The paper introduces IonoDGNN, a dynamic graph neural network that models the ionosphere over ionospheric pierce points, conditioning forecasts on future satellite trajectories (ephemeris conditioning). It predicts ionospheric irregularities, measured by the Rate of TEC Index, up to two hours ahead using GNSS data from Singapore, achieving a Brier Skill Score of 0.55 and a PR-AUC of 0.77—significant improvements over persistence and interpolation baselines. The model remains robust under simulated coverage dropouts, leveraging spatial message passing to maintain predictive skill on affected nodes.
By Mert Can Turkmen, Eng Leong Tan, Yee Hui Lee
arXiv:2606. 29644v1 Announce Type: new Abstract: Earth system infrastructures relying on satellite-based technologies, such as Global Positioning System (GPS) communications, are affected by ionospheric Total Electron Content (TEC) gradients.
By Stephen Tete, Carl Shneider, Maxime Cordy, Claudio Cesaroni, Andreas Hein, Vasily Petrov
arXiv:2608.12685v2 Announce Type: replace-cross
Abstract: An ``Attention Residual U-Net'' method is described for probabilistic quantitative precipitation forecasting (PQPF) that predicts the hourly...
By Thomas M Hamill
arXiv:2606. 18436v1 Announce Type: cross Abstract: Sparse point observations are increasingly available for precipitation nowcasting, but it is unclear how much they improve dense radar-field forecasts.
By Oph\'elia Miralles, M\'at\'e Mile, Christoffer Artturi, Thomas Nipen, Ivar Seierstad
arXiv:2607. 05100v1 Announce Type: cross Abstract: Data-driven models now rival numerical weather prediction in the medium range, but extending them to sub-seasonal lead times raises challenges absent at shorter horizons.
By Jakob Schloer, Steffen Tietsche, Christopher D. Roberts, Lorenzo Zampieri, Simon Lang, Gert Mertes, Gareth Jones, Matthew Chantry, Frederic Vitart
Data-driven models now rival numerical weather prediction in the medium range, but extending them to sub-seasonal lead times raises challenges absent at shorter horizons. Errors accumulate over long autoregressive rollouts, systematic biases grow with lead time, and several years of data must be held out for independent verification, even though machine-learning models otherwise benefit from longer training records.