I didn't say proxies are perfectly accurate, mhaze.

//Multiple independent proxies allow us to bound how fast temperatures and sea levels changed in the past.//

I said they allow us to bound rates of change. That's standard scientific practice. Uncertainty limits precision, it doesn't erase the ability to distinguish orders-of-magnitude differences. If your position is that proxies are too unreliable to constrain past rates in any meaningful way, that would invalidate most of paleoclimate science - including the paper you're citing.

//No, the graph does that.//

Graphs don't "say" things. Authors do. Figures are interpreted in context, and Clark et al. do not state that modern oceans are "at close to unprecedented lows". If you think a specific figure supports that claim, quote the authors' interpretation - not your own.

//No it isn't about that at all.//

Exactly - which is the point. You've been trying to use a deep-time equilibrium reconstruction to make claims about modern climate risk that the paper itself is not making.

//Rubbish. Now you're just making it up. Where does it say that?//

I'm not inventing anything.

The paper explicitly treats ocean temperatures over millions of years under near-equilibrium conditions and distinguishes that from transient behaviour under different forcing regimes. That slow-versus-rapid distinction is foundational to climate dynamics, not a rhetorical invention.

//Nowt to do with the oceans… signposts on the road to the unravelling.//

You're free to call them signposts, but they're political and cultural observations. They don't alter ocean heat content, radiative imbalance, or the physical constraints we were discussing.

If the argument is now that proxies can't constrain past rates, graphs speak without interpretation, and physics should give way to political mood shifts, then we're no longer talking about climate science at all.

Proxies are inherently fuzzy and can't give accurate readings to allow science to determine rates of change at the levels required in your assertions.

To know if the rate of change of this or that measure (eg temperature) in the past century is unusually high or low or whatever, you need to know what the rate of change in previous centuries was. But proxies can't give you that. Its what the rooly trooly scientists call data resolution and the resolution back into the past is not high. For example, in his paper (Marcott et al) Marcott confirms that the data resolution for the most recent past is at best 120yrs and beyond further back than 1500AD its around 300 years. Its impossible to say if current waring is unprecedented if you can't know what previous decadal rates were.

Same with this paper. The data is useful in extended time-frames but useless as regards small decadal periods.

"Graphs don't "say" things."
They do to people who can read them.

"Clark et al. do not state that modern oceans are "at close to unprecedented lows".
No they don't because that wasn't what their paper was about. But anyone looking at the data can see that we are close to unprecedented lows. I know you don't want it to be true, but that's what the data shows. Read the graph.

"I'm not inventing anything."

Well you invented the claim that the paper talks about "ocean behav[ing] very differently when changes happen slowly over millions of years than when it's forced rapidly, as it is today."
Stand by for JD's word salad in response.

"They don't alter ocean heat content, radiative imbalance, or the physical constraints we were discussing."

I was talking about signposts in the unravelling of the great climate scare. You don't want to see the signposts so tried to divert.

You're confusing resolution with constraint, mhaze.

//Proxies are inherently fuzzy and can't give accurate readings to allow science to determine rates of change…//

Lower-resolution proxies smooth short-term variability; they don't make rapid change invisible or unknowable. That bias works against detecting fast rates, not in favour of inventing them. If modern, instrumentally measured rates still exceed proxy-bounded past rates, that's a meaningful result.

//It's impossible to say if current warming is unprecedented if you can't know what previous decadal rates were.//

That's simply not true. You don't need decadal resolution everywhere to identify orders-of-magnitude differences in rate. Marcott explicitly notes resolution limits — and for that very reason does not claim to resolve short-term spikes. Those limits don't imply modern rates are unknowable; they imply past rapid changes would be smoothed out, not exaggerated.

//Anyone looking at the data can see that we are close to unprecedented lows.//

If that were the authors' conclusion, they would have said so. Graphs don't interpret themselves, and Clark et al. do not claim modern oceans are unusually cold. That interpretation is yours, not theirs.

//Where does it say that?//

The distinction between slow, near-equilibrium change and rapid, forced response isn't a slogan in the paper because it's basic climate dynamics. Clark et al. explicitly analyse long-term equilibrium behaviour under very different boundary conditions. Using that to dismiss modern transient forcing is the category error.

You're free to talk about "signposts", but that's a political narrative. It doesn't alter ocean heat content, energy imbalance, or the physical constraints we've been discussing.

