Postdoc: Weakly supervised ML-based Earth observation for climate extreme impact quantification
The postdoc will develop an advanced Earth observation (EO) machine learning (ML) framework to identify, geolocalize, and quantitatively estimate the physical and socio-economic impacts of climate extremes such as floods, storms, wildfires, and landslides, using satellite image sequences. A primary bottleneck in climate adaptation is that historical impact data is plagued by annotation sparsity, which this project aims to address by developing annotation-efficient, weakly supervised ML-based Earth observation models that leverage unstructured data as a source of scalable distant supervision. The framework will incorporate uncertainty quantification techniques to ensure that AI-derived impact estimates are reliable and actionable for decision-makers. This work is hosted by RISE / Climes, a research and training platform building an interdisciplinary field on how climate extremes affect people, ecosystems and infrastructure.
This position is located in Gothenburg, Sweden.
Key qualifications:
- A PhD in Computer Science, Machine Learning, Data Science, Applied Mathematics, Physics, or a highly quantitative equivalent.
- Strong theoretical understanding and hands-on coding mastery of deep learning architectures using PyTorch or TensorFlow, with a proven track record in computer vision or spatio-temporal modeling. Experience working with machine learning for remote sensing data.
- Excellent communication skills in English (written and spoken) are required.
- Nice-to-haves: Prior experience handling multi-spectral geospatial formats, radar (SAR) data, or standard satellite pipelines (Sentinel, Landsat/HLS, MODIS/VIIRS). Familiarity with Parameter-Efficient Fine-Tuning (PEFT/LoRA frameworks), out-of-distribution detection, change-detection tasks, or single-pass Uncertainty Quantification methods (SNGP, Evidential DL, Conformal Prediction). Peer-reviewed publications in top-tier ML conferences or major remote sensing journals. Proficiency in Swedish or other Nordic languages is a welcome plus for stakeholder communication.
Deadline: September 15th, 2026