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:20250822T115806Z
LOCATION:Room 5.2D02
DTSTART;TZID=Europe/Stockholm:20250617T130000
DTEND;TZID=Europe/Stockholm:20250617T133000
UID:submissions.pasc-conference.org_PASC25_sess134_msa139@linklings.com
SUMMARY:RAIN: Reinforcement Algorithms for Improving Numerical Weather and
  Climate Models
DESCRIPTION:Pritthijit Nath, Henry Moss, and Emily Shuckburgh (University 
 of Cambridge) and Mark Webb (Met Office)\n\nThis study explores integratin
 g reinforcement learning (RL) with idealised climate models to address key
  parameterisation challenges in climate science. Current climate models re
 ly on complex mathematical parameterisations to represent sub-grid scale p
 rocesses, which can introduce substantial uncertainties. RL offers capabil
 ities to enhance these parameterisation schemes, including direct interact
 ion, handling sparse or delayed feedback, continuous online learning, and 
 long-term optimisation. We evaluate the performance of eight RL algorithms
  on two idealised environments: one for temperature bias correction, anoth
 er for radiative-convective equilibrium (RCE) imitating real-world computa
 tional constraints. Results show different RL approaches excel in differen
 t climate scenarios with exploration algorithms performing better in bias 
 correction, while exploitation algorithms proving more effective for RCE. 
 These findings support the potential of RL-based parameterisation schemes 
 to be integrated into global climate models, improving accuracy and effici
 ency in capturing complex climate dynamics. Overall, this work represents 
 an important first step towards leveraging RL to enhance climate model acc
 uracy, critical for improving climate understanding and predictions. Code 
 accessible at https://github.com/p3jitnath/climate-rl.\n\nDomain: Chemistr
 y and Materials, Climate, Weather, and Earth Sciences, Computational Metho
 ds and Applied Mathematics\n\nSession Chairs: Riccardo Balin (Argonne Nati
 onal Laboratory) and Alessandro Rigazzi (HPE)\n\n
END:VEVENT
END:VCALENDAR