An international consortium improving our understanding and representation of gravity waves in climate models through machine learning and balloon observations
Our new review paper "Machine Learning for Climate Physics and Simulations" with Pedram @turbulentjet, Aditi, Maike, Raffaele & Balaji is online @AnnualReviews 🎉. Beyong accelerating simulations, can ML help us understand climate physics? We highlight the recent progresses & challenges. Share with us your favorite #AI4climate papers!
doi.org/10.1146/annurev-conm…
Open-source arxiv version here: arxiv.org/abs/2404.13227
Lead PI Aditi Sheshadri sits down with host Russ Altman on The Future of Everything podcast to discuss how new data and techniques are reshaping the future of climate projection. Check it out here: engineering.stanford.edu/new…@StanfordEng
Excited that our paper on uncertainty quantification of subgrid-scale parameterizations for gravity waves has been published! Work with Aditi Sheshadri funded by @DataWave_VESRIagupubs.onlinelibrary.wiley.…
📢 Are you an Engineer with an interest in Climate Science?
We're looking to hire an Engineering Lead to steer our highly-specialised RSE team
More information can be found here: jobs.cam.ac.uk/job/45880/
Closing date 17 May 2024
youtube.com/watch?v=nSLOp5sS…
@hpcchris @dorchard
EGU Spotlight: DataWave PIs Claudia Stephan and Ulrich Achatz will be convening a session Tuesday called "Internal Gravity Waves." DataWave ECRs Yanmichel A., Sothea H., Felix J., Ray C., Iman T., and Aman G. will be presenting posters/talks during the session. Stop by!
Check out the new paper on "offline-online learning" of parameterizations for climate/turbulence to address shortcomings of supervised learning led by NWRA scientist Hamid Pahlavan published in @theAGU Geophysical Research Letters: doi.org/10.1029/2023GL106324#NWRA_papers
ALT The a posteriori performance of the CNN (a) in big-data regime, (b) in small-data regime and (c) after on-line re-training the CNN initially trained in the small-data regime. The period (τ) and amplitude (σ) of the true QBO are 28 - 0.7 and 21 - 0.3 m/s, respectively. The probability density functions (PDFs) of F and u become more similar to the true QBO after online re-training the CNN initially trained in the small-data regime. Copyright (2024) American Geophysical Union.
A postdoc position is available in the climate dynamics group at Stanford with Aditi Sheshadri. Projects could include gravity waves (datawaveproject.github.io/), machine learning parameterizations, etc. Message me or email Aditi if you want to know more: eddy.stanford.edu/
Check out new research from our Stanford team comparing Bayesian History Matching to Ensemble Kalman Inversion when calibrating a gravity wave parameterization. @HeIsRobKing @lau_mansfieldessopenarchive.org/doi/full/…
Jobs Alert! Check out our website for new atmospheric dynamics positions open for student research assistants, postdocs, and phd students across our different DataWave research groups. Please share! datawaveproject.github.io/jo…
AGU '23 is all wrapped up! We had a great week full of DataWave posters, talks and sessions. Read more about it on our blog datawaveproject.github.io/bl…
Our group is moving to The University of Chicago! @UChicago. Excited to expand our work on weather extremes, scientific ML, climate change & turbulence with amazing new colleagues in Geophysical Sciences @GeoSci_UChicago, Computational & Applied Math, AI Science @DSI_UChicago ...
Why shouldn’t we just write climate models and machine learning algorithms in the same coding language? In an interview, ICCS software engineer Jack Atkinson talks about the challenges and opportunities in his latest project with @DataWave_VESRIiccs.cam.ac.uk/news/iccs-sof…
PREPRINT ALERT: Explainable Offline-Online Training of Neural Networks for Parameterizations: A 1D Gravity Wave-QBO Testbed in the Small-data Regime, Pahlavan et al, doi.org/10.48550/arXiv.2309.…@RiceUniversity
@APS_GPC virtual seminar is back for its 2nd year! Our upcoming speakers @alli_wing , @turbulentjet, and @GeoffVallis will delve into the outstanding questions and challenges in climate physics. Everyone is welcome to join via the registration link in the thread. :D