And here are a few more photos of the model.

Two versions of my favourite ship, RV #Polarstern: one that I worked on during an #Arctic expedition in 2022 and one that I have just finished building from bricks. Evidently, I'm not too old for this!

I am happy to have contributed to @noragourmelon.bsky.social new paper benchmarking 22 deep learning models for SAR glacier front monitoring. Best AI model averaged 221 m error vs 38 m for humans – careful validation is still essential before replacing manual delineation. doi.org/10.1109/TPAM...

Just wrapped up an exciting trip to the ESA CCI colocation meeting at ECSAT! This was a fantastic occasion to meet the European climate data community, exchange insights across different ECV projects (and programs), and bridge the gap between satellite observations and climate modelling.

Last Friday, my amazing colleagues at @tudresden.bsky.social gave me this very fancy hat! 🎓 #PhDone I’m incredibly grateful to them and to everyone else who has supported me over the past few years. Very excited for the next adventures to come.

Just wrapped up an exciting trip to the #CopernicusClimate GA! This was a fantastic occasion to meet the C3S team and connect with the wider climate data community. Additionally, I had the change to present on ice sheet monitoring and how our ECV program is facilitating and advancing these efforts.

This week I am excited to start a new chapter @copernicusecmwf.bsky.social where I will be working as a Scientific Officer in the Climate Intelligence team. In this new role, I'll be developing, evaluating and promoting the use of satellite-based ECV datasets and related services.

The year 2025 is off to a very exciting start for me. Last week, I finally submitted my PhD thesis. It has been a wonderful 5 years @tudresden.bsky.social and @awi.de and I am very grateful to many colleagues and friends who have supported me along the way.

This dataset covers the period from 2013 to 2023 and includes thousands of calving front entries, achieving a sub-seasonal temporal resolution. Here is an example for the Larsen-B embayment.

Building on our large base of Greenland training data, we applied our proven deep learning processing system to 42 key glaciers on the Antarctic Peninsula Ice Sheet.

Last week I had the opportunity to present our latest research to the ice altimetry community at the #CRYO2ICE symposium. I talked about AWI-ICENet1, our new AI-based retracking algorithm for ice altimetry (doi.org/10.5194/tc-1...). I took home fantastic feedback and a lot to think about.

Home sweet home after our exciting field campaign in East Antarctica carrying out geodetic and geophysical measurements. Our results will contribute to a better understanding of the interactions between solid earth, ice sheet and ocean. See our reports for details: tu-dresden.de/bu/umwelt/ge...

Our project concentrates around a field camp at Gaussberg. During this time we will carry out a wide range of geodetic, geophysical and photogrammetric measurements. This survey will allow us to determine changes in ice height and speed over a period of 120 years. 3/4

Gaussberg was discovered in 1902 by the first German Antarctic Expedition under Erich von Drygalski, who named it after his expedition ship "Gauss". Remarkably, during this expedition an extensive surveying programm of Gaussberg and its surrounding glaciers was carried out. 2/4

Tomorrow it finally starts! On board RV Polarstern, I'm heading for East Antarctica. This expedition will link ice sheet instabilities, ocean circulation and geoscientific fieldwork. As part of this, our research project is focussing around one site in particular – the extinct volcano Gaussberg. 1/4