Sign in

Tom Andersson 🌍

@tom-andersson.bsky.social
1K followers 192 following 39 posts

Research Engineer at Google DeepMind; Building AI for weather & tropical cyclone forecasting; WMO Young Scientist of the Year 2022; he/him

PostsRepliesMedia
Tom Andersson 🌍 @tom-andersson.bsky.social · 10/08/2026
❤️
010
Tom Andersson 🌍 @tom-andersson.bsky.social · 21/07/2026
I'm looking forward to seeing how our models perform this 2026 NH cyclone season. It was great to see the uptake of 'FNv3' by the community in 2025, particularly WeatherNext helping the NHC predict Hurricane Melissa's historic landfall in Jamaica: deepmind.google/blog/how-wea...
deepmind.google
AI breakthrough: WeatherNext predicts Hurricane Melissa
Discover how our WeatherNext AI model helps the National Hurricane Center predict Hurricane Melissa's Category 5 landfall in Jamaica
000
Tom Andersson 🌍 @tom-andersson.bsky.social · 21/07/2026
I hope this makes our Google AI weather models even more accessible and useful, and we'd love to hear your feedback, either direct to weather-lab@google.com or through the site's feedback button in the upper right!
110
Tom Andersson 🌍 @tom-andersson.bsky.social · 21/07/2026
Note: WN2's cyclone track outputs are a specialised output of the model trained on historical cycloen data, whereas the background grid fields are 0.25 deg channels trained on analysis data, so they won't always agree on tropical cyclone wind strength.
100
Tom Andersson 🌍 @tom-andersson.bsky.social · 21/07/2026
You can also look at older global forecasts initialised over the past 15 days, allowing you to scrub through and see how predictions have evolved for emerging weather events. (For cyclone tracks, you can still view historical forecasts back to 2022).
100
Tom Andersson 🌍 @tom-andersson.bsky.social · 21/07/2026
When I was testing the UI internally in late June, WN2's uncertainty for the path of Typhoon Bavi meant 10% of members had Taipei staying dry, while 10% predicted torrential rain. Bavi passed northeast of Taiwan 10 days later and the actual rainfall was near the upper estimates.
100
Tom Andersson 🌍 @tom-andersson.bsky.social · 21/07/2026
By popular demand, we've updated our interactive Google Weather Lab platform to show WeatherNext 2 ensemble mean gridded forecasts alongside our tropical cyclone predictions, including point forecasts for: temp, precip, wind speed, and sea level pressure: deepmind.google.com/science/weat...
150
Tom Andersson 🌍 @tom-andersson.bsky.social · 18/07/2025
Glad you like Weather Lab, Matt! We also look at all responses we get via the feedback button if you have ideas about how we can make the site more useful.
120
Tom Andersson 🌍 @tom-andersson.bsky.social · 27/03/2025
We're hiring a Research Engineer based in San Fran or MTV, California, to join our sustainability & weather teams at GDM. Seeking strong engineers at the interface of ML, sustainability, weather, dynamical systems, and/or remote sensing: boards.greenhouse.io/deepmind/job...
boards.greenhouse.io
Research Engineer, AI for Sustainability
Mountain View, California, US
0132
Reposted by Tom Andersson 🌍
cira-csu.bsky.social @cira-csu.bsky.social · 16/01/2025
📸: Our researcher Kate Musgrave presenting on AI weather predictions #AMS2025 | @ametsoc.bsky.social
1131
Reposted by Tom Andersson 🌍
Ferran Alet @ferranalet.bsky.social · 14/01/2025
I'll be presenting GenCast, recently published in Nature, tomorrow Tuesday 8h45AM at #AMS2025 in room 339. GenCast is a diffusion model that outperforms ENS, the top operational ensemble forecast, giving skillful probabilistic forecasts up to 15 days ahead. ams.confex.com/ams/105ANNUA...
092
Tom Andersson 🌍 @tom-andersson.bsky.social · 12/01/2025
Ferran's GenCast talk on Tuesday morning at #AMS2025 is not one to miss! ams.confex.com/ams/105ANNUA...
ams.confex.com
GenCast: Diffusion-based Ensemble Forecasting for Medium-range Weather
Weather forecasts are fundamentally uncertain, so predicting the range of proba...
0171
Tom Andersson 🌍 @tom-andersson.bsky.social · 10/01/2025
I'll be at #AMS2025 next week alongside some other Google DeepMind colleagues behind GenCast/GraphCast. Excited to meet people, discuss ML for weather, and learn!
