Glaciers May not Melt in Warmer Weather
The Science™, hard at-work, proposes that a warmer world might not lead to the extinction (sic) of glaciers by--pointing to new evidence from Svalbard at a time when the Arctic was 8-9 degrees hotter
In case you missed it in the flurry of commentary on current events™ in recent weeks, this little webzine of mine has a strong tradition of reading mostly convoluted academic prose (which is what I do in my offline job) and present you with ‘translations from the academese’ of papers and studies I found interesting.
Today’s posting does that, and I’m doing so because I just read a quite…interesting paper by a bunch of Arctic climate researchers that is, well, quite something: opening with statements like ‘observations needed to calibrate models are scarce’, researchers around Andreea Gabriela Auer et al. look at how a glacier located on Svalbard (the Åsgardfonna) did under ‘warmer-than-present Holocene Thermal Maximum conditions’: spoiler alert—the Åsgardfonna ‘survived and may have advanced despite warmer condition’, which is why I thought the paper merits attention.
It’s an English-language paper, hence the only additions to me quoting from it are emphases added and, of course, my [snark].
‘Hydroclimate intensification likely aided glacier survival on Svalbard in the Early Holocene’
A paper by Andreea Gabriela Auer et al., which appeared Communications Earth & Environment, vol. 6, art. no. 100 (2025); available in full via https://doi.org/10.1038/s43247-025-02064-z (published 11 Feb. 2025).
Abstract: ‘observations needed to calibrate models are scarce’
Accelerated Arctic warming and wetting has global impacts, as the region’s glaciers and ice caps respond to variations in temperature and precipitation, impacting global sea-level change. But as the observations needed to calibrate models are scarce, predictions cannot confirm if increases in snowfall can help offset melt. Here, we analyze two 14,000-year-long glacier-fed lake sediment records from the Svalbard archipelago to examine the response of a resilient ice cap (Åsgardfonna) to warmer-than-present Holocene Thermal Maximum conditions. End-Member Modelling allowed us to unmix the diluted grain size signal of rock flour – a widely used proxy for past glacier change, and surface runoff – an indicator of hydrological intensification. Our findings reveal that Åsgardfonna survived and may have advanced despite warmer conditions, possibly due to enhanced snowfall driven by sea-ice loss. This suggests that future increases in precipitation could moderate glacier retreat in similar settings.
Quite a bit of information that runs against the now-hyperbolic to the extreme climate catastrophists’ gloomy predictions of the imminent, well, end-of-the-world (as we know it).
Let’s turn to the main parts of the study, then.
Arctic climate change rates outpace the global average1, as observations reveal that the region warms and wets more than twice as fast as the global average2,3,4.…
The regions’ numerous glaciers and ice caps (GICs) are already dominant drivers of on-going sea-level rise8,9, and respond to changes in both temperature and precipitation10. Therefore, the counteracting impacts of warmer (melt) as well as wetter (accumulation) conditions on the future evolution of Arctic glaciers can have global societal consequences11.
One of the core aspects of critique™ in the humanities (my own field being History) is the simplistic question, ‘well, nice case study, but how does it relate to the grander scheme of things?’ Often when reading such papers, I think, well, a good point made here—but is it possible to replicate that study, esp.—as is very, very often the case with polar panic papers—there would be a prime comparative antipode, i.e., compare Arctic and Antarctic to each other.
While I wouldn’t call myself a climatologist™, I do read a lot and I’m a very fast reader in both English and German (600+ words/minute), I do wonder why the most obvious comparison—Arctic and/vs. Antarctic isn’t undertaken very often, which is a topic we’ll return to in a moment.
Here are some of the problems with such studies, though:
Predictions of Arctic glacier mass balance and the associated sea-level contributions remain uncertain, as projections of the amount and phase of future precipitation change remain poorly constrained, in part because of a lack of robust observations. Arctic meteorological station data are scarce, and snowfall is often mismeasured because of wind-driven undercatch12. As a result, simulations underestimate observed precipitation increases, while disagreeing on the seasonal distribution of future change4,6,13,14.
