BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260724T151409Z
LOCATION:Bldg. 6 - 103
DTSTART;TZID=Europe/Stockholm:20260701T140000
DTEND;TZID=Europe/Stockholm:20260701T160000
UID:submissions.pasc-conference.org_PASC26_sess162@linklings.com
SUMMARY:MS5E - Foundation Model for Weather and Climate
DESCRIPTION:Access the recording\n\nOrganizer(s): Xavier Lapillonne (Meteo
 Swiss), Ilaria Luise (ECMWF), and Sebastian Schemm (University of Cambridg
 e)\n\nIn recent years weather forecasting and to a lower extent climate mo
 deling have been undergoing a revolution driven by the emergence of machin
 e learning-based models. After having been successfully developed and used
  in the field of Large Language Model, the foundation model concept is bei
 ng applied to machine learning-based weather-forecasting. The goal is to c
 apture rich, multi-scale representations of the Earth system across space 
 and time by training on diverse datasets. These models can then be applied
  to a wide range of tasks, much like traditional equation-based Earth syst
 em models. The first large foundation models are now emerging, and their p
 otential applications are actively being explored. The talks in this minis
 ymposium will define the concept of foundation models for weather and clim
 ate, discuss both their development and applications, and provide a compre
 hensive overview of the current state of the field.\n\nWeatherGenerator: A
  Foundation Model for Weather and Climate\n\nThe WeatherGenerator project 
 aims to develop a foundation model for the European community that combine
 s a wide variety of datasets to improve a broad range of applications acro
 ss spatial and temporal scales. The WeatherGenerator model is a highly mul
 timodal transformer capable of ingesting and pred...\n\n\nJulian Kuehnert 
 (ECMWF)\n---------------------\nEarth System Foundation Model - Heterogene
 ous Data Integration and Forecasting\n\nThe talk introduces the Earth Syst
 em Foundation Model (ESFM), a fully open foundation model tailored for wea
 ther and climate modeling tasks. Built upon the Swin Transformer backbone 
 of the pioneering Aurora model, ESFM introduces several key extensions tha
 t allow it to process diverse and heterogene...\n\n\nFirat Ozdemir (Swiss 
 Data Science Center); Yun Cheng and Salman Mohebi (Swiss data science cent
 er); Fanny Lehmann, Simon Adamov, Langwen Huang, Leonardo Trentini, Oliver
  Fuhrer, Torsten Hoelfer, and Siddhartha Mishra (ETH Zurich); Sebastian Sc
 hemm (University of Cambridge); Benedikt Soja (ETH Zurich); and Mathieu Sa
 lzmann (Swiss data science center)\n---------------------\nClimate in a Bo
 ttle: Towards a Foundation Model for Climate\n\nAI emulators offer a path 
 to compressing, boosting limited ensembles, and improving the latency of i
 nteracting with petabyte-scale climate prediction data. However, prevailin
 g auto-regressive paradigms offer limited flexibility, and are challenging
  to train on climate time horizons due to drifts, in...\n\n\nKarthik Kashi
 nath (NVIDIA Inc.)\n---------------------\nPanel Discussion Foundation Mod
 el for Weather and Climate\n\nThe panel discussion will explore the challe
 nges and opportunities of using foundation models for weather and climate.
 \n\n\nXavier Lapillonne (MeteoSwiss)\n\nDomain: Climate, Weather, and Eart
 h Sciences\n\nSession Chair: Xavier Lapillonne (MeteoSwiss)
END:VEVENT
END:VCALENDAR
