HIDALGO 2 Wildfires deployment 2024

• Name: HIDALGO 2 Wildfires deployment 2024 , HiDALGO2
• EuroHPC machine used: Discoverer
• Topic: Computer and information sciences; Physical sciences; Earth and related environmental sciences

Overview of the project

The wildfires use case is part of the HIDALGO2 Center of Excellence project, funded by the European Commission. Two different scenarios has been defined:

Scenario 1. Landscape level. Simulations of wildfire progression, energy release and coupled atmosphere-fire interactions, including release and dispersion of smoke. This scenario takes advantage of the existing coupled models such as WRF-LES-SFIRE, adapting their use to high resolution, for which a careful parameterization of the models is needed to allow the calculation of local scale convective processes. In addition to this, we are also working on coupling our own model, Fire Spread Engine, with WRF. FSE is a fire simulation model that runs on CPU or GPU.

Scenario 2. Settlement level. The main goal is to provide simulations of the fire behavior and fire front-low atmosphere interactions, including perturbation of the air flow and the local effect of buildings (hard geometric bodies) and vegetation (porous media).

In light of the above, the objective is to provide computational capacity for the simulation of wildfire-atmosphere interactions and smoke dispersion at several scales aimed at the assessment of risk and potential impacts over populated areas. Integrated models and results in visually-oriented platforms for analysis, training and education to advance in the understanding and characterization of fire behavior and atmosphere interactions, in particular the fire-induced wind flow patterns and fire behavior at mesoscale and microscale in the vicinity and inside of wildland-urban interface (WUI) areas.

The aim is to progress in the field of model integration under a common, interoperable framework coupling and improving the tools needed for the modeling of lower atmosphere dynamics and the interaction with large fire fronts, for which we plan to use WRF-LES-SFIRE for large fires modeling and WRF outputs as inputs for a coupled OpenFOAM/fireFOAM modelling in WUI areas. These physical models require a detailed description of the governing factors, at several scales, and the underlying formulation of the processes, thus involving highly turbulent thermo- fluid dynamics, which CFD solvers partially calculate at the expense of high computational resources.

How did EPICURE support the project and what were the benefits of the support?

EPICURE provided technical support that enabled the successful deployment and operation of WRF-Sfire across all the EuroHPC JU machines. The assistance included resolving operational issues when utilising thousands of cores, selecting the appropriate partition to run the model, optimising CPU usage for I/O operations in Discoverer and Karolina, and troubleshooting when using a large number of nodes associated with memory management malfunction in LUMI.

With the support received from EPICURE, the HiDALGO2 project was able to save time solving deployment issues and resources optimisation. Additionally, the project gained experience in installation processes using EasyBuild and EESSI.

Additional references

Caballero, D., Salazar, L.L, Rivera, A., Torres, L. (2025) Simulation of Wildfires Using EuroHPC Resources: Challenges and Opportunities. R. Wyrzykowski et al. (Eds.): PPAM 2024, LNCS 15581, pp. 3–16, 2025. https://doi.org/10.1007/978-3-031-85703-4_1