Nature Publishes Pseudoscientific Extreme Weather Attribution Study Which Blames 2023/24 Ocean Warming Spike On 'Climate Change'

Another day, another climate alarmist pseudoscientific scam, this time published in ‘prestigious’ journal Nature.

The HTDS (Hunga Tonga Derangement Syndrome) is strong in this one!

It comes from academics at the University of Berne, Switzerland, so not the usual suspects, but it’s still junk science - climate scam propaganda dressed up as real science.

In structure, the entire paper is basically just another extreme weather attribution study, using dodgy statistics and overheated climate models forced with unrealistic emissions scenarios. It comes to the dubious conclusion that, without man-made climate change, the 2023/24 extreme ocean warming - though still unlikely in today’s warmer world - would have been “impossible” absent the long term warming trend since 1850 - thus tacitly attributing the event to ‘climate change’. This is bog standard extreme weather pseudoscientific attribution procedure.

Here we construct observation-based synthetic time series to show that a jump in global sea surface temperatures that breaks the previous record by at least 0.25 °C is a 1-in-512-year event under the current long-term warming trend (1-in-205-year to 1-in-1,185-year event; 95% confidence interval). Without a global warming trend, such an event would have been practically impossible. Using 270 simulations from a wide range of fully coupled climate models, we show that these models successfully simulate such record-shattering jumps in global ocean surface temperatures, underpinning the models’ usefulness in understanding the characteristics, drivers and consequences of such events. These model simulations suggest that the record-shattering jump in surface ocean temperatures in 2023–2024 was an extreme event after which surface ocean temperatures are expected to revert to the expected long-term warming trend.

The authors mention a low planetary albedo as being put forward as a possible cause, but Hunga Tonga doesn’t even get a mention (I have argued numerous times previously that HTHH disrupted global circulation patterns and more recently that the marked decline in low level cloud cover during 2023/24 may be a consequence of that disturbance to global circulation).

Over summer 2023, the margin by which SSTs exceeded the previous record that occurred in 2015–2016 increased to 0.2–0.3 °C (Fig. 1a). Overall, the annually averaged SSTs from April 2023 to March 2024 based on NOAA Optimum Interpolation (OI) Sea Surface Temperature (SST) V2.1 (NOAA OISST V2.1)¹ were 0.25 °C larger than the previous record SSTs when averaged over the same months of the year. This global record-shattering jump in SSTs (‘record-shattering jump’ is here defined as a record-breaking jump in annual (April to March) and globally averaged SSTs that exceeds previous records by at least 0.25 °C, as observed in 2023–2024) coincides in time with record atmospheric surface temperatures in late 2023^3,4,5,6 and early 2024^7,8. Moreover, the record-shattering jump in SSTs is believed to be responsible for the global atmospheric record surface warming^5,6, although surface temperature extremes over land and ocean are not necessarily related^9,10. The record-shattering jump in globally averaged SSTs has been a subject of much attention in the scientific community^11,12 and the general public^13,14. It has, for example, recently been argued that part of the jump was caused by low albedo owing to reduced low-cloud cover¹⁵. Since mid-July 2024, globally averaged SSTs are no longer record-breaking but still remain warmer than in any year before the jump in 2023 (Fig. 1a).

It hasn’t just been ‘argued’ that record warmth in 2023/24 was primarily caused by a reduction in low level cloud cover, it has been definitely demonstrated, using actual CERES flux data, that an increase in short wave solar radiation explains most or all of the jump initial jump in temperature in 2023 (and virtually all global warming since the beginning of the century for that matter).

But here we have a team of scientists poo-pooing that ‘theory’ and claiming, with the use of dodgy statistics and climate models, that the warming was caused by climate change! Tell me again, who are the anti-science nutters in this game of climate change cat and mouse that has been ongoing between sceptics and the cheerleaders of establishment science for nigh on 30 years now?

Owing to the unprecedented margin in the record-breaking jump in SSTs in 2023–2024, this jump in SSTs came as a surprise for the public and the scientific community. This event has raised questions about the likelihood of such a jump and whether jumps of this size are simulated in climate models². The failure of state-of-the-art climate models to reproduce events such as the jump in SSTs in 2023–2024 would consequently question the ability of these models to assess future risks associated with anthropogenic climate change²⁹. However, if climate models are able to simulate such record-shattering jumps in globally averaged SSTs, they would deliver analogues to study the evolution of the ongoing event and to see whether temperatures decrease again or remain high. In addition, the climate models could be used to identify the drivers of such record-shattering jumps in SSTs, and their consequences for other parts of the climate system and the marine ecosystem.

Not enough real data so they used dodgy statistical methods to reconstruct their sea surface temperature time series instead:

Owing to the relatively short observational SST record, return periods of rare extreme events, such as the record-shattering jump in globally averaged SSTs observed in 2023–2024, cannot be directly inferred from that observational record. To quantify the likelihood of a record-shattering jump in globally averaged SSTs that exceeded the last record by at least 0.25 °C, as in 2023–2024, we constructed synthetic time series of 100 million years using an autoregressive model of order one (AR(1)) using observation-based estimates of the trend, autocorrelation and standard deviation of annual globally averaged SSTs with respective uncertainties (see Methods for a detailed description of how these values and their uncertainties are quantified).