Proxies are inherently fuzzy. Always have been. Whatismore, proxies are inherently regional. What is the proxy data for the middle of the Pacific Ocean 2000 ft down in 1000AD? there isn't any and therefore its inferred from other data ie inherently fuzzy.

As to Marcott, well you're making it up again. He specifically says he can't say that the recent warming is unprecedented because the data from previous periods isn't good enough to know what changes occurred over periods on 1 century. The proxy data smooths things so that they can 'know' what happened over a period of 300 years, but they can't know if that was a rapid change or a gradual change.

"The distinction between slow, near-equilibrium change and rapid, forced response isn't a slogan in the paper because it's basic climate dynamics"

Oh so when you said that "a central point of the paper is that the ocean behaves very differently when changes happen slowly over millions of years than when it's forced rapidly, as it is today" you really meant that it wasn't a central point, just something everyone knows.

Making up the claim and then making up further claims to defend the original made-up claims. Kudos.

"You're free to talk about "signposts", but that's a political narrative. It doesn't alter ocean heat content, energy imbalance, or the physical constraints we've been discussing."

You've been discussing. I've been discussing the signposts on the road to the unravelling of the great climate scare. Read the first post

"The study by Shaun Marcott and colleagues, published in Science, aimed to reconstruct global and regional temperatures over the past 11,300 years using 73 proxy records from diverse sources like marine sediments, ice cores, and lake deposits. This Holocene reconstruction revealed a gradual cooling trend after an early thermal maximum, but it also highlighted significant limitations in proxy data resolution, particularly for earlier periods, underscoring the challenges in discerning fine-scale climate variability.

Central to these caveats is the temporal resolution of the proxies. The median sampling resolution across the dataset was 120 years, with individual records ranging from 20 to 500 years. This coarseness inherently smooths out high-frequency signals, such as decadal or centennial fluctuations. Marcott et al. explicitly noted that their statistical averaging procedure eliminates variability at periods shorter than 300 years, making it impossible to resolve short-term events reliably in older epochs. For instance, processes like sediment mixing in marine cores and bioturbation further attenuate rapid changes, leading to a muted representation of past dynamics.

Dating uncertainties compound this issue. Chronological errors, often from radiocarbon calibration, increase with time depth, potentially shifting events by decades or centuries. The authors accounted for this via Monte Carlo simulations, estimating fundamental dating errors of 120–150 years, but acknowledged that such imprecision grows in pre-modern periods, blurring resolution further. Proxy calibration non-stationarity—where relationships between proxies and temperature vary over time due to confounding factors like precipitation or CO2—adds another layer of ambiguity.

Spatial coverage also declines backward in time, with fewer proxies available before 4,500 years ago, leading to regional biases and inflated uncertainties in global estimates. Critiques, such as those from Steve McIntyre, pointed to artifacts in the reconstruction, like an apparent 20th-century uptick not supported by raw data, stemming from redating and sparse modern proxies.

Ultimately, Marcott et al. demonstrates that while proxy data provide valuable long-term insights, their resolution in previous periods remains unclear due to inherent smoothing, dating errors, and incomplete coverage. This cautions against overinterpreting fine-scale warming rates in deep time, emphasizing the need for higher-resolution proxies and refined methods to clarify Holocene climate history."

mhaze,

This reply is essentially a restatement of the same argument you've made before: that proxy limits imply rates are unknowable and modern change may therefore be unremarkable.

We’ve been here before - twice - including with the Marcott summary you’ve just reposted, and the pattern hasn’t changed.

In earlier threads on Marcott, this followed the same arc: resolution limits -> therefore rates are unknowable -> therefore modern change might be normal -> therefore concern is hysteria.

http://forum.onlineopinion.com.au/thread.asp?discussion=10604&page=0
http://forum.onlineopinion.com.au/thread.asp?discussion=10434&page=0

When higher-resolution work was introduced to address exactly that gap (Neukom et al.), it was ignored, sidelined, or declared irrelevant, while Marcott continued to be cited for claims it explicitly did not make.

You're now doing the same thing again: leaning on low-resolution or deep-time reconstructions to argue that rates can't be constrained at all, while declining to engage with the literature that exists specifically to constrain them.

That directly answers the points you've just raised about proxy "fuzziness", resolution, and alleged unknowability. Those limits were already accounted for last time - and they didn't do the work you want them to do now, either.

If you want to argue that no proxy-based reconstruction can meaningfully bound rates of change, then that case needs to be made consistently - including for the papers you cite. Otherwise, we're just replaying an old disagreement under a new heading.