Category: Publications2025

Exploration of short-term predictions and long-term projections of Barents Sea cod biomass using statistical methods on data from dynamical models

Koseki M, Sandø AB, Ottersen G, Årthun M, Stiansen JE 2025: Exploration of short-term predictions and long-term projections of Barents Sea cod biomass using statistical methods on data from dynamical models. PLoS One. https://doi.org/10.1371/journal.pone.0328762

Summary: This study aims to explore how well simple statistical modeling can generate short-term predictions and long-term projections of the total biomass of the Northeast Arctic stock of Atlantic cod (Gadus Morhua) inhabiting the Barents Sea. We examine the predictability of statistical models only based on hydrographic and lower trophic level biological variables from dynamical modeling. Simple and multiple linear regression models are developed based on gridded variables from the regional ocean model NEMO-NAA10km and the ecosystem model NORWECOM.E2E. This includes the essential environmental variables temperature, salinity, sea ice concentration, primary production and secondary production. The regression models are statistically evaluated to find variables that can capture variability in Barents Sea cod biomass. Finally, future total cod stock biomass is projected by applying the best found regression models to the range of downscaled IPCC climate scenarios from the coupled Intercomparison Project Phase 6 (CMIP6 Shared Socioeconomic Pathways; SSP1–2.6, SSP2–4.5, SSP5–8.5). Our prediction models are based on variables that affect cod both directly and indirectly. We find that several regression models have high prediction skill and capture the variations in total stock biomass of the Northeast Arctic cod well. Our results suggest that increased ocean temperature and abundant zooplankton may lead to a large cod stock. However, even if total stock biomass has a positive trend with an increase in copepods in the highest warming scenario SSP5–8.5, we found that it has a negative trend in the low emission scenario SSP1–2.6 when the regional ocean and ecosystem models show weak cooling and reduced zooplankton. We show that variability in essential environmental variables can provide a remarkably good first approximation to cod dynamics. However, to resolve the full picture other factors like fishing and natural mortality also need to be addressed explicitly.

Link to publication. You are most welcome to contact us or the corresponding author(s) directly, if you have questions.

Predicting Atlantic and Benguela Niño events with deep learning

Bachèlery ML, Brajard J, Patacchiola M, Illig S, Keenlyside N 2025: Predicting Atlantic and Benguela Niño events with deep learning. Sci. Adv.. https://doi.org/10.1126/sciadv.ads5185

Summary: Atlantic and Benguela Niño events substantially affect the tropical Atlantic region, with far-reaching consequences on local marine ecosystems, African climates, and El Niño Southern Oscillation. While accurate forecasts of these events are invaluable, state-of-the-art dynamic forecasting systems have shown limited predictive capabilities. Thus, the extent to which the tropical Atlantic variability is predictable remains an open question. This study explores the potential of deep learning in this context. Using a simple convolutional neural network architecture, we show that Atlantic/Benguela Niños can be predicted up to 3 to 4 months ahead. Our model excels in forecasting peak-season events with remarkable accuracy extending lead time to 5 months. Detailed analysis reveals our model’s ability to exploit known physical precursors, such as long-wave ocean dynamics, for accurate predictions of these events. This study challenges the perception that the tropical Atlantic is unpredictable and highlights deep learning’s potential to advance our understanding and forecasting of critical climate events.

Link to publication. You are most welcome to contact us or the corresponding author(s) directly, if you have questions.