Required Education : Master 2 or Engineering schools
Start date : 1 October 2021
Mission duration : 36 months
Deadline for applications : 1 July 2021
Salary : about 2525 € gross/month
The present project aims at improving the biomass fuel representation in the coupled atmosphere/fire model, Meso-NH/Blaze, and at analyzing its sensitivity to an ensemble of plausible biomass fuel characteristics. This will involve improving coupling of land surface properties as represented in BLAZE and SURFEX (the land surface component of Meso-NH), and the development of scaling relationships to relate high-resolution fire simulation in Meso-NH/BLAZE to large-scale SURFEX model outputs and databases. The resulting ensemble will be used to design a strategy to improve current wildfire risk assessment tools based on the coupled system Meso-NH/Blaze.
Candidates should have a good scientific and technical background, with skills in programming, and interests in fluid mechanics, atmospheric sciences and land surface modeling. No specific knowledge on wildland fires is needed, but the motivation to interact with different scientists across disciplines (CERFACS, CNRM and beyond) would be advantageous. Familiarity with Python and data science experience would be an asset. Written and spoken English is desirable.
The selected candidate will become a PhD student registered at Université de Toulouse for three years and employed by CERFACS. The proposed salary corresponds to the standard conditions at CERFACS.
If you are interested in this PhD offer, please send your CV along with a cover letter and 1-2 recommendation letters before July 2021 to:
- Mélanie Rochoux, firstname.lastname@example.org
- Benjamin Sanderson, email@example.com
- Patrick Le Moigne, firstname.lastname@example.org
Predicting wildfire behavior is a research area that has been attracting growing interest in recent years due to the large number of extreme fire events, also referred to as megafires, that occurred in multiple ecosystems. Extreme fire behavior  has been partly attributed to climate change, which has the potential to modify the production of large amounts of flammable fuels, and enhance synoptic atmospheric patterns that are critical for wildfires. Mediterranean-type ecosystems may be particularly exposed to intense wildfires under climate change . Recent events like 2019/2020 Australian bushfires  have shown that climate models do not reliably represent the fire risk with strong implications for public safety, ecosystem persistence and air pollution . Better representing fire processes in numerical models at the scale of an event is therefore invaluable to improve wildfire risk assessment.
Wildfire behavior can be modeled at a range of scales, from landscape scales (from a few tens of m2 up to several km) to global climate scales (where wildfires are statistically represented in grid cells representing hundreds of km). Landscape-scale models represent a propagating fire front over the land surface, which separates burnt and unburnt vegetation. The local spread rate of the fire front is parameterized with respect to biomass fuel properties, terrain slope and surface wind conditions. The large heat fluxes released by the wildfires can modify the flow dynamics and induce retroactive feedback on the fire behavior. The fire-induced flow can be captured by coupling the fire behavior model with an atmosphere model at high resolution (10-100 m), such as in Meso-NH/Blaze. In this two-way atmosphere/fire coupling system, the fire processes are represented based on the fire front paradigm, but the atmospheric processes are fully resolved using Navier-Stokes equations .
The coupled model Meso-NH/Blaze has been validated against data from an experimental grass fire known as FireFlux I. There is now a need to extend the coupled model to a variety of case studies, involving different types of biomass fuels and different atmospheric conditions, in order to further evaluate its capacity to represent the complexity of the processes driving wildland fires.
Figure – Simulation of the FireFlux I controlled burn using Meso-NH/Blaze . Left: Visualization of the fire front propagation at the land surface in red and of the smoke plume (gray scale). Right: Instantaneous snapshot of the vertical wind and density anomaly at 10-m horizontal resolution. Source: PhD thesis Aurélien Costes (CNRM/CERFACS).
Main scientific issues and objectives
One current limitation of the coupled model Meso-NH/Blaze is the simplified representation of biomass fuel properties (dead/live fuel moisture content, surface loading, bulk density…) and of the vegetation. Given these limitations, the present PhD thesis has two main objectives.
- Improving the representation of the biomass fuels in Blaze, to have a physically consistent representation with a land surface model accounting for the carbon cycle such as SURFEX/ISBA [6,7] and benefit from soil moisture and leaf area index information that are assimilated in SURFEX/ISBA 
- Analyzing the sensitivity of the coupled system Meso-NH/Blaze (in terms of fire front propagation and fire-induced atmospheric flows) to biomass fuel types for different fire intensities and atmospheric conditions using ensembles
- Designing a strategy to derive fire risk indices from the coupled system Meso-NH/Blaze
This work will be done in close collaborations between CERFACS (Mélanie Rochoux, Benjamin Sanderson), Météo-France (Patrick Le Moigne, Christine Lac, Valéry Masson, Mathieu Regimbeau) and Université de Toulouse (Rosie Fisher).
 Werth et al. (2016). Technical report. https://www.fs.usda.gov/treesearch/pubs/50530
 Ruffault et al. (2018). Nat. Hazards Earth Syst. Sci. https://doi.org/10.5194/nhess-18-847-2018
 Boer et al. (2020). Nat. Clim. Change. https://doi.org/10.1038/s41558-020-0716-1
 Sanderson and Fisher (2020). Nat. Clim. Change. https://doi.org/10.1038/s41558-020-0707-2
 Costes et al. (2021). In review for publication in Fire Saf. J. https://hal.archives-ouvertes.fr/hal-03012049/
 Masson et al. (2013). Geosci. Model Dev. https://doi.org/10.5194/gmd-6-929-2013
 Delire et al. (2019). J. Adv. Model Earth Sy. https://doi.org/10.1029/2019MS001886
 Barbu et al. (2014). Hydrol. Earth Syst. Sci. https://doi.org/10.5194/hess-18-173-2014