Our team and mission

ESA maintains a world-leading Science Programme with missions in heliophysics, planetary science, astrophysics and fundamental physics. Its mission is to 'empower Europe to lead space science'. The Department for Science and Operations hosts the scientists and engineers that oversee the space missions, from study to end of operations. It develops the science operation systems for the missions and operates the missions in space, as well as archiving and curating their data during operations and beyond. Our main objective is to maximise the scientific output of the missions for the benefit of humankind.

Gaia is ESA’s Billion Star Surveyor mission aimed at creating an astrophysical catalogue of unprecedented completeness and quality. Four highly successful data releases (DR1, DR2, EDR3 and DR3) have been published so far and two more are planned (DR4 and DR5) by the end of the decade. A record number of scientific publications has already resulted from the public (E)DR1-3 data, and by this measure, Gaia is considered the most productive space science mission ever. For an overview of the mission, see: https://www.esa.int/Science_Exploration/Space_Science/Gaia and https://gaia.esac.esa.int

Gaia’s data processing task is large and complex and undertaken by the pan-European Data Processing and Analysis Consortium (DPAC) of which ESA’s Science Operation Centre (SOC) is an integral part. The SOC is located at ESA’s European Space Astronomy Centre (ESAC) near Madrid and composed of a team of about 25 engineers and scientists working on specific data processing and other mission aspects. If selected for the position, you will work as part of the Gaia SOC with access to all mission experts, both local and in the wider DPAC community.

One of the central elements of the data reduction is the Astrometric Global Iterative Solution (AGIS). AGIS produces and delivers the core astrometry data products of Gaia. The AGIS team members are distributed across ESAC, Lund University, the University of Heidelberg and Technische Universität Dresden. The generation of the solution takes place at ESAC. The team includes global experts in astrometry and calibration, software engineers and experts in data science. We are currently working on the production of the solution for Gaia DR4 and are already looking into improving the calibration model for DR5, planned for release by the end of 2030. 

You are encouraged to visit the ESA website: http://www.esa.int

Field(s) of activity/research for the traineeship

One of the most challenging tasks that we need to tackle in the software producing Gaia’s astrometric solution is creating a calibration model accurate enough to take into account the subtle effects which may have an impact on the quality of the solution at the micro-arcsecond level. These can have a real physical cause (imperfections in the hardware or CCD manufacturing/behaviour) or can be caused by limitations of the Point Spread Function models needed to extract basic input data for AGIS from Gaia’s raw observation data.

Among the AGIS outputs we use to assess the quality of the solution are residuals. These are the differences between the values predicted by our observation models and the actual observation data itself. Residuals are a collection of many data points (one for each Gaia observation) and depend on hundreds of parameters. The residuals can be very noisy up to a certain level, but hidden in the noise one can find dependencies with some of the other parameters to which we have access (field of view, CCD, time, etc.). Many of the residual parameter correlations are known and we correct them by introducing changes in our model. However, residuals can be correlated with one or several parameters in a huge number of ways, offering possibilities to apply new techniques such as machine learning, regression analysis, recursive partitioning or neural networks to discover unknown dependencies.

As a Young Graduate Trainee, you will be given the task of developing a framework which takes the residuals as input and uses state-of-the-art machine learning techniques to analyse them, thereby discovering the patterns and correlations between them and the parameters we know about. The goal is to find unknown correlations. Once we discover a new correlation, we will try to explain its origin and consider how it can be modelled.

Once the framework is in place, we would like to use it regularly when solutions are generated after updating the model or when new data is processed. The output of your work will therefore contribute to the improvement of the Gaia astrometric catalogue which will be the reference catalogue in astronomy for the decades to come.

Technical competencies

Behavioural competencies

Result Orientation

Operational Efficiency

Fostering Cooperation

Relationship Management

Continuous Improvement

Forward Thinking

Education

You should have just completed or be in the final year of your master’s degree in a scientific or engineering field.

Additional requirements

You should have a strong background in data science and/or data mining with some experience in the management of large datasets using Python and familiarity with the libraries commonly used in machine learning, AI or pattern discovery in data mining.

You should have good interpersonal and communication skills and should be able to work in a multi-cultural environment, both independently and as part of a team.

The working languages of the Agency are English and French. A good knowledge of one of these is required. Knowledge of another Member State language would be an asset.

During the interview, your motivation and overall professional career goals will also be explored.