So mentioning multiple countries for Uranium misses the problem that also the Kazakh Uranium is controlled by Russia (Rosatom). Main problem of nuclear itself beside the security issues is that it is structurally incompatible to highly volatile renewables. Nuclear is generally not forward pointing.
That study is not really a surprise: agupubs.onlinelibrary.wiley.com/doi/full/10.... AI depends on training data as any other statistical prediction and training only on yesterdays weather/climate will deliver you yesterdays weather/climate.
Submit! 😊
The story is a bit different. The Problem with the bombs in Germany during WW2 was that about 10-20% did jot explode and that currently about 5000 a year need to be diffused (about half of them in NRW). Main issue was that the cities were rebuild over the unexploded bombs.
I’m grumpy about this paper, which has been poorly communicated in a way that’s taken up a bunch of my time this week and is the result of framing decisions in the paper itself.
We would need an independent scientific email service. Once you get a PhD it should provide you with a livelong address and you can then redirect it to wherever you like. No hassle anymore with institution change or loosing job. Sad that such a service has not emerged in the 1990s.
Walking in a winter wonderland… my amateurish measurement gave 24cm this by morning on the balcony, and who would have thought that even CPH can have road trouble (poor guy who got stuck this morning for 20 minutes just as he wanted to drive the 20 m from the street to the car park).
An interesting week of discussions came to an end in Lecce. Now it means going back to the much more grey North.
End of year look at paper projects (* = first author):
published 2025: –
in review: PL, NW, ND
in write up: HY*, BG, KA, EA, KP
in work: TD*, IA*, HA + some on the horizon #researchlife
And back home again. Perhaps the last time with the old IC 3, as they will be replaced in the next months by the IC 5.
Were the 15% statements by the students themselves or were it identified by a checking-software?
Never forget!
Well, there might be many who do not even remember all the one-off accounts they did over the years. Especially in case of email adresses weren't there apple who created one-off emails on the spot?
Not sugar-coating it: We have a reviewing crisis!
It needs nowadays months to find reviewers, with potential reviewers exhausted and those suffering are mainly early career. Just waiting with a paper more than 3 months for a 2nd reviewer, with 12 unsucessful search rounds done. #fridayfrustration
Cyber security is a tricky field, but when providers still work with outdated guidelines it becomes almost impossible to keep things save. #fridayfrustration
Three exciting days at the workshop for annual to multi-decadal prediction in beautiful Bologna.
It confirmed a theory of mine before I could submit it. Well, might be an easier review now as it won't be just a speculation (but yeah, still needs a bit of time).
Depends. The argument of the KI publishers is generally that it is not possible to extract source data. This court case claims otherwise. For everything KI is currently used, which has potentially data security implications, there might be new challenges ahead. KIs fault: Gives source data for free.
The consequences can be wider: Anybody publishing a KI result publicly might be liable, if what has been published might stem from a book or similar. As some KI learn sources behind paywalls the user is unable to check this. Consequently: Any posting of KI results might be a risk. We‘ll see.
In this case it is about republishing without licence. So if a song (or its lyrics) gets published in the web without getting the ok from the licence holder this is copyright infringement and you become liable. Same with any streaming service and in this case KI repeating close enough lyrics online.
Europe’s #heatwaves are not only getting hotter — they’re becoming more chaotic.
A new #NatureCommunications study by Goratz Beobide-Arsuaga (ESRAH, Univ. Hamburg) shows how soil moisture fuels extremes in Central and Northern Europe.
🔗 www.esrah.uni-hamburg.de/about/news/0...
🌊🧊🇦🇶
I am happy and proud to present our #platformist team effort out in @nature.com!
www.nature.com/articles/s41...
We examine if, when, and why #Antarctic #ice shelves will no longer be viable, at the latest, due to changes in #atmosphere and #ocean conditions.
A little 🧵 for the experts...
1/7
Yes, the 6-9 m is a good estimate, but there are many uncertainties around it. It is probably best seen as the agreed compromise, as it is complicate to get indicators in line with the models (my old take on it: academic.oup.com/gji/article/...). Not even talking about the statistical assumptions.
Might it mean that libraries in DK are able to afford more books now?
Advertising this PhD student position once more (deadline 30 August) workingat.nioz.nl/o/phd-positi...
We are looking for someone with a MSc in physical oceanography, fluid dynamics or physics to study how ocean eddies contribute to the Atlantic Meridional Overturning Circulation
Yes. Or simpler: taking n years, calculate the anomaly towards the mean by ENSO state (so their annula cycle, probably from Aug/Sep each year starting) and then reranking them after they are corrected. Might be a relative straight rise (as it would also correct for annual cycles).
How would the table look like, if we would calculate the anomalies towards years with El Niño, La Niña or Neutral (whatever fits the actual year)?
The second link delivers "Page not found".
Best time to cycle the roads in IRL (some might say the only) is Sunday mornings. The elderly are stitting in the church and the others negotiate their pints from the night before. So it is quiet on the roads and halfway safe (unless someone is late to church or comes too late home from the pints).
The history and rivalary between cars and bikes in Ireland is a complex one. Mixed with often much too narrow roads and the wide spread lack of cycle paths (due to the complex history) lead to the rest.