to present a series of maps showing fire behavior projections in several isochrones according to the storyline defined on an input stage, for the range of different weather conditions provided and with a geographical scope, which comprehends the Iberian Peninsula (Portugal and Spain).

Many iterations over the same initial points of ignition could be executed, varying, not only weather or fuel settings, but expected combat actions that will be performed on the wildfire scenario proposed.

These results will provide a useful management, prevention and training tool for Public Administrations and Emergency professional, employing the emergencies and wildfire management system EMERCARTO (developed by Tragsa Group) as the user interface.