BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Short-term\, high-resolution sea ice forecasting with diffusion mo
 del ensembles - Andrew McDonald\, University of Cambridge and British Ant
 arctic Survey
DTSTART:20250212T130000Z
DTEND:20250212T140000Z
UID:TALK227839@talks.cam.ac.uk
CONTACT:Dr Birgit Rogalla
DESCRIPTION:Sea ice plays a key role in Earth’s climate system and exhib
 its significant seasonal variability as it advances and retreats across th
 e Arctic and Antarctic every year. The production of sea ice forecasts pro
 vides great scientific and practical value to stakeholders across the pola
 r regions\, informing shipping\, conservation\, logistics\, and the daily 
 lives of inhabitants of local communities. Machine learning offers a promi
 sing means by which to develop such forecasts\, capturing the nonlinear dy
 namics and subtle spatiotemporal patterns at play as effectively—if not 
 more effectively—than conventional physics-based models. In particular\,
  the ability of deep generative models to produce probabilistic forecasts 
 which acknowledge the inherent stochasticity of sea ice processes and repr
 esent uncertainty by design make them a sensible choice for the task of se
 a ice forecasting. Diffusion models\, a class of deep generative models\, 
 present a strong option given their state-of-the-art performance on comput
 er vision tasks and their strong track record when adapted to spatiotempor
 al modelling tasks in weather and climate domains. In this talk\, I will p
 resent preliminary results from a IceNet-like [1] diffusion model trained 
 to autoregressively forecast daily\, 6.25 km resolution sea ice concentrat
 ion in the Bellingshausen Sea along the Antarctic Peninsula. I will also t
 ouch on the downstream applications for these forecasts\, from conservatio
 n to marine route planning\, which are under development at the British An
 tarctic Survey (BAS). I welcome ideas and suggestions for improvement and 
 look forward to discussing opportunities for collaboration within and beyo
 nd BAS.\n \n\n[1] Andersson\, Tom R.\, et al. "Seasonal Arctic sea ice for
 ecasting with probabilistic deep learning." Nature communications 12.1 (20
 21): 5124. https://www.nature.com/articles/s41467-021-25257-4
LOCATION:BAS Seminar Room 2\; https://ukri.zoom.us/j/96472472041
END:VEVENT
END:VCALENDAR