Observed SIC for certain regions can vary substantially from one year to another. (a) SIC observed for September 2007, when the extent of the sea ice was lower than usual. (b) SIC observed for September 2013, when the extent of the sea ice was greater than usual. (c) Area where the observed SIC differed the most between September 2007 and September 2013 is highlighted (white), used as the ROI for gradient attribution.
Image:
Harald L. Joakimsen (UiT)

Observed SIC for certain regions can vary substantially from one year to another. (a) SIC observed for September 2007, when the extent of the sea ice was lower than usual. (b) SIC observed for September 2013, when the extent of the sea ice was greater than usual. (c) Area where the observed SIC differed the most between September 2007 and September 2013 is highlighted (white), used as the ROI for gradient attribution.

Collaborative paper with The Alan Turing Institute and the British Antarctic Survey accepted in IEEE Geoscience and Remote Sensing Letters

We are thrilled to announce that the paper "Interrogating Sea Ice Predictability With Gradients", a collaborative paper between Visual Intelligence, The Alan Turing Institute and the British Antarctic Survey, has been accepted in IEEE Geoscience and Remote Sensing Letters.

Collaborative paper with The Alan Turing Institute and the British Antarctic Survey accepted in IEEE Geoscience and Remote Sensing Letters

We are thrilled to announce that the paper "Interrogating Sea Ice Predictability With Gradients", a collaborative paper between Visual Intelligence, The Alan Turing Institute (AT) and the British Antarctic Survey (BAS), was accepted in the journal IEEE Geoscience and Remote Sensing Letters on February 14th 2024.

The paper focuses on interrogating the effect of the IceNet's, a state-of-the-art ice prediction model, input feature with a gradient-based analysis, which takes advantage of the developments within the deep learning literature to open the so-called "black box".

The authors' analysis focuses on the unusually large sea ice extent event in September 2013 and indicates that IceNet places a strong emphasis on previous observations of SIC, linear trends, and seasonal components when making predictions. They further identify which input features are most influential for the prediction and also which spatial location these measurements are particularly influential.

The authors of this particular work from the Visual Intelligence side are post doc. Luigi Luppino (UiT), PhD students Harald Lykke Joakimsen and Iver Martinsen (UiT), and Robert Jenssen (UiT). From the UK side, the authors are Scott Hosking (BAS/AT) and Andrew McDonald (BAS).

You may read the paper via IEEE Explore.

Abstract

Predicting sea ice concentration (SIC) is an important task in climate analysis. The recently proposed deep learning system IceNet is the state-of-the-art sea ice prediction model. IceNet takes high-dimensional climate simulations and observational data as input features and forecasts SIC for the next 6 months over a spatial grid over the northern hemisphere. The model has proven to be particularly good at predicting extreme sea ice events compared with previous dynamical models, but lacks interpretability. In the original IceNet paper, a permute-and-predict approach was taken for assessing feature importance. However, this approach is not capable of revealing whether a feature contributes positively or negatively to the final prediction, nor can it reveal the importance of features over the spatial grid of predictions. In this letter, we take steps to instead interrogate the effect of the IceNet input feature with a gradient-based analysis, taking advantage of developments within the deep learning literature to open the so-called black box. Our analysis focuses on the unusually large sea ice extent event in September 2013 and indicates that IceNet places a strong emphasis on previous observations of SIC, linear trends, and seasonal components when making predictions. In our analysis, we identify which input features are most influential for the prediction and also which spatial location these measurements are particularly influential.

Latest news

Anders Waldeland receives the Digital Trailblazer Award 2025

December 4, 2025

Congratulations to Senior Research Scientist Anders Waldeland, who was awarded the Digital Trailblazer Award 2025 at the Dig X Subsurface conference in Oslo, Norway.

sciencenorway.no: AI can help detect heart diseases more quickly

December 3, 2025

Researchers have developed an artificial intelligence that can automatically measure the heart's structure – both quickly and accurately (Popular science article on sciencenorway.no)

State Secretary Marianne Wilhelmsen visits SFI Visual Intelligence and UiT

November 26, 2025

State Secretary Marianne Wilhelmsen visited UiT The Arctic University of Norway to learn more about SFI Visual Intelligence and UiT's AI initiatives in education and research.

TV2.no: Sier Elon Musk er smartere enn Leonardo da Vinci

November 25, 2025

KI-chatboten Grok har fortalt brukere at verdens rikeste mann er både smartere og sprekere enn noen andre i verden – inkludert basketballstjernen LeBron James og Leonardo da Vinci (Norwegian news article on tv2.no)

Successful science communication workshop at Skibotn

November 21, 2025

The Visual Intelligence Graduate School gathered our early career researchers for a 3-Day Science Communication workshop at Skibotn field station outside of Tromsø, Norway.

uit.no: UiT og Aker Nscale sammen om storsatsing på kunstig intelligens

November 19, 2025

Onsdag inngikk Aker Nscale og UiT Norges arktiske universitet en ti-årig samarbeidsavtale for å utvikle og styrke kompetansemiljøene for kunstig intelligens i Narvik og Nord-Norge. Aker Nscale garanterer for 100 millioner kroner i avtaleperioden (news story on uit.no)

Two fruitful days at The Alan Turing Institute's headquarters

November 17, 2025

Centre Director Robert Jenssen and PhD Candidate Lars Uebbing had two fruitful days together with researchers at The Alan Turing Institute's headquarters in London

Anders Waldeland nominated for the Digital Trailblazer 2025 Award

November 12, 2025

Senior Research Scientist Anders Waldeland is nominated for the Digital Trailblazer 2025 Award. The winner is announced at the Dig X Subsurface conference in Oslo, Norway in December.

AI can help detect heart diseases more quickly

November 7, 2025

Visual Intelligence researchers have developed an AI to automatically measure the heart's structure – both quickly and accurately. They believe it can help doctors detect and treat cardiovascular diseases faster.

How can PET and AI help detect prostate cancer earlier?

November 5, 2025

Samuel Kuttner and Elin Kile presented research on PET and artificial intelligence at evening seminar on early detection of prostate cancer organized by the Norwegian Prostate Cancer Assocation.

Visual Intelligence represented at Svarte Natta 2025

October 29, 2025

Centre Director Robert Jenssen represented Visual Intelligence at Svarte Natta 2025 – North Norway's journalist and media conference organized by the Norwegian Union of Journalists.

My Research Stay at Visual Intelligence: Aitor Sánchez

October 5, 2025

Aitor Sánchez is a PhD candidate at the Intelligent Systems Group of the University of the Basque Country in Spain. He visited Visual Intelligence in Tromsø from March to June 2025.