If Influence Functions are the answer, then what is the question?
In our new work with @isaac_o_reid, we revisit this question, and give a new distributional perspective under which influence functions (IFs) turn out to make a lot of sense for deep learning!
Can we use ML onboard spacecraft to identify pollution?
Our Starcop 2.0 team has been pushing the edge of what’s possible, on the edge.
Join us Aug 8 as they share how AI + hyperspectral data can spot hard-to-see atmospheric anomalies 🌍✨
🔗 RSVP: https://t.co/yHieLOCSol
🏔️☔️ How can we improve daily precipitation mapping across complex mountain regions? Our new work @EGU_HESS tackles this by leveraging regional climate models, in-situ observations and a probabilistic machine learning framework — check it out! https://t.co/RSfrjVByDT
Precipitation prediction from large-scale climatic features over the Upper Indus Basin using Gaussian processes – part of the Environment and Sustainability seminar series
🕒 1pm | 📅30 October | 📍 Online | ➡️ Register now: https://t.co/N2BdwzZuYv
#Sustainability#DataScience
Discussion open for comments on @EgUsphere: Probabilistic precipitation downscaling for ungauged mountain sites: a pilot study for the Hindu Kush Karakoram Himalaya https://t.co/RTvVm4wCc0
🚨New preprint🎉
"Generative Data Assimilation of Sparse Weather Station Observations at Kilometer Scales"
https://t.co/FbfxNyeDvA
An absolute pleasure working on this during my internship @nvidia with @NoahBrenowitz @SciPritchard@JaideepPathak@tropmetpie and others 🧵
📄📌PRESS RELEASE:
Alarmingly low snow levels this year in Hindu Kush Himalaya threaten water security for hundreds of millions.
2024 - an "extraordinary below normal snow year" according to ICIMOD’s Snow Update Report, and could have devastating consequences for large downstream population centres that are reliant on mountain water.
Snow persistence (the time snow remains on the ground) has plummeted, and falls are particularly steep in the western Hindu Kush Himalaya region, where snow’s contribution to water supply is the highest.
Persistence is down approx.
🔹30% in Helmand, Tarim, amd Amu Darya rivers;
🔹20% in Indus, and Ganges;
🔹15% in the Tibetan Plateau, Yangtze and Brahmaputra.
Only the Yellow River sees a dramatic increase in snow persistence: up approx. 20%
ICIMOD warns relevant water management officials to initiate drought management strategies and calls for rapid decarbonization of economies.
Read the full report: https://t.co/fmtj0MqzOt
Read the press release: https://t.co/wAAVKZK68E
#ClimateCrisis #WaterScarcity #SG1 #SaveOurSnow #HKH #SnowUpdate2024
@Dr_Msher@aisha4climate@DrMiriamJackson@SunitaMQ@NakulChettri5@PhilippusWester@jakobfsteiner
⚡️ Lightning talks from @KenzaxTazi, @GironaMata, Ilenia Manco, and Daria Botvynko spanning geographies from the Himalayas to the Alps to the Mediterranean to the deep ocean!
⌚️ Day 2 is speeding along here at #ClimateInformatics 2024, hosted by @turinginst.
Presenting our work on probabilistic downscaling of daily precipitation in ungauged mountain regions at #EGU24 this afternoon! Catch our talk in Room 2.17 at 17:25 CEST. 🏔️🌧️ @EuroGeosciences
Environmental observations are messy. Could a single ML architecture fill gaps, increase granularity, and tell us where more data is needed?
I'll be discussing this & more at the NERC 'AI for Environmental Science' webinar this Friday - register here: https://t.co/eIJasepxNX
15 days to the submission deadline for Climate Informatics 2023!
📆 Submission Deadline: 26 January
🧑🏫 Conference: 19-21 April
📌 Location: Cambridge, UK
📚 Fields: Climate + ML + AI
🖥️ Website: https://t.co/iUOP11D8Q8
@Cambridge_ICCS@EnvDataScience@CambridgeUP#CI2023
Per fi ha arribat el dia!
A 2/4 de 7 a la Taverna de la @FMediterrania presentem #epursiballa i ens hi acompanyaran en @candidxeremier i els #Ballaveu .
Encara queden entrades!
Ah, i tindrem discs (@Seedmusic_eu )
Delighted to share my first paper, out today in Nature Communications: https://t.co/NS79mXc49o. Beyond grateful to my brilliant coauthors who helped me take this much further than I could alone. Truly hope this makes a positive difference in the Arctic. @BAS_News@turinginst
Àvia sana de 92 anys. Fractura de fèmur. Urgències. De moment, 6 dies ingressada a @harnaulleida i encara ESPERANT ser operada. Indignació!! Vergonya!! Això no es culpa de la covid @salutcat@icscat Que passa amb la gent gran en aquest país? @maticatradio
És a dir, si una dona de 90 anys immòbil en un siti de llit amb un fèmur trencat, sense poder menjar ni beure per si troben un forat, no és greu, aleshores què ho deu ser. Què pretenen estalviar-se l'operació?
Qui són els responsables de tanta ineptitud i caradura?
La raó que els donen per explicar els constants endarreriments, amb tot el que comporta, abstinències de 12 hores, gairebé un dia sencer sense menjar ni beure, és que estan desbordats i que aquest no és un cas urgent.