Calendar

Instead of the workshop that was initially planned, we have opted for a smaller format with two invited talks and a discussion session: Huiji Lee: Alkalinity Redistribution Leads to Weakened Global Ocean CO₂ Uptake Pushpak Nadar: Enhanced Antarctic Intermediate Water Ventilation During Past AMOC Weakening and Implications for Thermocline Biogeochemistry Both speakers will present their work in a way that is accessible to a broad audience. After the talks, Are will lead a discussion session, followed by a joint lunch. We also hope that Ralph Keeling will be able to join us. Time: 23 June, 11:00–13:00 Lunch: Provided free of charge If you would like to attend, please fill out the registration form by 12 June (so that we know how much lunch we should order): Carbon Theme meeting, 23.06., 11-13h – Fill out form

Dear all, The next BCCR Monday Seminar will be given by Eurico Noleto-Filho from the Global group. He will present his work on "Ask the Climate: Voice and Chat AI Agents for Ocean Data”. The seminar will take place 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 Climate prediction models are powerful tools for projecting future environmental changes, yet their complexity (high-dimensional datasets, spatial maps, and probabilistic forecasts) makes them difficult to interpret for non-specialists. This limits meaningful engagement by researchers, policymakers, and the public, and constrains how climate information feeds into applied sectors such as fisheries and energy. Recent advances in large language models (GPT-5.4, Claude Sonnet 4.2), real-time voice interfaces (OpenAI Realtime API), and agentic coding tools (Claude Code, OpenClaw) have reshaped what is possible in this field. These systems no longer simply describe model outputs; they actively query datasets, execute code, and reason over climate data in dialogue with the user. This shift has prompted us to rethink how LLMs can be embedded in climate model workflows, and how voice and chat modalities can broaden the reach of climate science. In this talk, we present the latest advancements in our NorCPM-AI app, including new chat and voice agent features. We discuss the state of the art of these agentic and multimodal capabilities, their remaining limitations in numerical reasoning and uncertainty communication, and case studies linking NorCPM forecasts to fisheries and renewable energy. Speaker information I am a biologist with extensive experience in applying artificial intelligence, statistical modeling, and web application development to marine resource management initiatives (Blue Economy). My expertise includes Machine Learning, Hierarchical Bayesian Models, Spatiotemporal Models, and web app development, with proficiency in R and Python. I have strong programming skills and the ability to design innovative data analytics applications for governmental, corporate, and academic purposes. Additionally, I am the author of several web applications for marine resource data analytics, modeling, and forecasting. Join Zoom Meeting https://uib.zoom.us/j/68304284910?pwd=2IgsDMWHuJlQw3XFHSTo3OoGBsRrhz.1 Meeting ID: 683 0428 4910 Password: 7pwZK4mG