Environmental Modeller

Environmental Modeller needed to contribute to the development and maintenance of the Agricarbon Ecosystem Model (AEM) using Python.

Model Components & Integration:

Working with agricultural ecosystem models (AEM) including plant growth models, soil organic carbon models, soil water models, mineral nitrogen models, and grazing models.

Implementing and maintaining the integration between different AEM components, ensuring seamless data flow between plant growth, soil carbon, water, nitrogen, and livestock models within the Bayesian data assimilation framework.

Technical Development:

Contributing to the development and maintenance of the Bayesian data assimilation framework that underpins the AEM, ensuring robust uncertainty quantification and model calibration.

Configuring, running, and extending existing model components such as LINTUL-5 (arable crops), LINGRA (grass), RothPC-N (soil organic carbon and nitrogen), developing Python implementations that maximise the benefit of our access to the world's largest soil carbon database.

Evaluating and implementing machine learning and statistical models using Python libraries to enhance overall accuracy and predictive power, potentially as part of ensemble modelling approaches.

Code Quality & Collaboration:

Ensuring all modelling code meets high standards for reliability, performance, and maintainability, with comprehensive testing and documentation.

Working closely with our Principal Soil Modeller to translate scientific requirements into robust technical solutions, providing programming expertise to support complex modelling challenges.

Data & Validation:

Designing and implementing automated testing frameworks to validate and improve model performance, ensuring statistical rigour in all implementations.

Communication & Documentation:

Producing comprehensive technical documentation, code comments, and user guides for all modelling implementations.

Supporting collaborative research initiatives by providing technical implementation of novel modelling approaches and contributing to peer-reviewed publications.

Communicating technical modelling concepts and results to both technical and non-technical audiences, including partners and stakeholders.

Skills and Experience:

• Must have:
• Advanced Programming Skills: Extensive experience in Python programming for data science and environmental modelling, including proficiency with scientific libraries (NumPy, SciPy, Pandas, scikit-learn, GeoPandas) and Bayesian statistical libraries (PyMC or similar).
• Environmental Modelling Experience: Proven experience developing and working with ecosystem models or related areas.
• Data Science Proficiency: Extensive experience with machine learning techniques and their application to environmental data, including model validation and statistical analysis.
• Code Quality Focus: Experience with software development best practices including version control (Git), testing frameworks, and code documentation.
• Problem-Solving Skills: Excellent analytical and problem-solving abilities with extreme attention to detail and a rigorous approach to model development.
• Educational Background: Master's degree or PhD in Data Science, Environmental Science, Computer Science, or related field with a strong focus on modelling and programming.

Nice to have:

• Experience with Bayesian methods and data assimilation frameworks.
• Familiarity with Soil carbon (e.g. RothC) and crop growth models (e.g. LINTUL, WOFOST, DSSAT, APSIM) or grassland (e.g. LINGRA) models, and/or integrated agricultural system models.
• Knowledge of nitrogen cycling and soil-plant-atmosphere interactions.
• Familiarity with data assimilation using satellite-derived data (e.g. Leaf area index, canopy cover).
• Experience with cloud computing platforms for large-scale data processing (AWS, Azure, GCP).
• Track record of peer-reviewed publications in relevant fields.
• Geospatial data handling experience (e.g., GeoPandas, DuckDB, etc.).
• Familiarity with containerisation and deployment technologies (Docker).