Ken Rice
@kenrice.bsky.social
Interested in #scicomm, in particular about astronomy and climate change. Professor of Computational Astrophysics and Head of Institute for Astronomy, Univ. of Edinburgh. Views own, of course.
Indeed, which is why I think it's useful to highlight this pattern.
October 12, 2025 at 7:30 AM
Indeed, which is why I think it's useful to highlight this pattern.
Reposted by Ken Rice
I discussed the issues a couple of years ago @realclimate.org www.realclimate.org/index.php/ar...
and the latest barrage of op-eds adds nothing to their argument.
and the latest barrage of op-eds adds nothing to their argument.
RealClimate: Watching the detections
RealClimate: The detection and attribution of climate change are based on fundamentally different statistical frameworks and shouldn't be conflated.
www.realclimate.org
October 9, 2025 at 8:57 PM
I discussed the issues a couple of years ago @realclimate.org www.realclimate.org/index.php/ar...
and the latest barrage of op-eds adds nothing to their argument.
and the latest barrage of op-eds adds nothing to their argument.
I'm not quite sure what you're suggesting then. I don't think anyone argues that energy balance estimates aren't useful. It's mostly a suggestion that Nic Lewis has tended to be over-confident about his analyses and that - as many have suggested - these estimates have tended to be biased low.
September 20, 2025 at 4:03 PM
I'm not quite sure what you're suggesting then. I don't think anyone argues that energy balance estimates aren't useful. It's mostly a suggestion that Nic Lewis has tended to be over-confident about his analyses and that - as many have suggested - these estimates have tended to be biased low.
But it's not just uniform vs Jeffreys. There are other priors and there are other ways to estimate ECS. As Andrew points out, if your method suggest a good chance of the ECS being < 2K, then there's probably an issue with the method, or it's at least worth considering this.
September 20, 2025 at 2:02 PM
But it's not just uniform vs Jeffreys. There are other priors and there are other ways to estimate ECS. As Andrew points out, if your method suggest a good chance of the ECS being < 2K, then there's probably an issue with the method, or it's at least worth considering this.
Credit to @atomskssanakan.bsky.social for providing the data.
September 19, 2025 at 7:40 PM
Credit to @atomskssanakan.bsky.social for providing the data.