AI4PEX
banner
ai4pex.bsky.social
AI4PEX
@ai4pex.bsky.social
270 followers 280 following 63 posts
We are a consortium of 19 institutions from all over Europe involved in the Horizon Europe funded project: Artificial Intelligence and Machine Learning for Enhanced Representation of Processes and Extremes in Earth System Models. https://ai4pex.org
Posts Media Videos Starter Packs
💡We’re thrilled to announce the Joint Winter School “AI for Earth Hazards & Climate Extremes”!

📍 NTUA Athens | 📅 March 16–20 2026

A week where AI meets climate science. Stay tuned!

#MeDiTwin #AI4PEX #AIWinterSchool2026
☁️ 🌍 Cloud alert!
Warming oceans = fewer marine stratocumulus clouds → less sunlight reflected → more warming. Spotlighted in a new #EUMETSAT case study.
These subtle clouds play a huge role in our climate system — but they’re tricky to model.

Check out the full case study here 👉
EUMETSAT - User Portal
EUMETSAT User Portal
buff.ly
🌍 Earth system models have come a long way. #AI4PEX improves cloud formation, ocean-atmosphere interactions, and biological processes in the carbon cycle to get a better grip on what the future holds and guide policy decisions.

👉 Stay tuned for exciting updates!

#ClimateScience #Innovation #ESM
💡New paper!
Ocean models often struggle to simulate accurate #SouthernOcean circulation, resulting in unrealistic warm conditions. Check out how Leith viscosity parameterisations in NEMO improves the Antarctic circumpolar circulation to model more realistic #oceantemperatures.

👉 buff.ly/HWCbk71
Reposted by AI4PEX
🆕 A quick guide to an elusive concept!

💡 #Storylines can help #climate scientists better convey a message and help average people understand the challenges they may face in the future due to #ClimateChange.

▶️ What's a Storyline? (Three different approaches): youtu.be/E0SzovQjQOc
What's a Storyline? (Three different approaches)
YouTube video by EERIE Project
youtu.be
Reposted by AI4PEX
Traditional #ClimateModels have long relied on physics. What happens when statistics steps in?

Vincent Verjans (@bsc-cns.bsky.social) is testing how statistical techniques can complement predictions from climate models. 📊🌦️

Read more: expect-project.eu/news/hindcas...

#EXPECTproject #HorizonEU
Reposted by AI4PEX
Climate models can be a challenge to understand, but a new app from the NorCPM-team is using AI to make complex data understandable.

- The idea is that anyone could use it, says Eurico Noleto-Filho, a postdoc with the University of Bergen and the Bjerknes Center. 🧪

bjerknes.uib.no/en/news/brin...
Bringing Climate Models to Everyone
How AI is making complex data understandable
bjerknes.uib.no
Reposted by AI4PEX
Project Spotlight #6: ELLIS Unit Jena @ai4pex.bsky.social 
🌍 AI+Earth observations to advance climate science
⚡Better process representation
🌱Future climate extremes & risks
🛠️Fast, transferable tools for resilience
👨‍🔬Coordinator: Nuno Carvalhais (@bgc-jena.mpg.de)
🔗 ai4pex.org
🌱 IPSL-CM: Paris-built climate–biosphere integration links LMDZ atmosphere, ORCHIDEE land, NEMO ocean, and PISCES biogeochemistry.

Our updates: enhance ORCHIDEE land processes & develop LMDZ-ORC coupling for sharper climate–carbon feedbacks.

#ClimateModel #IPSLCM #ORCHIDEE #ESM
🇪🇺 EC-EARTH: a European community climate model coupling ECMWF’s IFS atmosphere, NEMO ocean & LPJ-GUESS vegetation.

We’re upgrading LPJ-GUESS land processes and vegetation dynamics to better capture ecosystem–climate feedbacks.

#ECEarth #ESM #ClimateModeling #ClimateScience #LPJGUESS
🌍🌊 CNRM-ESM: coupling climate physics & ocean biogeochemistry.

