3 days, 9 models, 1 goal: benchmarking mortality in #EarthSystemModels 🌍
From data & model overviews to hands-on coding sessions, we worked toward a shared framework, common metrics, and first results for a joint paper.
Great collaboration across teams & time zones! 🚀
#AI4PEX postdoc position available at @unituebingen developing #MachineLearning approaches to improve #EarthSystemModels!
👉Join our team: https://uni-tuebingen.de/de/290466
🌍 New series: 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, CNRM-ESM, EC-EARTH, IPSL-CM.
đź’ˇ 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: https://www.sciencedirect.com/science/article/pii/S1569843225002109
📢 News from #nextGEMS: Our latest #PolicyBrief is out—“Reaching the Last Mile with Km-Scale #EarthSystemModels.”
It highlights how km-scale models deliver sharper #climate insights, especially on extreme weather and regional projections. But better data ≠better #policy—yet.
🔑 Key proposals:
Support transdisciplinary #collaboration
Remove barriers to data access
Recognise & retain early-career talent
Build public-private partnerships for future-ready infrastructure
We call on policy-makers and funders to help ensure this cutting-edge science reaches its societal potential.
đź“Ą Read the brief via our Media Library
👀 A second brief is coming soon—stay tuned!
🚀 First General Assembly for #AI4PEX in Lund!
Last month partners from across Europe met to advance #EarthSystemModels with #AI, #ML & #EarthObservation. 4 days packed with:
✨ Shaping benchmarking tools
đź› Reviewing progress & syncing milestones
🌍 Boosting outreach & community building
🤝 Hands-on collaboration
Big thanks to Lund University for hosting & to all driving this forward!
👉 www.ai4pex.org
The added value of machine learning for weather and climate applications is measurable through performance metrics, but explaining it remains challenging, particularly for large deep learning models. Inspired by climate model hierarchies, we propose that a full hierarchy of Pareto-optimal models, defined within an appropriately determined error-complexity plane, can guide model development and help understand the models' added value. We demonstrate the use of Pareto fronts in atmospheric physics through three sample applications, with hierarchies ranging from semi-empirical models with minimal parameters to deep learning algorithms. First, in cloud cover parameterization, we find that neural networks identify nonlinear relationships between cloud cover and its thermodynamic environment, and assimilate previously neglected features such as vertical gradients in relative humidity that improve the representation of low cloud cover. This added value is condensed into a ten-parameter equation that rivals deep learning models. Second, we establish a machine learning model hierarchy for emulating shortwave radiative transfer, distilling the importance of bidirectional vertical connectivity for accurately representing absorption and scattering, especially for multiple cloud layers. Third, we emphasize the importance of convective organization information when modeling the relationship between tropical precipitation and its surrounding environment. We discuss the added value of temporal memory when high-resolution spatial information is unavailable, with implications for precipitation parameterization. Therefore, by comparing data-driven models directly with existing schemes using Pareto optimality, we promote process understanding by hierarchically unveiling system complexity, with the hope of improving the trustworthiness of machine learning models in atmospheric applications.
What are the climate models we are working on? Land model number 3:
ORCHIDEE (Organising Carbon and Hydrology In Dynamic Ecosystems), the land surface model of the IPSL #EarthSystemModels, where our focus is especially on the hydrological processes.
👉 Learn more: https://buff.ly/83X23ho
What are the climate models we are working on? Let’s start with the land models:
JULES, the Joint UK Land Environment Simulator: a community land surface model that can be coupled to the Met Office Unified Model where we will build high-dimensional emulators to calibrate land surface behavior.
👉Learn more: https://buff.ly/ZJY5eZJ
AI4PEX
💧🌏 Greg Cocks
nextGEMS