070
Tom Andersson 🌍 @tom-andersson.bsky.social · 06/01/2025
Welcome, Jeff, and thanks for the links!
030
Tom Andersson 🌍 @tom-andersson.bsky.social · 29/12/2024
Can everyone come over from LinkedIn now 😅 There’s still more of the ML/Earth sciences community active on there than BlueSky I feel
070
Tom Andersson 🌍 @tom-andersson.bsky.social · 29/12/2024
dynamical tests would be in my top 3 too :-)
030
Tom Andersson 🌍 @tom-andersson.bsky.social · 27/12/2024
See: twitter.com/elonmusk/sta...
twitter.com
Elon Musk on X: "Mars will be called the “New World”, just as America was in past centuries. Such an inspiring adventure!" / X
Mars will be called the “New World”, just as America was in past centuries. Such an inspiring adventure!
000
Tom Andersson 🌍 @tom-andersson.bsky.social · 27/12/2024
Yeah ofc there are indirect benefits from crewed space exploration and non-Earth-related spacecraft, even if just for the sake of knowledge I’m speaking more to a shift in myself while at uni, realising there are too many urgent problems on Earth for some Muskian Mars colonisation project lol
120
Tom Andersson 🌍 @tom-andersson.bsky.social · 27/12/2024
I love how William Shatner from Star Trek went to space and realised how horrible it is in contrast to our beautiful home planet. My life goal used to be to help get humans to Mars. I’m so glad I realised how special Earth is and now work on better living here rather than leaving 🌍
160
Tom Andersson 🌍 @tom-andersson.bsky.social · 24/12/2024
See you at AMS!
100
Tom Andersson 🌍 @tom-andersson.bsky.social · 23/12/2024
It's @ecmwf.bsky.social keeping the tradition this year 😉 bsky.app/profile/rasp...
131
Tom Andersson 🌍 @tom-andersson.bsky.social · 23/12/2024
I ran into this as well - there is a separate button for uploading animations (next to the static image one). Nice work by the way!
000
Reposted by Tom Andersson 🌍
Stephan Hoyer @handle.invalid · 19/12/2024
Can incorporating AI improve precipitation in global weather and climate models? Yes! In the latest NeuralGCM paper, we show that training on satellite-based precipitation results in significant improvements over traditional atmospheric models: arxiv.org/abs/2412.11973
arxiv.org
Neural general circulation models optimized to predict satellite-based precipitation observations
Climate models struggle to accurately simulate precipitation, particularly extremes and the diurnal cycle. Here, we present a hybrid model that is trained directly on satellite-based precipitation obs...
1345
Tom Andersson 🌍 @tom-andersson.bsky.social · 10/12/2024
My personal opinion (and I think general consensus) is that while purely data-driven modelling already excels from nowcasting to medium-range timescales, it is not the right paradigm for climate forecasting. Very fast ML/physics hybrid GCMs like NeuralGCM will be the way to go. cc @stephanhoyer.com
030
Tom Andersson 🌍 @tom-andersson.bsky.social · 10/12/2024
Actual rollout schematic animation here:
020
Tom Andersson 🌍 @tom-andersson.bsky.social · 10/12/2024
Looks like Bluesky has a separate button for videos that I missed 🙃 Actual Milton animation here:
140
Reposted by Tom Andersson 🌍
Stephan Hoyer @handle.invalid · 09/12/2024
Interested in AI weather/climate modeling at #AGU24? I'll be giving an overview talk on NeuralGCM at 11:30am Wed at the Google booth, and an talk on modeling precipitation with NeuralGCM at 4:25pm Wed in the session A34A.
1358
Tom Andersson 🌍 @tom-andersson.bsky.social · 10/12/2024
It's been an honour to work on this study led by Ilan Price with such a talented team ✨: Alvaro Sanchez Gonzalez, Ferran Puig, Andrew El-Kadi, Dominic Masters, Timo Ewalds, Jacklynn Stott, @shakirm.bsky.social, Peter Battaglia, Rémi Lam, & Matthew Willson
220
Tom Andersson 🌍 @tom-andersson.bsky.social · 10/12/2024
Like its predecessor (GraphCast), the weights & code of GenCast have been made publicly available: github.com/google-deepm... We’re looking forward to seeing how the community builds on this!
github.com
GitHub - google-deepmind/graphcast
Contribute to google-deepmind/graphcast development by creating an account on GitHub.