Built by Météo-France/CNRM, it couples atmospheric dynamics (CNRM-CM) to NEMO–PISCESv2 ocean module capturing nutrient-plankton-carbon cycles.

Our updates: boost atm. physics, ocean dynamics & carbon–nutrient cycles.

#ClimateModeling
🇬🇧🌊 UKESM: The UK’s flagship #ESM linking atmosphere, ocean, land & ice with interactive carbon & chemistry cycles.

What we do: upgrade ocean processes in HadGEM & NEMO, enhance JULES land surface, & improve climate–carbon feedbacks.

#UKESM #ClimateModeling #ClimateScience
🌍🌐 ICON-ESM: Climate modeling on a whole new grid

The first fully coupled #ESM on an icosahedral grid — no pole problem, uniform global resolution & ready for ML-powered #ClimateScience.

What we do: add ML parameterisations & improve JSBACH land processes for more realistic vegetation dynamics.
🌍 Meet the #EarthSystemModels powering our project!

Over the next weeks we’ll share 5 unique #ESMs — what makes them special, the science they enable, & how we’ll advance them.

🚀 ICON-ESM, UKESM, EC-EARTH, CNRM-ESM, IPSL-CM.

#ClimateModeling #AI4Climate
💡 New paper!
Climate extremes are impacting forests across Europe. Drought–heat compound events diminish resilience and prolong recovery times, with temperate and boreal forests lacking adaptations leading to physiological damage that increasing sensitivity.

👉 Find out more: buff.ly/w6ChLXa
💡 New paper!
Droughts are complex hydrometeorological phenomena driven by large-scale land–atmosphere feedbacks. Using a new parametric calibration algorithm improves Deep Learning framework reliability in drought detection as showcased on European drought events.

👉 Find out more: buff.ly/llm8RRK
💡 New paper!
Find out how recent significant improvements of #ESMValTool allows a more effective evaluation of complex #ESMs and high-resolution models while needing less computational resources.

👉 Join the discussion: buff.ly/CyycyeP
🚀 First GA for #AI4PEX in Lund!
Last month partners from across Europe met to advance #EarthSystemModels with #AI, #ML & #EarthObservation - tackling ✨ Benchmarking 🛠 Syncing milestones 🌍 Boosting outreach 🤝 Hands-on collaboration.

Big thanks to Lund University for hosting!
👉 www.ai4pex.org
🚀 Excited to share our team’s contributions at the ESA Living Planet Symposium next week in Vienna! Catch us across multiple sessions on carbon fluxes, EO data consistency, AI for climate science & more.

#LPS25 🌍🛰️ #EarthObservation #EarthSystemModeling #CINEA
Fourth and last day of our #AI4PEX GA is again reserved for workshops. Moderated by Fangfei Lan from UNIL we spend the morning focusing on Neural Operator Learning. In the afternoon, we build our own hybrid model using the SINDBAD modeling framework, led by Nuno Carvalhais from MPG-BGC.

#CINEA
AI4PEX GA Day 3 we are mainly focusing on the technical and methodological aspects: How to improve our benchmarking? And what can explainable AI do to improve our parametrizations? Highlight was the keynote by Dave Lawrence from UCAR on how #ArtificialIntelligence is entering the #ESMs development.
#AI4PEX GA Day 2: starting with in-depth scientific talks and posters on our main achievements so far, before discussing in more detail next steps for the atmosphere, ocean and land teams.
In the evening we went to Kulturen Open Air Museum, immersing in the history of the region.

#CINEA #HorizonEU
Today, we started with a workshop on Doing Interdisciplinarity to set the stage for an effective communication across disciplinary boundaries, moderated by Catrin Finkenauer from @utrechtuniversity.bsky.social.
Today we're kicking off the 1st #AI4PEX General Assembly in Lund, Sweden, hosted by @lund-university.bsky.social. Within the next 4 days we will talk about achievements, discuss methods and plan research of next year, while getting closer, building trust and turning our team into a project family.