Project “Would the Northern European Enclosure Dam really protect Sweden from sea level rise? (NEEDS)”, funded by FORMAS grant 2020-00982
Ca 4 million SEK; I am the PI. Co-I is Heather Reese, Department of Earth Sciences, GU.
January 2021 – December 2025

The aim of the project is to determine what causes flooding in Sweden: remote effects that could be blocked out by distant sea walls, or local effects such as precipitation or tides. We’re using a combination of all data sources (in-situ, remote sensing, models) that we analyse with help from Machine and Deep Learning.

Team:

• Postdoc Zhi-Bo Li, since September 2024;
• Postdoc Lea Poropat, June 2021 – September 2023;
• Myself, only for supervision at first, now working 100% of my research time on this project since November 2024;
• Master’s student Linn Carlstedt, May 2022 – March 2023 (thesis completed with honours);
• Master’s student David Ek, April 2021 – April 2022 (thesis completed with honours);
• Heather Reese;
• Collaboration with Dani Jones and Scott Hosking from the British Antarctic Survey AI Lab.

Publications:

• Li et al. [preprint] in which we quantify trends in the winds that we found to be most responsible for sea level extremes, expanding to the entire northern hemisphere.
• Heuzé et al. (2025) in which we determined the drivers of extreme sea level events around the wider North Sea – Baltic region, using LSTM and Random Forest.
• Poropat et al. (2024) in which we use Gaussian mixture modelling to determine the areas of the wider North Sea whose sea level covary, at the monthly time scale.
• Linn Carlstedt’s Master’s thesis, in which she determined the drivers of hourly sea level variability in the North Sea in general and on the Dutch coast in particular, using an LSTM network.
• David Ek’s Master’s thesis, in which he determined the drivers of monthly sea level variability in the small but extremely complex “belt” region between Denmark and Sweden, using traditional statistics and machine learning.

Conference, seminar, or outreach presentations related to this project:

• 22 Sept 2025, “Hvordan kan Machine Learning bidrage til styrket klimatilpasning i Danmark?” – DMI webinar for the Danish government
  Lea Poropat, talk: Finding drivers of sea level variability with machine learning
• 18 Sept 2025, SMHI professor inauguration, Göteborg (Sweden)
  Céline Heuzé, panel discussion
• 13 May 2025, Climate AI Nordics workshop, Göteborg (Sweden)
  Céline Heuzé, poster: Drivers of high frequency extreme sea level around Northern Europe – Synergies between recurrent neural networks and Random Forest
• 27 April – 2 May 2025, EGU25, Vienna (Austria)
  Zhi-Bo Li, poster: Is Europe becoming stormier? Extratropical cyclone clustering over the last century 
• 8-12 May 2023 Liège Colloquium on Ocean Dynamics – Machine Learning and Data Analysis in Oceanography, Liège (Belgium)
  Lea Poropat, talk: Finding drivers of sea level variability with neural networks
• 24-28 April 2023, EGU23, Vienna, (Austria)
  Lea Poropat, talk: Finding regions of similar sea level variability with the help of a Gaussian Mixture Model
  Linn Carlstedt, talk: Drivers of sea level variability using neural networks
• Vetenskapsfestivalen 2023, Göteborg (Sweden)
  Lea Poropat, outreach presentation: Investigating sea level changes: Insights from Artificial Intelligence (Soapbox Science)
  Lea Poropat, outreach presentation: Studying sea level changes with the help of artificial intelligence (Science Roulette)
• 10 March 2023, RCG seminar, Göteborg Universitet (Sweden)
  Lea Poropat, seminar: Unsupervised classification of the ocean based on satellite altimetry data
• Forskar Fredag 2022, Dalenium Science Centre (Sweden)
  Lea Poropat, outreach presentation: The sea and the climate – What drives the sea level variability along the Swedish coast?
• 23-27 May 2022, EGU22, Vienna (Austria)
  Lea Poropat, talk: Drivers of sea level variability in the Baltic/North/Nordic Seas using neural network
  David Ek, talk: Different drivers of Sea Level Variability at the North – Baltic Sea transition
• 16-18 May 2022, Swedish Climate Symposium, Norrköping (Sweden)
  Lea Poropat, talk: Drivers of sea level variability along the Swedish coast using neural networks
• 3 February 2022, GEO seminar, Göteborgs Universitet (Sweden)
  Lea Poropat, seminar: What drives regional sea level variability? Using neural networks o predict low-frequency sea level variations