And this is what the dodgy statistics revealed:

Based on these synthetic observation-based time series, the record-shattering global jump in 2023–2024 was a 1-in-512-year event (mean estimate) under the current long-term warming trend (1-in-205-year to 1-in-1,185-year event; 95% confidence interval based on uncertainties of the observation-based trend, standard deviation and autocorrelation estimates; Fig. 2 and Methods). This result is qualitatively insensitive to the choice of the autoregressive model, to the methods that are used to estimate the trend, the autocorrelation and the standard deviation of annual globally averaged SSTs, as well as to the observation-based SST dataset that is used for the analysis (Methods). Without underlying warming, a record-shattering jump as observed in 2023–2024 is practically impossible. We found indeed no record-shattering jumps in our synthetic time series without a long-term warming trend, irrespective of the variability or autocorrelation characteristics (Extended Data Fig. 2).

Well there you go. Are you not convinced? OK, maybe not, so to nail the attribution, the authors turn to the all singing-all dancing, wondrous, miraculous climate models (again, standard fare for an extreme weather attribution analysis) and they find that they too ‘prove’ that such extreme jumps in ocean temperatures are now rare, but ‘expected’ because of demonic man-made global warming - although they are only half as likely as the statistical analysis shows.

Having quantified an observation-based estimate of the return period of such record-shattering jumps in globally averaged SSTs, we show that such jumps—exceeding the previous record by at least 0.25 °C—were simulated 11 times across 270 simulations from 35 different state-of-the-art climate models (Methods) between 2000 and 2040. These four decades encompass the time when the record-shattering SST jump occurred in the real world (Fig. 3a and Extended Data Fig. 1). As the 270 climate model simulations here comprise a total of 11,070 years, the likelihood for a single year to experience a record-shattering global jump in SSTs as observed in 2023–2024 in the climate models is 0.1%, making the record-shattering SST event in 2023–2024 a 1-in-1,006-year event in climate models (1-in-563-year to 1-in-2,016-year event; 95% confidence interval using the Pearson–Clopper confidence interval for binomial experiments; Methods and Fig. 3b). Although the return period estimate based on climate models is approximately twice as large as the observation-based estimate of the return period, it lies within the confidence interval of the observation-based estimate and the confidence intervals of both estimates largely overlap. The difference in the return period might be owing to lower warming trends and higher autocorrelations in the models compared with the observation-based estimates (Methods). However, the observation-based estimate of the return period is also highly uncertain as it is difficult to estimate the real trend, variability and autocorrelation over a short time series that is strongly influenced by natural climate variability so that the return period of such events might well be 1,006 years as estimated from the models. Overall, the probability of a simulated record-breaking jump in SSTs with a certain magnitude decreases if that magnitude increases. For example, jumps that exceed 0.2 °C are 6–7 times more likely than jumps exceeding 0.25 °C, whereas jumps that exceed the previous record by a larger magnitude than 0.25 °C become less likely but are not impossible in climate model simulations (Fig. 3b). Under high-emissions scenarios, the probability of record-shattering events increases (Extended Data Fig. 3a), in line with higher rates of warming. However, under strong mitigation scenarios, the warming trend reduces in the future and there are no simulated record-shattering events simulated across the model ensemble (Extended Data Fig. 3b).

Once again. we see from the final sentence above that the authors are abusing “extremely unlikely” high end emissions scenarios to dishonestly advocate for carbon reduction policies. The authors conclude:

Based on long synthetic time series with temporal characteristics that match the available observations, we estimated that the observed record-shattering jump in globally averaged SSTs in 2023–2024 was a 1-in-a-512-year extreme event (1-in-205-year to 1-in-1,185-year event; 95% confidence interval) based on current warming rates. Such a jump would not have been possible without anthropogenic warming. We have further shown that climate models indeed simulate such global (60° S–60° N) annual record-shattering jumps in SSTs that exceed the previous records by at least 0.25 °C, like the global jump in SSTs that was observed in 2023–2024. Moreover, the estimated return period of these events in climate models (1-in-a-1,006-year events in models) is within the confidence interval of the observation-based estimate of the return period. Furthermore, in these models, the simulated SST anomalies drop below record levels between May and September in the year after the jump in SSTs had started, consistent with the time when observed globally averaged SSTs stopped being record-breaking (July 2024). On the basis of the simulated record-shattering jumps, we conclude that it is likely that SSTs will return to pre-jump levels before September 2025 [well that’s very convenient isn’t it?]. In the few simulations that do not simulate a return to pre-jump levels, SSTs revert to the expected warming trajectory over the following years. Thus, SSTs have not shifted to a higher or accelerated warming trajectory after a record-shattering SST jump in the models. The ability of climate models to simulate both the magnitude of the SST jump and the timing of the decline of positive SST anomalies enhances confidence in their use for future studies to understand the length, intensity and drivers of such extreme events, and to quantify their impact on regional weather systems and their potentially devastating consequences for terrestrial and marine ecosystems, and their services.

My conclusion: more pseudoscientific, hocus-pocus, extreme weather/climate scam bollocks.