252
Tom Andersson 🌍 @tom-andersson.bsky.social · 10/12/2024
A GenCast ensemble member takes 8 minutes on a TPU chip, versus hours on a supercomputer for physics-based models. This opens up the possibility of large ensembles (eg 1000s of members), which could better estimate risks of extreme events. We don't yet know how much this will help.
120
Tom Andersson 🌍 @tom-andersson.bsky.social · 10/12/2024
Cyclone max wind speeds are still underestimated, but this performance on tracks is really promising. One recent devastating cyclone was Hurricane Milton, which caused >$85 billion in damages. GenCast predicted ~70% probability of landfall in Florida 8.5 days before it struck.
120
Tom Andersson 🌍 @tom-andersson.bsky.social · 10/12/2024
We also extracted cyclone tracks from GenCast and ENS and compared them with ~100 cyclones observed in 2019. GenCast's ensemble mean cyclone track has a 12-hour position error advantage over ENS out to 4 days, and more actionable track probability fields out to 7 days.
140
Tom Andersson 🌍 @tom-andersson.bsky.social · 10/12/2024
For example, we created a dataset of simulated wind power data at wind farm sites across the globe, and found that GenCast outperforms ENS by 10–20% up to 4 days ahead. This is promising, because better weather forecasts can reduce renewable energy uncertainty and accelerate decarbonisation.
150
Tom Andersson 🌍 @tom-andersson.bsky.social · 10/12/2024
It’s vital that we ensure these new ML weather systems are safe and reliable. One thing I'm proud of is our range of evaluation experiments: per-grid-cell skill & calibration, spatial structure, renewable energy, extreme cold/heat/wind, and the paths of tropical cyclones (i.e. hurricanes).
130
Tom Andersson 🌍 @tom-andersson.bsky.social · 10/12/2024
GenCast uses diffusion to generate multiple 15-day forecast trajectories for the atmosphere. It assigns more accurate probabilities to possible weather scenarios than the SoTA physics-based ensemble system from ECMWF, across a 2019 evaluation period.
130
Tom Andersson 🌍 @tom-andersson.bsky.social · 10/12/2024
So excited to share our Google DeepMind team's new Nature paper on GenCast, an ML-based probabilistic weather forecasting model: www.nature.com/articles/s41... It represents a substantial step forward in how we predict weather and assess the risk of extreme events. 🌪️🧵
nature.com
Probabilistic weather forecasting with machine learning - Nature
GenCast, a probabilistic weather model using artificial intelligence for weather forecasting, has greater skill and speed than the top operational medium-range weather forecast in the world and provid...
210816
Tom Andersson 🌍 @tom-andersson.bsky.social · 08/12/2024
Here’s one go.bsky.app/9T8hneP
251
Tom Andersson 🌍 @tom-andersson.bsky.social · 01/12/2024
For the ~8 billion minus 200 people who haven't cited it yet, what are you waiting for! www.nature.com/articles/s41...
nature.com
Seasonal Arctic sea ice forecasting with probabilistic deep learning - Nature Communications
Accurate seasonal forecasts of sea ice are highly valuable, particularly in the context of sea ice loss due to global warming. A new machine learning tool for sea ice forecasting offers a substantial ...
010
Tom Andersson 🌍 @tom-andersson.bsky.social · 01/12/2024
Quite happy with IceNet getting its 200th citation a little after its 3rd birthday ❄️🎂💻 IceNet began around the time when applying ML to Earth sciences was the stuff of hushed conversations in corridors. A few years later and its potential to revolutionise the field is indisputable. Lots more to do!
1111
Tom Andersson 🌍 @tom-andersson.bsky.social · 24/11/2024
And when you’re done building or reinforcing your visual intuition for transformers, check out this short note by my old supervisor Rich Turner for a very clear description of the maths behind the architecture! arxiv.org/abs/2304.10557
arxiv.org
An Introduction to Transformers
The transformer is a neural network component that can be used to learn useful representations of sequences or sets of data-points. The transformer has driven recent advances in natural language proce...
010
Tom Andersson 🌍 @tom-andersson.bsky.social · 24/11/2024
Beautiful introduction to transformers in this lecture by 3blue1brown / Grant Sanderson: youtu.be/KJtZARuO3JY?...
youtu.be
Visualizing transformers and attention | Talk for TNG Big Tech Day '24
YouTube video by Grant Sanderson
141
Tom Andersson 🌍 @tom-andersson.bsky.social · 20/11/2024
🌎
011
Tom Andersson 🌍 @tom-andersson.bsky.social · 20/11/2024
Thanks! Suggest adding @raspstephan.bsky.social who I’m trying to convince to move over here :)
010