From a9f96ceda3370a52d632cf0614ec5b982ea5d153 Mon Sep 17 00:00:00 2001 From: Erik van Sebille Date: Thu, 3 Sep 2026 09:40:42 +0200 Subject: [PATCH] Adding Donner paper --- src/data/papers-citing-parcels.ts | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/src/data/papers-citing-parcels.ts b/src/data/papers-citing-parcels.ts index 8ea417d..db5819a 100644 --- a/src/data/papers-citing-parcels.ts +++ b/src/data/papers-citing-parcels.ts @@ -3168,4 +3168,14 @@ export const papersCitingParcels: Paper[] = [ abstract: 'Antarctic krill are central to the Southern Ocean ecosystem and support a commercial fishery. Late larval stages of Antarctic krill depend on sea ice for food and shelter during winter. As climate change alters sea ice habitats and polar ecosystems, understanding krill transport and retention is essential for ecosystem-based fishery management. These late-stage larvae have been observed to perform reverse diel vertical migration (DVM), moving from beneath the sea ice during the day to the upper water column at night. Variability in ocean currents and sea ice drift, together with behaviours such as DVM, can strongly influence krill transport and retention. To investigate the impact of this behaviour, we conducted Lagrangian particle tracking experiments using 32 years of daily sea ice and ocean velocities from a 0.1° global ocean-sea ice model (ACCESS-OM2-01). Virtual particles were released in the south-west Atlantic sector of the Southern Ocean and tracked over winter (April-October), a key period prior to spring recruitment. To assess sensitivity to DVM, we applied five scenarios in which particles spent different proportions of each day influenced by ocean currents and sea ice drift under sea ice. Transport driven solely by ocean velocities resulted in the highest proportion of particles on the Antarctic Peninsula shelf by winters end, with low interannual variability (55.5% ± 7% of released particles). Increasing exposure to sea ice drift reduced shelf retention and increased variability (33.6% ± 10.9%). Our results show that including sea ice interactions enhances overwinter connectivity from the Antarctic Peninsula to South Georgia and between management areas, highlighting sea ice-driven northward transport of young krill. As krill depend on declining sea ice habitats, connectivity across the Antarctic Peninsula and Scotia Sea may shift in the future. Understanding interannual transport variability between fishing regions improves stock assessment and supports sustainable management.', }, + { + title: + 'CoDiT: conditional diffusion models for multi-day ocean drifter trajectory prediction', + published_info: 'Machine Learning: Earth, 2, 025012', + authors: + 'Donner, C, S Goshtasbpour, E Dalsasso, M Romero, M Béchaz, M Rußwurm, R de Vries, D Tuia, M Volpi (2026)', + doi: 'https://doi.org/10.1088/3049-4753/ae9bc5', + abstract: + 'Accurate multi-day forecasting of floating-object trajectories on the ocean surface is critical for applications ranging from search-and-rescue to environmental tracking. This task remains however challenging due to the complex interplay of influencing factors such as ocean currents and winds. In this work, we frame trajectory prediction as a denoising task and present Conditional Diffusion models for Trajectories (CoDiT), which adapts the denoising diffusion framework, originally developed for image synthesis, to the problem of trajectory forecasting. CoDiT generates realistic trajectory forecasts, conditioned on heterogeneous context data: ocean currents and winds from reanalysis products, bathymetry, and the initial position. We train and evaluate CoDiT on two global, specialized datasets focusing on the open ocean and coastal regions, using GPS trajectories from the Global Drifter Program as ground truth. We compare CoDiT rigorously against various baselines, including a convolutional neural network that predicts velocity fields, and physical forecasts generated directly from the current and wind fields. Quantitative evaluations show that CoDiT achieves the lowest position error across both datasets and all forecast horizons, and the best probabilistic forecast quality among all methods, as measured by the energy score. Notably, in the coastal setting, CoDiT is the only method to surpass the naive persistence baseline in position error.', + }, ]