A course dedicated to emerging data-driven approaches in geosciences

Today, geosciences generate ever-increasing volumes of data from field measurements, sensors, imagery, numerical models and geographic information systems. Making effective use of these data has become a key challenge for understanding natural processes, managing resources and supporting decision-making.

In this context, Data Science and Artificial Intelligence methods are playing an increasingly important role. However, their implementation requires a sound understanding of data, models and their limitations.

By the end of the programme, all participants will be able to understand data-processing workflows, critically interpret results. Professionals will be better equipped to effectively collaborate with data scientists and modelling specialists; doctoral candidates and students will have the foundations to start applying these methods in their own research.

Skills and autonomy for research and professional practice

Upon completion of the course, participants will be able to:

• understand the main stages of a data-processing workflow, from acquisition to modelling;
• identify the most appropriate methods for a given geoscientific problem;
• assess the performance and limitations of machine learning models;
• critically interpret results produced by artificial intelligence methods;
• communicate effectively with data scientists, modellers and data engineers;
• take their first independent steps in applying data science and AI methods to their own research or professional projects.

The objective is not to train AI specialists in a few days, but rather to provide participants with the foundations needed to integrate these tools effectively into professional and research contexts.

A progressive programme combining theory and practical applications

Delivered over five days and comprising 27 hours of instruction, the course combines lectures and feedback from professionals working in geoscience and enrinonmental sectors.

Topics covered include:

• fundamentals of the data ecosystem;
• data cleaning and preparation;
• basics of databases and data management;
• exploratory data analysis and data visualisation;
• principles of machine learning;
• fundamentals of deep learning;
• critical evaluation of models;
• introduction to digital twins and the integration of physical knowledge into AI models.

Instructors

The course is coordinated by Anne-Laure Argentin, lecturers at ENSEGID - Bordeaux INP and researcher at the EPOC Laboratory (CNRS UMR 5805). Classes will be delivered by ENSEGID faculty members and researchers, and will be complemented by guest speakers from industry and research, illustrating real-world applications in fields such as hydrology, water resource management, the oil and gas sector, and environmental modelling, providing participants with practical insights into current challenges and professional practices.

Practical information

• Dates: 4 - 8 January 2027

• Location: On-site at ENSEGID, Pessac (France)

• Format: In-person - 27 hours of training

• Language: English (course materials in English)

• Number of participants: 15–20

• Application deadline: 30 October 2026

Applications will be reviewed on a rolling basis - early submission is encouraged, particularly for participants travelling from abroad.

Fees

• PhD students (all institutions) and self-funded participants: €450
• Professionals funded by their organisation: €700

Registration fees include course materials, access to the datasets used during the training sessions, and a certificate of attendance.

Application requirements

Applicants are invited to submit:

• Full name and affiliated institution/organisation;
• Status: PhD student, professional, or other;
• Research topic or area of expertise;
• Python proficiency level: beginner, intermediate, or advanced;
• Curriculum vitae (CV);
• Cover letter outlining your current research or professional context, your motivation for attending the course, and the objectives you hope to achieve through participation.