Missions

Arboviral diseases such as dengue, Zika, or chikungunya are transmitted by mosquitoes of the Aedes genus and affect millions of people every year, currently mainly in urban areas of intertropical regions. In a context of rapid expansion of Aedes albopictus in France, increased arboviral risk, and growing demand for robust and transferable operational tools, UMR IDEES is developing spatial modelling and simulation tools for vector risk assessment.
This effort is grounded in a spatially explicit mechanistic model implemented in R (aspatial version) and Scala (spatially distributed version), simulating at a daily time step the population dynamics (eggs, aquatic stages, adults) represented as interconnected stocks. Transitions (development, mortality, emergence, reproduction) are driven by daily environmental forcings (temperature, rainfall, evaporation) that determine water availability, as well as by the density of potential breeding sites influencing local carrying capacity. The system is distributed across micro-zones characterised by land-use attributes (building density, vegetation, more or less open spaces) and socio-demographic variables.
Developed and calibrated in Bangkok for Aedes aegypti within the ANR MO3 project (2019–2023), the model demonstrated good predictive performance for daily and weekly dengue cases based on two years of sentinel hospital data. As part of the ANRS SEA-ROADS programme (2024–2027), coordinated by Institut Pasteur, the objective is to continue calibration and sensitivity analysis, and to assess the model's transferability to other Southeast Asian metropolitan areas for which vector time series will be available (field collections 2025–2026 within the project). The aim is to consolidate a methodological foundation that can be reused in other geographical contexts, including in France.

Activités

- Continue implementing and maintaining both versions of the model (aspatial R; spatially distributed Scala), including testing and documentation;
- Parameterise biological processes (development, mortality, emergence, reproduction) based on the scientific literature and studies carried out by the project partners;
- Calibrate the model for Bangkok and other SEA-ROADS cities (Hanoi, Ho Chi Minh City, Vientiane, Phnom Penh) using collected entomological time series;
- Perform global sensitivity analyses (Sobol, FAST, and equivalents) to identify dominant parameters and key interactions;
- Quantify uncertainty (parameters, data, structure) and produce uncertainty intervals for key outputs (abundance, biting rate);
- Conduct temporal and spatial validation; leave-area-out validation; inter-city transferability tests;
- Assess predictive performance (errors, bias, ability to reproduce seasonality, peaks, and environmental gradients);
- Test intervention scenarios (breeding-site reduction, larvicidal/adulticidal treatments, timing);
- Ensure reproducibility (data → model → outputs pipeline, HPC execution);
- Establish data–model governance (traceability, metadata, data quality, rights management, sharing with partners);
- Document and disseminate (technical documentation, articles/methods).

Compétences

1. Statistics, calibration, and uncertainty
- Applied statistics (GLMMs, hierarchical models).
- Calibration methods for mechanistic models.
- Global sensitivity analyses (Sobol, FAST, Morris).
- Uncertainty propagation and spatiotemporal validation.
2. Scientific computing
- Advanced scientific programming in R / Python (data pipelines, modelling, visualisation).
- Programming in Scala (or equivalent language) for distributed and high-performance models.
- Version control (Git), documentation.
- Knowledge of high-performance computing (HPC), parallelisation, and automated scenario execution.
3. Entomological modelling
- Biology and ecology of mosquito vectors (Aedes aegypti, Aedes albopictus).
- Mechanistic modelling of population dynamics (compartment/stock-based models).
- Parameterisation and interpretation of entomological indicators (abundance, larval indices, etc.).
4. Cross-cutting and operational skills
- Management of (spatial) big data (quality, metadata, interoperability).
- Interdisciplinary work (geographers, entomologists, epidemiologists, computer scientists, public health actors).
- Ability to translate a scientific model into operational tools (maps, indicators, dashboards).
- Scientific and technical writing (articles, deliverables, user guides).

Contexte de travail

The CNRS research unit IDEES is a multidisciplinary and multi-site SHS laboratory (Caen, Le Havre, and Rouen) that plays a major role in the regional scientific landscape and is involved in numerous national (SFR CIST, GdR MAGIS, Réseau des ISC, SFR SCALE) and international research networks. The unit brings together approximately 120 faculty members, researchers, research engineers/associates, and doctoral students, working across areas recognised in France and internationally, such as spatial modelling and analysis, health and risk, transport and port environments, information and communication technologies (ICT), and socio-territorial transformations.
The recruited individual will be based on the Rouen campus, under the hierarchical supervision of Eric Daudé (CNRS Research Director) and Romain Reuillon (CNRS Research Scientist). Together with Alexandre Cebeillac and Sébastien Rey-Coyrehourq (Research Engineers), the team specialises in geomatics, spatial modelling and simulation, and model exploration and evaluation (OpenMole platform https://openmole.org/).
The recruited individual will also work directly with the project's entomologists, based at Institut Pasteur in Paris (Rick Paul) and Cambodia (Sébastien Boyer), as well as with scientific partners and public health ministries from the four countries involved in the project.

Contraintes et risques

None