Get Started
Research questionHow can neural ocean forecasts use cyclone tracks reliably?Rare event indicators may be mostly absent during training but become active at forecast time. A model can then treat their activation as a distribution shift and produce confident errors near the event.
Machine Learning
Neural and Evolutionary Computing
Latest papersRecent research connected to this question, newest first.Too Rare to Learn: Prescribed Cyclone Tracks Degrade a Bay of Bengal Ocean EmulatorThe study compares otherwise matched U-Nets trained on GLORYS12 reanalysis, with and without prescribed cyclone-track channels, using withheld Bay of Bengal cyclones across multiple seeds. Its evidence is limited to this regional dataset, architecture, event-conditioning design, and forecasting task.research paper · Sep 4, 2026
Related questions
How can multimodal models improve tropical-cyclone track and intensity forecasts across weather foundation models without retraining?How can online time-series forecasters adapt to evolving, unseen patterns without costly continual parameter updates?How can storm-surge emulators generate diverse spatial scenarios for hypothetical storms without costly physics simulations?How can AI weather models produce high-resolution global forecasts without relying solely on analysis data?
Home
Topics
Search
Library