That sounds…like a teeny-weeny problem in terms of both epistemology and whatever climatologists™ may call ‘methodological considerations’. Here’s what the paper discussed today did:
Reconstructions of glacier-climate change from past warm periods can help answer how the uncertain interplay between warming and hydrological intensification may affect the future evolution of Arctic GICs. In this context, the Early Holocene (11.7 - 8.2 ka BP [this means ‘kilo-annum’ (1,000 years) BP (before-present), i.e., 11,700-8,200 years before 1950, thus covering the period from the end of the last ice age; ‘Before Present’ relates to 1950 as the standardised indicator due to the advent of C-14 radiocarbon dating)]) is highly relevant as there is ample evidence that conditions were warmer and wetter than today [huhum, we gotta talk about all these pesky ancestors of ours and their CO2 emissions, then, right?]15,16,17,18. Also, unlike geological archives from prior interglacials eroded by subsequent ice advances19, Early Holocene deposits have a high preservation potential. Critically, recent evidence suggests that some Arctic glaciers (temporarily [so, you’re suggesting that there were no glaciers left on earth for a period after 8,200 before present?]) survived these conditions20…
We present a lake sediment-based reconstruction that resolves glacier variations throughout the culmination of Early Holocene warming on Svalbard. This archipelago is exceptionally sensitive to drivers of amplified Arctic glacier-climate change as reflected by record-breaking warming27, the largest regional sea-ice loss28, and high projected precipitation changes6. To warrant a continuous and representative Early Holocene glacier signal, we target 1) two lakes – validating and upscaling site-specific findings using two ~14 ka long sediment records that are 2) both fed by Åsgardfonna - an ice cap that can survive warmer-than-present conditions based on observational, modelling, and isostatic evidence29,30,31.
There’s also talk about ‘human-relevant changes’, which are resolved (‘rigorously’, it is claimed) by a variety of approaches; I’ve omitted that half-paragraph here.
From the ‘Results and Discussion’ Section
Setting
Our study sites, lakes Berglivatnet and Lakssjøen (79°42’N, 15°53’E), are both located on the eastern shore of Wijdefjorden in northeast Spitsbergen, Svalbard (Fig. 1). Lake Berglivatnet, the main focus of this study, measures 1.2 km2, while our secondary site, Lake Lakssjøen, covers 3.6 km2…Situated at 46 and 74 meters above sea level (m a.s.l.), respectively, both basins lie above the postglacial marine limit in contrast to previously investigated lakes in the area33,34. An index point at Gråhuken places the local marine limit at 42 m a.s.l.35…
Meltwater from one of the largest ice caps on Spitsbergen, 1230 km2 Åsgardfonna37, drains into both lakes. Lake Berglivatnet is fed by the Berglibreen valley glacier, while Lake Lakssjøen receives input from an unnamed lobe of the NW sector of Åsgardfonna (Fig. 2a). In contrast to glacial lakes previously investigated [this is important as the two glacial lakes investigated here have no comparison either: in terms of methodology, this is a case study of case studies] to constrain the past behavior of the ice cap34, the glaciers that feed both lakes are classified as non-surge-type38, and thus more likely to respond primarily to climate fluctuations26. Satellite imagery reveals that Berglibreen has retreated ~0.23 km since 193639 [lol, satellite images from as far back in time as 1936? You’ve gotta be kidding me here…] - mere decades after the local culmination of the Little Ice Age (LIA)40. In contrast, the unnamed glacier that feeds Lake Lakssjøen has remained stationary since32 (Fig. 2) [ahm, come again: ‘the unnamed glacier….remained stationary since’ (1936)? How TF is that even possible, let alone a plausible proposition?]
Yet, while these anomalies are certainly…odd (as the selection of the two case studies appears to be not exactly too rigorous: they can’t be compared to each other in the same study, let alone to, say, other glaciers elsewhere), the true zinger, I would argue, is this segment:
Beyond instrumental observations, marine sediment core data suggest that the adjacent Wijdefjorden deglaciated prior to the Younger Dryas period41 (YD: 12.9 - 11.7 cal. ka BP - calibrated kiloannum before present), while high sedimentation rates suggest particularly rapid glacier loss between 12.1 and 9.9 cal. ka BP42. Complementing this evidence, lake sediment data from nearby Lake Femmilsjøen34 indicate the Åsgardfonna ice cap had retreated behind its present-day extent by 10.1 cal. ka BP [i.e., it was warmer back then, hence that makes sense]. Iron fluxes into Wijdefjorden reveal limited glacial activity during the subsequent Middle Holocene43, and basal ice-core ages suggest Åsgardfonna melted completely prior to 6 ka BP44[so, what’s driving that kind of warming 6,000+ years ago? I mean, if we were to believe the CO2 narrative hook, line, and sinker, pre-industrial CO2 levels were lower than they are now…], although it should be noted that this chronology has been challenged45. Therefore, evidence of the Early Holocene survival of Åsgardfonna remains ambiguous. However, glacier models and space-for-time substitution do suggest that the ice cap can survive Representative Concentration Pathway (RCP) 8.5 21st century warming of ~8 °C30,31
RCP 8.5 is typically referred to as ‘business-as-usual’ and leads straight to hell, so to speak. It’s the conventional talking point to advance the alarmist agenda, for which this TED Talk by James Hansen from 2011 or 2012 might serve as a pars pro toto.
