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:20230831T095755Z
LOCATION:Sertig
DTSTART;TZID=Europe/Stockholm:20230628T140000
DTEND;TZID=Europe/Stockholm:20230628T160000
UID:submissions.pasc-conference.org_PASC23_sess109@linklings.com
SUMMARY:MS6C - Implementation and Validation of Hybrid Earth System and Ma
 chine Learning Models
DESCRIPTION:Minisymposium\n\nThe transformative power of ML unleashed in E
 arth System modelling is taking shape. Recent advances in building hybrid 
 models combining mechanistic Earth system models grounded in physical unde
 rstanding and machine learning models trained from huge amounts of data sh
 ow promising results. Full replacement models such as FourCastNet have als
 o made continuous progress and shown impressive performance. Hybrid models
  that include, for instance, ML based parameterizations, can provide a sub
 stantial speed-up or qualitative improvement if trained on high-resolution
  data compared to parameterizations based on few data points. However, the
  ongoing implementation of such models poses technical challenges, while t
 he acceptance of hybrid approaches and full replacement models critically 
 depends on the means scientists have for validating them. This minisymposi
 um will discuss the technical challenges in building hybrid models, such a
 s the coupling of Earth system and machine learning model components via e
 mbedded Python or MPI, and report on accomplished milestones and benchmark
  results. For further adoption of ML methods, it is also crucial to establ
 ish the credibility of ML-enhanced Earth system models through statistical
  reproducibility. The symposium will discuss arising questions around the 
 validation of hybrid approaches, including metrics for training and evalua
 tion that emphasize spatial features.\n\nGetting the Best of Both Worlds: 
 Bridging Fortran and Pytorch for the CPU and the GPU HPC Simulations\n\nOv
 er the past years, machine learning (ML) has been attracting rapidly incre
 asing interest in the computational science. Many developers are adding ML
  models to their traditional simulation pipelines, and yet more are willin
 g to follow. However, integration of such models into an existing applicat
 io...\n\n\nDmitry Alexeev and Markus Hrywniak (NVIDIA Inc.)\n-------------
 --------\nCapabilities, Limitations and Validation of Statistical Earth Sy
 stem Models\n\nThe last year has seen tremendous progress on deep learning
  based forecasting models trained on reanalysis data. An extension of thes
 e to Earth system models (ESMs) and to training on observational data for 
 obtaining predictions that are not bound by the reanalysis process (and he
 nce classical model...\n\n\nChristian Lessig (Otto-von-Guericke-Universita
 t Magdeburg)\n---------------------\nData-Driven Cloud Cover Parameterizat
 ions for the ICON Model\n\nA promising approach to improving cloud paramet
 erizations in climate models, and thus climate projections, is to train ma
 chine learning (ML) algorithms on coarse-grained output of global storm-re
 solving model simulations. The ICOsahedral Non-hydrostatic (ICON) modeling
  framework enables simulations ...\n\n\nArthur Grundner (DLR, Columbia Uni
 versity); Tom Beucler (University of Lausanne); Pierre Gentine (Columbia U
 niversity); Marco Giorgetta (Max Planck Institute for Meteorology); Fernan
 do Iglesias-Suarez and Rémi Kazeroni (DLR); and Veronika Eyring (DLR, Univ
 ersity of Bremen)\n---------------------\nBuilding Data-Driven Surrogate M
 odels at Appropriate Granularity Using High-Resolution Climate Simulation 
 and Coupling Library\n\nWhen building a climate model emulator using machi
 ne learning, we are concerned about the generalization performance of the 
 surrogate model and the extrapolation of the simulation results. It is ver
 y problematic to evaluate an emulator that implicitly includes too many ph
 ysical processes and has bee...\n\n\nHisashi Yashiro (National Institute f
 or Environmental Studies); Takashi Arakawa (CliMTech Inc., The University 
 of Tokyo); and Shinji Sumimoto and Kengo Nakajima (The University of Tokyo
 )\n\nDomain: Climate, Weather and Earth Sciences\n\nSession Chair: Tobias 
 Weigel (German Climate Computing Centre, DKRZ)
END:VEVENT
END:VCALENDAR