My first time at #EGU in Vienna, so many great people and exciting research. Amazing ML for Climate Science session. If you want to have some more info on my talk: arxiv.org/abs/2208.05424 #egu2023 #ESAatEGU

High-res climate simulations are essential but computationally expensive. In this work, we present an FNO-based method to downscale climate data to arbitrarily high resolution.(arxiv.org/abs/2305.14452) Our method beats state-of-the-art even on resolutions it wasn't trained on. 🧵

Über #Aerosole hat sich #Corona verbreitet und sie beeinflussen sogar unser Klima. Paula Harder vom Bereich »High Performance Computing« forscht mithilfe von #MachineLearning daran, wie man diese Partikel besser in Klimamodellen erfassen kann. Mehr➡️ itwm.fraunhofer.de/projekt-aeroso…

LEAP welcomes paulaharder.bsky.social to our ML Journal Club this week! THURSDAY, 10/26 @ 3:30p (EDT) leap.columbia.edu/events/ Columbia University Columbia Climate School Columbia Engineering U.S. National Science Foundation Fraunhofer ITWM #climate #compsci #ML #AI #climateadaptation

The ML community has been calling for a large scale climate dataset. In our recent Neurips 2023 publication, we introduce ClimateSet: A ➡️ large-scale ➡️ consistent ➡️ ML-accessible climate model dataset 🌍. 1/8 arxiv.org/abs/2311.03721 climateset.github.io

🌎🌍🌏 Happy to share that our paper "Hard-Constrained Deep Learning for Climate Downscaling", excellently led by paulaharder.bsky.social, is now online at JMLR Journal of Machine Learning Research! jmlr.org/papers/v24/23-… This thread by Paula is a great summary 🧵

Happy that this great downscaling overview paper is out: Enhancing Regional Climate Downscaling through Advances in Machine Learning doi.org/10.1175/AIES-D… Honored to be part of this work led by Neelesh Rampal

Come by today 4.30-6.30 Hall B #4 "Hard-Constrained Deep Learning for Climate Downscaling" ICLR 2026 presenting the work below done at Mila - Institut québécois d'IA supervised by David Rolnick #iclr2024

It was a great pleasure to speak at the AI4DifferentialEquation Workshop yesterday ICLR 2026 and meet the AI4Diff community

Who is this climber? And what was her great ICLR poster about? Read our full review in the latest Computer Vision News: rsipvision.com/ComputerVision… Meet the scientist behind the science!!! #ICLR #MachineLearning #womeninSTEM

Congratulations to paulaharder.bsky.social for her successful PhD defense. Joint work at RPTU Fraunhofer ITWM and Keuper Labs on „Physics-Constrained Deep Learning for Accelerating Climate Modeling“