TC-Next: Zero-Shot Multimodal Cyclone Forecasting

TC-Next: Zero-Shot Multimodal Cyclone Forecasting

TC-Next is a multimodal deep learning model that forecasts tropical cyclone track and intensity for 6‑24 hour lead times by combining atmospheric forecast fields from a foundation model with GridSat infrared satellite imagery. Trained solely on GraphCast forecasts for the Western Pacific, it reduces track error by 15‑44% and intensity error by a factor of 3‑6 compared to the rule‑based tracker TempestExtremes, and maintains superior performance when applied zero‑shot to other forecast systems such as Pangu‑Weather, IFS HRES, and WeatherNext Cyclones. Ablation studies confirm that incorporating the additional satellite modality consistently improves tracking accuracy at all lead times and enhances intensity predictions, especially at longer horizons.

AIFS-TC: A simple correction competitive with the operational frontier for tropical cyclone intensity forecasting

arXiv:2608. 09959v1 Announce Type: cross Abstract: AI weather models are in the process of revolutionising weather forecasting.

Tianmu-TC: Physics-constraints Generative Artificial Intelligence for Global Tropical Cyclone Forecasting

Tianmu-TC is a physics‑constraints generative AI framework designed for global tropical cyclone forecasting. Trained on Western North Pacific data, it produces controllable outputs with reduced uncertainty, outperforming both deterministic and ensemble meteorological AI models as well as the ECMWF NWP system across global ocean basins. The model also demonstrates strong performance in challenging scenarios such as data sparsity, anomaly tracks, rapid intensification, and weakening, while maintaining significantly lower computational cost.

Deep Learning Imputation of Missing Radius of Maximum Winds (Rmax) Values in Tropical Cyclone Best-Track Data

arXiv:2608. 09683v1 Announce Type: new Abstract: Probabilistic coastal hazard assessments require accurate characterization of tropical cyclone (TC) parameters, yet datasets often contain missing records for the radius of maximum winds (Rmax), a key variable in Joint Probability Method analyses.

Resolving sources of uncertainty in AI weather forecasting

The paper introduces Pangu‑Bayes, a probabilistic forecasting hierarchy that separates atmospheric‑state uncertainty from learned‑model uncertainty as distinct stochastic variables, allowing cross‑flow perturbations of the evolving state with Bayesian parameter samples. In tests on 90 held‑out 2023 tropical cyclones, Pangu‑Bayes reduces track, pressure, and wind errors by 54.2%, 17.2%, and 24.9% respectively, and improves rapid‑intensification detection. The study finds that atmospheric‑state variability more consistently improves track prediction, while learned‑model variability more often enhances intensity prediction, demonstrating how model‑defined uncertainty resolution can be linked to target‑dependent value and dynamical interpretation.

VeinCast: Physics-Guided Dynamic Field Graphs with Graph-Conditioned Fusion for Global Medium-Range Weather Forecasting

arXiv:2608. 09286v1 Announce Type: cross Abstract: Global medium-range weather forecasting requires modeling structured yet state-dependent interactions among heterogeneous atmospheric fields.

WeatherNext 3: Increasing resolution and performance of global weather models with raw observations

WeatherNext 3 is a new AI‑driven global weather model that improves both spatial and temporal resolution by generating hourly forecasts at 0.1° resolution, matching the best physics‑based models. It incorporates low‑latency geostationary satellite data and learns to predict satellite‑derived precipitation, tropical cyclones, and station observations, enabling 2 m temperature and dewpoint predictions anywhere and anytime. By directly using raw observations instead of relying solely on analysis data, WeatherNext 3 sets a new state‑of‑the‑art for probabilistic medium‑range forecasting skill.

Breaking the Statistical Similarity Trap in Extreme Convection Detection

arXiv:2509. 09195v2 Announce Type: replace Abstract: Current evaluation metrics for deep learning weather models create a "Statistical Similarity Trap", rewarding blurry predictions while missing rare, high-impact events.

Too Rare to Learn: Prescribed Cyclone Tracks Degrade a Bay of Bengal Ocean Emulator

Neural ocean emulators that use prescribed cyclone tracks as input were tested in the Bay of Bengal. Two identical U‑Net models were compared: one with four cyclone‑track channels and one without. The storm‑conditioned model performed worse than persistence on every run, while the ocean‑only model outperformed persistence, indicating that the rare cyclone input caused the model to learn a misleading response.

Stable Attention Response for Reliable Precipitation Nowcasting

arXiv:2605. 13181v2 Announce Type: replace-cross Abstract: Precipitation nowcasting remains challenging due to the highly localized, rapidly evolving, and heterogeneous nature of atmospheric dynamics.

Pointwise is Pointless? A Multimodal Ablation Study for Precipitation Nowcasting with Graph Neural Networks

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.

WeatherNext: AI model achieves breakthrough in forecasting cyclones