GraphDOP: Towards skilful data-driven medium-range weather forecasts learnt and initialised directly from observations

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Alexe, Mihai, Boucher, Eulalie, Lean, Peter, Pinnington, Ewan, Laloyaux, Patrick, McNally, Anthony, Lang, Simon, Chantry, Matthew, Burrows, Chris, Chrust, Marcin, Pinault, Florian, Villeneuve, Ethel, Bormann, Niels, Healy, Sean
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866929642461462528
author Alexe, Mihai
Boucher, Eulalie
Lean, Peter
Pinnington, Ewan
Laloyaux, Patrick
McNally, Anthony
Lang, Simon
Chantry, Matthew
Burrows, Chris
Chrust, Marcin
Pinault, Florian
Villeneuve, Ethel
Bormann, Niels
Healy, Sean
author_facet Alexe, Mihai
Boucher, Eulalie
Lean, Peter
Pinnington, Ewan
Laloyaux, Patrick
McNally, Anthony
Lang, Simon
Chantry, Matthew
Burrows, Chris
Chrust, Marcin
Pinault, Florian
Villeneuve, Ethel
Bormann, Niels
Healy, Sean
contents We introduce GraphDOP, a new data-driven, end-to-end forecast system developed at the European Centre for Medium-Range Weather Forecasts (ECMWF) that is trained and initialised exclusively from Earth System observations, with no physics-based (re)analysis inputs or feedbacks. GraphDOP learns the correlations between observed quantities - such as brightness temperatures from polar orbiters and geostationary satellites - and geophysical quantities of interest (that are measured by conventional observations), to form a coherent latent representation of Earth System state dynamics and physical processes, and is capable of producing skilful predictions of relevant weather parameters up to five days into the future.
format Preprint
id arxiv_https___arxiv_org_abs_2412_15687
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle GraphDOP: Towards skilful data-driven medium-range weather forecasts learnt and initialised directly from observations
Alexe, Mihai
Boucher, Eulalie
Lean, Peter
Pinnington, Ewan
Laloyaux, Patrick
McNally, Anthony
Lang, Simon
Chantry, Matthew
Burrows, Chris
Chrust, Marcin
Pinault, Florian
Villeneuve, Ethel
Bormann, Niels
Healy, Sean
Atmospheric and Oceanic Physics
Machine Learning
We introduce GraphDOP, a new data-driven, end-to-end forecast system developed at the European Centre for Medium-Range Weather Forecasts (ECMWF) that is trained and initialised exclusively from Earth System observations, with no physics-based (re)analysis inputs or feedbacks. GraphDOP learns the correlations between observed quantities - such as brightness temperatures from polar orbiters and geostationary satellites - and geophysical quantities of interest (that are measured by conventional observations), to form a coherent latent representation of Earth System state dynamics and physical processes, and is capable of producing skilful predictions of relevant weather parameters up to five days into the future.
title GraphDOP: Towards skilful data-driven medium-range weather forecasts learnt and initialised directly from observations
topic Atmospheric and Oceanic Physics
Machine Learning
url https://arxiv.org/abs/2412.15687