I do wish to note that I twice had the (questionable) pleasure of chatting with Prof. Stefan Rahmstorf, Germany’s leading climatologist, on the margins of some academic event in Zurich in 2015 and 2017, respectively. I recall a lunch conversation (he picked some Pacific fish dish, enquiring—entirely un-ironically—if the fish was ‘caught sustainably’) with him where he told me, 8-10 years ago, that ‘RCP 8.5’ would be ‘entirely un-realistic’. When I asked as to why it was brought up, then, Rahmstorf told me that this is done to ensure the public’s continued agitation and attention. But I digress.
Back to the findings of the paper we discuss today:
CT imagery, our granulometric data, measures of organic content (LOI, Inc./Coh.), as well as minerogenic proxies (MS, Ti, Ca), all indicate highly variable depositional conditions…
The ~14-10 cal. ka BP period was marked by regular ice margin oscillations in our study area, although the comparatively poor chronological control on unit 4 does not allow us to ascertain on what timescales…
We cannot attribute the aforementioned ice margin oscillations between ~14-10 cal. ka BP to either (internal) ice dynamics or (external) climate forcing…[this is hardly surprising given the study set-up]…organic content peaks at 10% LOI values also observed during the (Early) Holocene, while the terrestrial origin of the plant macrofossil dated ~12.5 cal. ka BP in this deposit also suggests climatic amelioration (see Suppl. Table 1). These findings contribute to a decades-long debate about the behavior of the GICs of Svalbard during the Younger Dryas70, and support recent evidence from glacially-derived (Fe)-oxide fluxes that show this stadial was marked by rapid ice retreat in Wijdefjorden43. From a paleoclimate perspective, the presented evidence for land vegetation (terrestrial plant fossils) and high productivity (LOI) helps explain why thermophilous species were already present at the onset of the Holocene in nearby Ringhorndalen71. Beyond our study area, the above (growing season) data complement a growing body of evidence that suggests that the Younger Dryas stadial was characterized by mild summers and severe winters72,73, rather than year-round cooling.
Oh, so we note that there’s no one-size-fits-all climate of the past. Why is there in the models and assumptions underlying The Science™ today?
But wait, there’s ‘more’ in this paper:
Glacier survival during the Holocene Thermal Maximum [that just means ‘it was warmer back then’ by about ‘4.9 degrees Celsius’]
Based on our observation-validated interpretation that EM 1 captures the influx of glacier flour and therefore records changes in glacier variability (see our Late glacial- Holocene lake evolution section in Results & Discussion), Fig. 7 shows that Åsgardfonna likely survived 1) the culmination of the Holocene Thermal Maximum (HTM) on Svalbard ~9.5 ka BP18,46,79,80, and 2) the subsequent local Holocene glacier minimum ~7 cal. ka BP29,81 (Fig. 7b)…
Importantly, given a response time of at least 880 years for a GIC the size of Åsgardfonna82, the observed centennial-scale EM 1 minima which vary between 70 – 280 years, are too brief for complete melt (Fig. 7d), even under on-going and predicted rates of change30. In support of this evidence, the modelled response of Åsgardfonna to RCP 8.5 forcing – which predicts summer temperatures that are ~8 °C higher than today and on-par with HTM estimates on Svalbard18,31 – reveals that ice remains in the sub-glacial catchments of Berglibreen as well as Lakssjøen by 2100 CE (Fig. 2 and Suppl. Fig. 9).
Did you catch the sleight-of-hand? RCP 8.5 indicates an increase of mean global surface temperatures by 2100 by about 8.5 degrees above pre-industrial levels, i.e., means isn’t an abrupt increase from, say, today until tomorrow. The paper is mum about how that incremental, if not necessarily abrupt, temperature increase may play out over the next 75 years (for the record, the mean temperature increase projections are: 2.0 (1.4 to 2.6) for 2046-65 and 3.7 (2.6 to 4.8) for 2081-2100).
And that’s before we even consider the little, if omitted, factoid that RCP 8.5 derives from the IPCC’s Assessment Report 5 (AR5) from 2014, which has been, in many ways, superseded by the Assessment Report 6 (AR6) from 2021. The latter doesn’t have any so-called ‘Representative Concentration Pathways’ (RCPs)—i.e., modelled scenarios of future developments—that have been supplanted (‘considered together’) with what the IPCC labels ‘Shared Socioeconomic Pathways’ (SSPs), which are:
Back to the discussed paper and its possible implications:
The [Holocene Thermal Maximum (HTM)] period is of particular relevance by providing a glimpse into the future, as summer surface temperatures were up to 9 °C higher than today (Fig. 7i)18. As such pronounced warming greatly enhances melt rates, we here investigate whether an increase in accumulation – the other major climatic driver of glacier change10– helped offset mass loss. To do so, we first cross-correlate glacigenic EM 1 and our EM 3-based runoff indicators, to account for offsets due to the afore-mentioned multi-centennial response time of Åsgardfonna82. In contrast, hydroclimate proxies like EM 3 have a near-instant response to changes in precipitation and/or runoff [i.e., accumulation offsets melting]. As can be seen in Suppl. Fig. 10, we find a robust correlation (ρ = 0.72, p < 0.0001), with a mean lag of 1.6 ka for EM1. This offset corresponds to the expected response times for GICs the size of Berglibreen ( ~ 5 km2) to NW Åsgardfonna ( ~ 50 km2)82, as shown by the inset of Suppl. Fig. 10…
Records indicate that winter precipitation increased between 9.5 and 7 cal. ka BP, in tandem with more variable sea ice conditions (Fig. 7g)…we hypothesize that Early Holocene changes of Åsgardfonna were driven by increases in snowfall.
