I det åpne hav og i dype fjorder skjuler det seg en helt spesiell type fisk.

Qidi holder prøveforelesning 22. mai kl. 10.15 over oppgitt emne: “Origins and evolution of weather forecasting before the age of operational numerical weather prediction”. I prøveforelesningskomitéen er: Harald Sodemann, leder Costijn Zwart Helene Asbjørnsen Veiledere: Thomas Spengler (hovedveileder) Clemens Spensberger (biveileder) Linus Magnusson (biveileder) Prøveforelesningen holdes på Foredragssal 200. Studenter er også velkomne!

Dear all, The next BCCR Seminar will be given by Will Chapman from the University of Colorado, Boulder. He will present his work on Towards coupled data-driven Earth system prediction. As next Monday is a public holiday, the seminar will be held next Wednesday in the usual BCCR seminar room (4th floor of the West wing) at 11:00. We hope to see you there! Best regards, Fiona and Johannes Abstract Recent advances in machine learning have enabled data driven models for weather and climate prediction that are approaching the skill of traditional numerical weather prediction systems. These approaches, ranging from hybrid physics and machine learning methods to full model emulation, offer dramatic reductions in computational cost and open the door to new experiments, large ensembles, and interactive workflows. In this talk, the speakers present recent work within NCAR’s CREDIT, Community Research Earth Digital Twin, framework, focusing on stable decadal scale autoregressive prediction. They introduce CAMulator, a data driven emulator of the NSF NCAR Community Atmosphere Model, CAM, and examine its architecture, training strategy, and long horizon behavior. Results show that the model reproduces key atmospheric dynamics while maintaining stability over extended forecasts. They then explore a pathway toward coupled data driven Earth system modeling by coupling CAMulator with a process based ocean model. This required building a robust interface between legacy Fortran infrastructure and modern Python based machine learning systems. They discuss both the technical challenges and the solutions that enabled this hybrid coupling, and present early results demonstrating the potential of such systems to accelerate Earth system prediction and experimentation. Speaker information Dr. William (Will) Chapman is an Assistant Professor in the Department of Atmospheric and Oceanic Sciences at the University of Colorado Boulder in Boulder, Colorado, USA. His research focuses on climate predictability, machine learning, and coupled Earth system dynamics, with an emphasis on improving weather and climate prediction through data-driven methods and numerical modeling. He leads research efforts machine learning for coupled Earth system modeling and has contributed to the development of advanced AI frameworks for climate model emulation and bias correction. He received his Ph.D. in Atmospheric Science from the Scripps Institution of Oceanography in 2022 and a B.S. in Environmental Engineering from the University of California San Diego. Dr. Chapman has held research positions at the National Center for Atmospheric Research (NCAR), including as a Project Scientist and Advanced Studies Program Postdoctoral Fellow, and collaborates broadly on advancing machine learning applications for Earth system science, including work on emulation frameworks and next-generation climate prediction systems. _________________________________________________________________________________________ Zoom link: https://uib.zoom.us/j/68304284910?pwd=2IgsDMWHuJlQw3XFHSTo3OoGBsRrhz.1 Meeting ID: 683 0428 4910 Password: 7pwZK4mG

As a part of the BCCR Training Programme in Machine Learning, there will be a seminar on AI/ML research at the BCCR on Monday, June 1st from 11:00 to 14:00. We will have two sessions of two talks each, with a free lunch provided in between. The speakers will present their work using neural networks, with a focus on what type of neural network they use and how they help them solve their problems. Please sign up for a free lunch here: https://forms.cloud.microsoft/pages/responsepage.aspx?id=vCSKZI2pJUCcYEjBmhQgaemxD9CtoRROnJCKyVCOholURTRMWUtRNkdHN0xCMUpNVzhEU1ZNSUdOTS4u&route=shorturl Here is the current programme: 11:00 - 11:30 - Kamilla Wergeland from NORCE and Småkraft AS: Improving Day-Ahead Production Forecasts for Run-of-River Hydropower Using Long Short-Term Neural Networks. 11:30 - 12:00 - Antoine Bernigaud from NERSC: Super-resolution of sea surface height: U-net VS GAN. 12:00 - 13:00 - Lunch break 13:00 - 13:30 - Yangfan Zhou from GFI: Using deep learning for precipitation downscaling in Norway. 13:30 - 14:00 - Julien Brajard from NERSC: End-to-end AI forecasting of Arctic Sea Ice from observations.