Newest papers

GraphCast-Mamba: Learning Non-Markovian Corrections for Weather Forecasting

Ilya Vilkoviskiy, Liang-Hong Mo, Suozhi Huang, Dmitry A. Abanin · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.28512 · cited by 0 works (OpenAlex) · open-access PDF

Russian

Нерешенные графики атмосферных процессов могут ввести историческую зависимость за пределами коротких входных окон, используемых многими моделями прогнозирования погоды. Мы вводим GraphCast--Mamba, который увеличивает замороженный, предварительно подготовленный $1\circ$ GraphCast спинной мозг с обучаемым Mamba коррекционным отделом, объединяющим графическое сообщение и слои Mamba. Отдел сохраняет временное состояние памяти на протяжении последовательных шестичасовых шагов прогнозирования, изучая корректировки, информированные временным контекстом за пределами входа GraphCast с двумя рамками. Обучение ограничено коррекционным отделом, используя данные ERA5 с 2015 по 2021 год. После 2500 обновлений оптимизатора, включая обогре

Machine translation (Meta M2M100); the English original follows.

Abstract

Unresolved atmospheric processes can introduce history dependence beyond the short input windows used by many weather forecasting models. We introduce GraphCast--Mamba, which augments a frozen, pretrained $1^\circ$ GraphCast backbone with a trainable Mamba correction branch combining graph message passing and Mamba layers. The branch retains a temporal memory state across successive six-hour forecast steps, learning corrections informed by temporal context beyond GraphCast's two-frame input. Training is confined to the correction branch, using ERA5 data from 2015--2021. After 2,500 optimizer updates, including warm-up, the selected model achieves substantial improvements relative to the uncorrected GraphCast backbone. It reduces the GraphCast loss (normalized weighted mean-squared error; MSE) by 14.42% when aggregated over the full ten-day forecast and by 21.66% at day 10 (240 h after forecast initialization). At this final forecast lead time, root-mean-square error (RMSE) decreases by 7.09--14.60% across all 11 evaluated variable groups, spanning temperature, wind, pressure, geopotential, specific humidity, and precipitation. These results demonstrate broad improvements in ten-day weather forecasts through a short correction-training schedule while retaining the pretrained backbone.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Context (archaeology), Computer science, Schedule, Meteorology, Weather forecasting, Forecast skill

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange