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:20241120T082408Z
LOCATION:HG E 3
DTSTART;TZID=Europe/Stockholm:20240603T120000
DTEND;TZID=Europe/Stockholm:20240603T123000
UID:submissions.pasc-conference.org_PASC24_sess163_msa150@linklings.com
SUMMARY:Coupling Machine Learning to Numerical (Climate) Models: Tools, Ch
 allenges, and Lessons Learned
DESCRIPTION:Minisymposium\n\nJack Atkinson (University of Cambridge)\n\nTh
 e rise of machine learning (ML) has seen many scientists seeking to incorp
 orate these techniques into numerical models. Doing so presents a number o
 f challenges, however. The Institute of Computing for Climate Science (ICC
 S) has explored this problem in the context of coupling ML components/para
 meterisations into climate models. In this talk we will explore a number o
 f challenges in this area, how ICCS has tackled them, and what has been le
 arnt in the process. We will present FTorch, a library developed by ICCS t
 o bridge the gap between Fortran (in which many large physics models are w
 ritten) and PyTorch (in which much ML is performed) and lower the technica
 l barrier to scientists seeking to leverage ML in their work. Case studies
  have been performed using both CPU and GPU architectures from laptops up 
 to HPC systems. We will reflect on the design challenges of coupling ML pa
 rameterisations to large numerical models and outline a framework guidelin
 es following software design principles to aid in this process. Finally we
  will discuss ongoing work using the Community Earth System Model (CESM) t
 o re-deploy a neural net trained using a high-resolution model with a diff
 erent grid and variables to a new setting.\n\nDomain: Chemistry and Materi
 als, Climate, Weather, and Earth Sciences, Engineering, Physics, Computati
 onal Methods and Applied Mathematics\n\nSession Chair: Jack Atkinson (Univ
 ersity of Cambridge)
END:VEVENT
END:VCALENDAR