Portrait of Stefan Graham

Stefan Graham

PhD Student · Acoustic Technology group, DTU Electro

Scientific machine learning for acoustics and wave-field modelling. Previously at the European Space Agency, working on the AIS payload of Sentinel-1C and -D.

About Me

I am a PhD student at the Technical University of Denmark (DTU), where my research focuses on the application of scientific machine learning in acoustics.

Before starting the PhD, I worked at the European Space Agency (ESA), where I led the commissioning activities for the Automatic Identification System (AIS) payload embarked on Sentinel-1C and -D.

In my spare time, I enjoy calisthenics, running, rock climbing and playing guitar.

Research Interests

  • Scientific machine learning
  • Acoustic wave-field modelling
  • Numerical methods for differential equations
  • Inverse scattering problems
  • Signal processing & wireless communication
  • Synthetic aperture radar (SAR) & InSAR

Publications

World map of AIS detections along Sentinel-1C acquisition passes, coloured by SAT-AIS, legacy AIS and common AIS, with the start and end of each acquisition marked.

Conference paperIGARSS · Washington DC · 2026

Sentinel-1C AIS Commissioning and Performance Results

S. Graham, A. Solana, E. Rondoni, F. Frattasi, T. Bollian, N. Miranda, D. Bibby, R. Torres

Introduces the AIS instrument on Sentinel-1C, its field of view and the detection results from commissioning acquisitions. It discusses AIS message collisions and radio-frequency interference, and presents an improved on-ground processing algorithm that significantly increases detection performance.

Map comparing vessels detected by Sentinel-1C and RCM-3 on 27 April 2025 over Central and South America: vessels seen only by Sentinel-1C, only by RCM, and by both, with both satellites' orbits.

Invited paperEUSAR · Baden-Baden · 2026

Sentinel-1C AIS Cross-Validation with the RADARSAT Constellation Mission

S. Graham, M. Rashid, E. Swift, C. Gierull, D. Geudtner, A. Boyer, M. Lapointe, M. Pinol Sole, S. Cote, R. Torres

As part of the Sentinel-1C AIS commissioning, its AIS data were cross-validated against the RADARSAT Constellation Mission using three co-located SAR and AIS acquisitions off the coast of Ecuador. The results show excellent payload performance and the benefit of multiple platforms for AIS detection.

A Turing machine running on a Sierpinski network: the coloured graph on the left and a plot of the machine's evolution on the right.

Summer school projectWolfram Community · 2023

Turing Machines on Graphs

S. Graham · Wolfram Summer School 2023

Ordinary Turing machines run on an infinite one-dimensional tape. This project explores Turing machines on finite cyclic graphs — a cycle, a torus and the Sierpinski network — and shows that the Sierpinski network gives the most complex and varied behaviour.

Conference paperIGARSS · Kuala Lumpur · 2022

Copernicus Sentinel-1 Next Generation Mission: Enhanced C-Band Data Continuity

R. Torres, D. Geudtner, M. Davidson, D. Bibby, I. Navas Traver, A. I. García Hernandez, G. Ladureé, J. Poupaert, M. Cossu, M. Touveneau, S. Graham

Presents the planned Copernicus Sentinel-1 Next Generation mission, which will continue C-band SAR data beyond 2030. It covers the envisaged SAR performance enhancements, candidate SAR system concepts, and the planned AIS and VHF Data Exchange System (VDES) payloads for maritime surveillance.

Figure 1.1 from the thesis: Sentinel-1 images are used as input to the persistent scatterer algorithm, which outputs a deformation map and a displacement time series for each point.

MSc thesisDTU Space · 2020

Implementation of a Persistent Scatterer InSAR Algorithm for the Measurement of Ground Subsidence

S. Graham · Supervised by J. P. Merryman Boncori

A Persistent Scatterer Interferometry (PSI) algorithm for Sentinel-1 data that maps linear ground deformation and the displacement history of individual points. It was validated on simulations, a corner-reflector experiment (0.6 mm RMS residual) and the sinking town of Thyborøn, Denmark (1.5 mm/year RMS against levelling).

Full list on Google Scholar and ORCID →

Experience & Education

Jan 2026 – Present

PhD Student — Technical University of Denmark (DTU)

Acoustic Technology group, DTU Electro, Denmark

  • Research into scientific machine learning for wave-field modelling.
2020 – Dec 2025

Performance Engineer — European Space Agency (ESA)

European Space Research and Technology Centre (ESTEC), Netherlands

  • Applied advanced wireless communication techniques to enhance AIS detection and demodulation.
  • Developed SAR-AIS data fusion tools based on constant false alarm rate (CFAR) hypothesis testing.
  • Led the AIS verification activities in preparation for the launch of Sentinel-1C/D, including implementation of the data validation toolboxes and management of external industrial partners.
  • Delivered algorithms to the ground segment team for processing the AIS data and defined the end-user product specification.
  • Initiated and coordinated collaborations with universities, research institutes and other teams within ESA, leading to improved knowledge exchange.
  • Defined and reviewed requirements for the Sentinel-1 Next-Generation mission (2032), ensuring alignment with technological advancements and mission objectives.
2014 – 2020

BSc & MSc, Earth and Space Physics and Engineering

Technical University of Denmark (DTU), Copenhagen

MSc thesis: Implementation of a Persistent Scatterer Interferometric-SAR (PSI) algorithm for the measurement of ground subsidence

BSc thesis: Optimal Interpretation and Decomposition of X-Ray Fluorescence Spectra