New book day!
10.03.2026 08:40
👍 20
🔁 5
💬 1
📌 0
New book day!
OK, that's one glowing AI article too many in Nature and one cosy fireside editorial too many from Science.
Can we have a prestige journal with a bit of attitude please, a little bite, a hint of tooth. Articles I want to read not shred. Papers the authors might actually have enjoyed writing.
I agree. A little help from the creators would be useful, but the vast bulk of the responsibility is with the user.
I had read that study. It is curious. Although am jealous of their ability to use 239,616 CMIP5 ensemble members... I have only used order of 100s ;-)
Yes. Just because I am concerned about one issue, doesn't mean such datasets are not useful for lots of other uses.
Just had a look at the pre-print, and I think they (like many others) don't appreciate that a dataset trained on climate models has limitations. e.g., can't use it in attribution or future warming constraints studies.
Do they say what the dataset could be used for, and what it couldn't be used for?
“A Dark Saint David’s Day"
In the early hours of 01 March 1927, an underground gas explosion at Marine Colliery, Cwm, killed 52 men. Rescue teams from Abertillery, Abercarn, Crumlin, Six Bells and more came to assist. The cause of the explosion was never found.
Ref: D6108
Great news from #Cornwall, where #geothermal energy and lithium are now being generated from deep, hot granite. ⚒️
www.bbc.co.uk/news/article...
In this week's Nature we argue there is a pressing need for a global assessment of avoidable climate-change risks. To understand the urgency of emissions reductions, policy makers and citizens need a full analysis of what is at stake including the worst-case scenarios. www.nature.com/articles/d41...
CHILDREN OF THE STONES (1977): ITV's landmark folk horror serial for children, and it's every bit as mature, dense, and scary as anything for adults. Happy day!
Front cover of first issue of 'Dragon user', the magazine for Dragon 32 - Wales built home computer.
Page describing lunar lander game, with the basic code the reader can type into the computer.
More code, with my alterations to enable it to run
Before vibe coding, github, or python ...
This was pretty much how I learned to programme, typing in code from a magazine and debugging the typos!
I still remember having to type this in each time to play the game, until we got a tape recorder to save it :-) #Dragon32
Front cover of first issue of 'Dragon user', the magazine for Dragon 32 - Wales built home computer.
Page describing lunar lander game, with the basic code the reader can type into the computer.
More code, with my alterations to enable it to run
Before vibe coding, github, or python ...
This was pretty much how I learned to programme, typing in code from a magazine and debugging the typos!
I still remember having to type this in each time to play the game, until we got a tape recorder to save it :-) #Dragon32
Remarkable that climate change isn't mentioned anywhere in this article on the fragility of the Dawlish stretch of the SW main line
BBC News - The saga of a £165m rail line that keeps causing travel chaos
www.bbc.co.uk/news/article...
A great photograph, of a landscape I used to know well. But it was the view from Waunlywd, near the Park Hotel, overlooking the Ebbw vale steel works. Cwmcarn is 10 miles down the valley.
For instance how many of the 30year trends in this plot are independent from each other (number of degrees of freedom)? It isn't even 2. It is very hard to interpret the variability of 30 year trends just based on 55 years of data.
Unconvinced by the Washington Posts' analysis here. One has to be very careful when comparing overlapping trends, of whatever length, and accounting for correlations/degrees of freedom.
www.washingtonpost.com/climate-envi...
"We need to be aware of the effect of preselection (“eyeballing” in our example) when evaluating the significance of certain results. "
"Because the study of the climate is empirical rather than experimental, we often find ourselves entertaining scientific hypotheses after we have already obtained and studied the data in some detail. "
Rediscovered this study about a pitfall of timeseries trend analysis.
"Visual determination of the starting time of the trend can lead to an incorrect conclusion that the trend is significant"
Relevant when looking at a global warming "hiatus" or "surge"
journals.ametsoc.org/view/journal...
Part of a series of negatives, some half plate glass negatives, some large format film negatives labelled ‘Time and motion studies’ which feature women with wires attached to their hands and arms which record the movements made. These two images are black and white half plate glass negatives ordered from the Kodak Research Laboratory, Harrow by WH Smith on 20th October 1959. British Library, Kodak A3038/20(1-2)
Colour print of a woman in a lab coat and goggles using glass laboratory apparatus. Photograph part of a series used for Kodak's internal newsletters. February 1975. British Library, Kodak A3146/27(8)
It's International Day of Women and Girls in Science.
Here are a few highlights from Kodak Research Laboratories in Harrow, documenting innovations led by women in as early as 1891.
#KodakHistory #Harrow #WomenInSTEM #Innovation #PhotographyHistory #HiddenHistories
Grayson Perry: Aspects of Myself is at RAMM until 3 May 2026.
📌 Find out more and join the thousands of people who have already visited at ramm.uk/Grayson-Perry
The observational dataset I put on the graph is HadCRUT5 from @colinmorice.bsky.social et al., agupubs.onlinelibrary.wiley.com/doi/full/10....
I wonder if they should reconsider renaming it "HardCRUT" as Durhasan et al do? ;-)
This suggests that the statistical model is "overfitting" to the observed temperatures, and that its skill in matching the past tells you nothing about future forecast.
Knowledge of human produced carbon dioxide and other gases is needed to make projections, with models of climate physics.
Purely statistical forecasts (no physics) of future global temperatures should be treated with a dose of salt.
This study from last year claims to have a statistical model that perfectly matches 1880-2022.
But if you add 2023, 2024, and 2025 the forecast fails.
link.springer.com/article/10.1...
Oh ffs, just do something else. Do ANYTHING else valuable. Work in a shop. Run a library. Become a personal trainer. Make pottery. Just anything else.
Disappointing to still see such solar-climate nonsense claims in scientific literature. The Maunder Minimum did NOT lead "to a drop in global average temperatures by between 1 and 2°C"
www.sciencedirect.com/science/arti...
I've updated my post on the funding crisis heading for British astronomy: docs.google.com/document/d/1.... UKRI made clear that astro and particle physics are being subject to unique cuts due to rising costs elsewhere. This looks like a deliberate decision not to fund fundamental physics. 🔭 🧪
