Happy to be a co-convener for the "ML for Climate Science" session together with @DWatsonParris, @NowackPeer@PaulaHarder8 & Tom Beucler at #EGU25. We are honored to have @lau_mansfield as a speaker. Looking forward to your abstract submissions.
Thrilled to unveil our latest work: multi-modal machine learning to forecast localized weather! We construct a graph neural network to learn dynamics at point locations, where typical gridded forecasts miss significant variation.
Paper: arxiv.org/abs/2410.12938
🧵
Very happy to share Henry's excellent work applying machine learning to achieve km-scale downscaling of rainfall in climate simulations! It's far cheaper than running km-scale regional models, opening up applications to e.g. studying extreme events in large simulation ensembles.
Excited to announce a pre-print of my first paper as part of my PhD at @BristolUni! We’ve created a generative ML emulator of a UK convection-permitting climate model (CPM). It’s able to produce simulations of high-res precipitation at far lower computational cost.
🔥Are you ready for the battle against megafires? 🔥
Wildfires are becoming more frequent and more catastrophic.
Discover our new masterclass by Pr. @atvoulgarakis@AXA Chair in Wildfires and Climate at @tuc_chania
➡️youtube.com/playlist?list=PL…
There's a been a lot of great discussion about this figure in the past day - and we should absolutely celebrate the incredible success of numerical weather prediction as a technology. But this figure doesn't tell a complete story. 1/n
Weather forecasting has come a long way.
The biggest improvements we’ve seen are for longer timeframes.
By the early 2000s, 5-day forecasts were “highly accurate” and 7-day forecasts are reaching that threshold today.
3 levels of data scientist job in the Earth System and Mitigation Science team at the @metoffice Hadley Centre. Link to apply: metoffice.gov.uk/about-us/ca…
Check out my new preprint on uncertainty quantification for a machine learning parameterization for gravity waves in a climate model 🌐🌎☁️ doi.org/10.22541/essoar.1709…
Really excited to share our new paper on climate-invariant machine learning science.org/doi/10.1126/scia… to solve extrapolation issues under climate change, led by the great Tom Beucler
Ein internationales Forscherteam mit Beteiligung von KIT-Informatikprofessor Peer Nowack hat einen Datensatz entwickelt, um mit Methoden des Maschinellen Lernens schnellere Berechnungen von Klimawandelszenarien zu ermöglichen. #KI#Klimainformatik.kit.edu/11147_135…
In search of a New Year's Research Resolution? Submit to #CI2024 in London hosted by @turinginst!
📆 Submission Deadline: 2 February
🧑🏫 Conference: 22-24 April
📌 Location: London, UK
📚 Fields: Earth Climate Weather ML AI Data
🖥️ Website: alan-turing-institute.github…
The 2023 Southern Hemisphere #ozonehole finally closed on 20 December, becoming the 7th longest-lived in the #CopernicusAtmosphere records. The animation shows the evolution from its early start in August to this late closure. Find out more:
atmosphere.copernicus.eu/lar…
Interested in the response of convection to climate and air pollutions changes in global km-scale models?
We are looking for a post-doctoral researcher in the @HorizonEU project @CleanCloud_HE with collaborators across Europe.
Apply by 31 January:
my.corehr.com/pls/uoxrecruit…
November updates for global (land ocean HadCRUT5), global ocean surface (HadSST4) and global land surface air (CRUTEM5) temperature anomalies are now available: crudata.uea.ac.uk/cru/data/t…
Year-to-date (Jan-Nov) anomalies for 2023 exceed all previous years @ClimateUEA_ @ueaenv
ALT Three global temperature anomaly timeseries, top for the full global (land ocean) temperature anomaly from HadCRUT5, middle for the global ocean surface temperature anomaly from HadSST4, and bottom for the global land surface air temperature anomaly from CRUTEM5.
New benchmark dataset for climate scientists and machine learners out now in Neurips! A definite read if you are interested in climate model emulation and large-scale data 👇
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/2311.03721climateset.github.io