Clouds represent a major source of uncertainty in Earth system models. In particular, the altitude and vertical distribution of clouds strongly influence their radiative properties and impact Earth’s climate. In this work, we derive 3D cloud maps from 2D satellite imagery and sparse measurements of atmospheric profiles. By encoding space and time, we create a ‘geospatially-aware’ model with reduced regional prediction biases.

Anna Jungbluth

Anna Jungbluth

Anna Jungbluth is a Research Fellow at the European Space Agency (ESA), working in the Climate Team on machine learning applications for Earth Observation. She holds a PhD in Physics from the University of Oxford, where her research focused on enhancing the efficiency of solar panels through charge-transfer state analysis in organic solar cells. Her passion for machine learning and space research began in 2018 when she won the UK finals of the ESA-sponsored Act in Space Hackathon. Later, she contributed to the NASA-funded Frontier Development Lab (FDL), developing machine learning models for solar research. Anna is also committed to mentorship and promoting diversity in STEM fields.

CONTACT INFO:

SUMMARY OF ACTIVITIES/INTERESTS

  • Earth Observation
  • Climate