And now for the meaty part:
To further investigate the role of hydrological intensification on the inferred HTM survival of Berglibreen and Åsgardfonna, and place these findings in a modern context, we compare our results with simulations of the Equilibrium Line Altitude (ELA) or snowline. Unlike modelled glacier dimensions, ELAs respond directly to a change in climate forcing (no response time)88. Available RCP 8.5 scenario simulations for 2071-2100 CE suggest the snowline of Åsgardfonna will rise above the modern topography89, ensuing disappearance of the ice cap. This contrasts with the presented EM 1-derived proxy evidence, which suggests glacier survival under similar ( ~ 8 °C) levels of summer (melt season) warming. Therefore, our results leave the possibility open that recent (CMIP5 and 6) generations of climate models still under-estimate changes in the phase and magnitude of Arctic precipitation change…
Finally, we note that both Principal Component Analysis (PCA) and End-Member Modelling Analysis (EMMA) evidence suggests that around 0.8 cal. ka BP, Berglibreen experienced the most prominent period of growth observed during the late Holocene (see Fig. 7c, d). In doing so, our findings support a growing body of regional evidence that suggests the classical Little Ice age (LIA: 0.7-0.1 cal. ka BP [1250-1850 AD]) did not mark the culmination of Arctic North Atlantic Neoglaciation20 and references therein95,104.
Bottom Lines
That’s quite a read, eh?
We note the scarcity of ‘observations needed to calibrate models’ as well as the main findings—the suggestion that high temperature may not kill all glaciers before 2100—but I’d argue that the implications are way, way worse than the authors consider.
For starters, I’m unsure that the two glacial lakes/glaciers are a good fit (as they’ve not been examined in such detail before): what are we comparing them to?
And that simple question takes us down a veritable rabbit-hole: are there antipodal (Arctic-Antarctic) comparisons?
I’ve prompted Grok in the following way:
go through the available high-profile journals, like Nature and Science, and look for papers on Arctic or Antarctic climate change, widely understood, in particular related to glaciers and ice caps (GICs)
Go through all abstracts and carefully read them—and list all papers that undertake a comparison of Arctic and Antarctic developments
Here are the three (!!!) such papers Grok identified:
Stokes et al., ‘Warming of +1.5 °C is too high for polar ice sheets’, Communications Earth & Environment volume 6, Article number: 351 (2025)
Hanna et al., ‘Mass balance of the ice sheets and glaciers – Progress since AR5 and challenges’, Earth-Science Reviews
Volume 201, February 2020, 102976
Shokr et al., ‘Why Does Arctic Sea Ice Respond More Evidently than Antarctic Sea Ice to Climate Change?’, Ocean-Land-Atmosphere Research
29 Mar 2023 Vol 2, Article ID: 0006, DOI: 10.34133/olar.000 [that’s my favourite]
For lack of time, I haven’t read any of them, but I found it weird that the principal comparison of north and south poles isn’t undertaken more often, if at-all. That said, I remain very wary of the BS masquerading as AI™ insights, hence do exercise caution. (I only skimmed the paper by Shokr et al., and I honestly don’t know why Arctic sea ice appears to be more sensitive to climate change than its Antarctic peer, but I suppose merely asking that kind of question would get anyone ostracised from the Church of Climatology.)
Speaking of cults, do not miss out on those grifters who, while the Science™ is apparently unable to tells us how, why, and to what extent glaciers may (or may not) melt, and how much of them, in warmer conditions, they claim that paying higher taxes will make the weather less ‘angry’ 75 years into the future:
To my mind, the most telling sign of Climatology actually being a cult is the below-linked episode of two researchers who put out a ‘working paper’ (an update to a refereed and published study) on the website of Statistics Norway only to be buried by state and corporate media bullying; eventually, Statistics Norway resolved this issue by notifying the public that no-one who’s not an expert™ will get to publish with them:
