arXiv Machine Learning By Kate Duffy, Elise Penn, Tristan Ballard, August Posch, Thomas J. Vandal

WeatherDreamer: Learning a World Model from Multimodal Earth Observations

Read the original on arXiv Machine Learning →

WeatherDreamer is a new end‑to‑end world model that learns the recursive filtering cycle of data assimilation directly from 18 global observing systems, without relying on traditional reanalysis. It infers a latent atmospheric state, propagates it hourly with learned dynamics, and corrects it each hour using new observations, enabling it to predict measurements for any observing system. In held‑out tests, its ensembles outperform NOAA’s GEFS for temperature, dew point, and wind out to 12 h, and its skill for precipitation, cloud‑top height, and solar irradiance matches or exceeds operational products.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

arXiv Machine Learning
Sep 4

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.

By Stephan Rasp, Boris Babenko, Dominic Masters, Andrew El-Kadi, Samier Merchant, Guy Shalev, Ilan Price, Fred Zyda, Remi Lam, Sasha Shysheya, Matthew Willson, Stratis Markou, Shreya Agrawal, Suhani Vora, Mohammed Alewi Hassen, Sunny Mak, Tom R. Andersson, Megan Bela, Akib Uddin, Nofar Peled Levi, Ben Gaiarin, Ferran Alet, Aaron Bell, Peter Battaglia, Alvaro Sanchez-Gonzalez
arXiv Machine Learning
Sep 4

From Nowcasting to Forecasting: Adapting a Reanalysis-Trained

The paper introduces CloudCast v2, a machine‑learning model that forecasts 12‑hour cloud‑cover from satellite‑derived initial conditions. Trained first on the Copernicus European Regional Reanalysis to learn cloud‑evolution dynamics, it is then adapted to real satellite data using conditional flow matching, a generative technique that conditions noise on observed cloud fields and NWP inputs. CloudCast v2 achieves a 10 % reduction in mean absolute error compared to its predecessor and surpasses it in spatial skill after 3–6 hours, extending useful forecasting beyond the typical 1–3‑hour nowcasting window while preserving satellite‑level spatial detail.

By Mikko Partio, Leila Hieta, Ossi Laine
arXiv Machine Learning
Jun 26

Huracan: A skillful end-to-end data-driven system for ensemble data assimilation and weather prediction

arXiv:2508. 18486v2 Announce Type: replace-cross Abstract: Over the past few years, machine learning-based data-driven weather prediction has been transforming operational weather forecasting by providing more accurate forecasts while using a mere fraction of computing power compared to traditional numerical weather prediction (NWP).

By Zekun Ni, Jonathan Weyn, Hang Zhang, Yanfei Xiang, Jiang Bian, Weixin Jin, Kit Thambiratnam, Qi Zhang, Haiyu Dong, Hongyu Sun
Hugging Face Trending Papers
Jul 6

AIFS-SUBS: Extending Data-Driven Forecasting to Sub-Seasonal Timescales

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.

Hugging Face Trending Papers
Sep 3

From Nowcasting to Forecasting: Adapting a Reanalysis-Trained

The paper introduces CloudCast v2, a machine‑learning model that forecasts 12‑hour cloud cover from observation‑based initial conditions. It is first trained on the Copernicus European Regional Reanalysis to capture cloud‑evolution dynamics, then adapted to satellite‑derived cloud fields using conditional flow matching. Compared to its predecessor, CloudCast v2 reduces mean absolute error by 10% over 1–12 h and surpasses it in spatial skill after 3–6 h, demonstrating that observation‑initialized forecasts can extend beyond the typical 1–3‑hour nowcasting window while preserving spatial detail.