Three years ago, when the Rosenthal et al. (2013) paper was made available online from the journal Science, there was a bit of a stir among the purveyors of the humans-control-climate-with-their-CO2-emissions paradigm. Climate activist Michael Mann, for example, was ostensibly quite unhappy that a Medieval Warm Period (with +0.65°C warmer-than-now intermediate [0-700 m] ocean temperatures) was prominently identified in the paper’s abstract, and that both the Medieval Warm Period and Little Ice Age were characterized as “global events” in the summary. These conclusions served to undermine his own tree-ring hockey-stick reconstructions that effectively made the inconvenient Medieval Warm Period disappear. None of the graphs from the paper depicted anything resembling a “hockey stick” for recent decades, which is probably why Mann claimed the paper was “suspect” and contained “a number of inconsistencies” and “debatable assumptions and interpretations”.
Indeed, the Rosenthal et al. (2013) ocean heat and temperature reconstruction depicted the last few centuries as still among the coldest of the last 10,000 years, with 21st century temperatures (2000-2010) slightly (just ~0.3°C) above the coldest years of the Holocene (the Little Ice Age), and a full 2.0°C colder than during the Early and Middle Holocene (10,000 to 6,000 years ago), when CO2 concentrations were in the 260 ppm range, or 140 ppm lower than today.
Below is an encapsulating graph from the 2013 paper with annotations. Notice that the amplitude and rapidity of the natural (pre-1950) warming and cooling phases sometimes exceed + or – 1°C per century, and how the post-1950 anthropogenic (presumed) warming is effectively undetectable and/or indistinguishable from natural variability.
Not inclined to disappoint those advocating for the anthropogenic climate change paradigm, the Rosenthal et al. scientists obligingly granted one single appeasing sentence, offering that the“modern rate of Pacific OHC change is, however, the highest in the past 10,000 years.” Got that? The increase in ocean temperatures in recent decades (reported by the IPCC to only be “about 0.015°C per decade” for the 0-700 m layer between 1971-2010) is claimed to be much faster than any other period during the Holocene.
“It is virtually certain that the upper ocean (0 to 700 m) warmed from 1971 to 2010. … The warming rate is 0.11°C per decade in the upper 75 m, decreasing to about 0.015°C per decade by 700 m.” (IPCC AR5 WG1 Chapter 3, page 263)
So how did Rosenthal et al. leap to their conclusion? By oddly deciding that it is scientifically defensible to directly compare an 8,000-year-long trend to a 55-year trend (1955-2010) — which, statistically, is much like comparing an 8-month-long El Niño-dominated temperature anomaly to a 37-year-long temperature trend. In other words, Rosenthal and co-authors were able to claim an unprecedented modern warming rate by cherry-picking an extremely short-term period (1950s-present) and by simultaneously ignoring all the other short term (<100 years) trends (from their own reconstructions) that greatly exceeded modern temperature changes in both rapidity and magnitude (as the graph above illustrates).
Interestingly, Bova et al. (2016) recently found that during the Holocene there were short-term “natural” ocean warming trends that achieved amplitudes of more than “2.0°C within 200 years” at the 0-1000 m depth, which is about 5 to 10 times the rate of warming (more than 0.1°C per decade) compared to the 0.015°C per decade warming rate in recent decades (1971-2010) at the shallower 0-700 m depth.
Rosenthal and colleagues were nonetheless able to get away with claiming that modern ocean temperature changes are among the fastest of the Holocene in their 2013 paper, which is likely what allowed the paper to be published in Science. And thus a seemingly throwaway sentence necessarily became the “take-home conclusion” for Michael Mann and his ilk.
A New Rosenthal (2017) Paper
A new Rosenthal et al. (2017) paper has many similar features to the 2013 version with some important differences.
Seemingly in response to those who claimed their last paper didn’t represent the global oceans well enough, the authors added Atlantic Ocean proxy records to their last Pacific Ocean reconstruction that extended from the North Pacific to Antarctic. Again they concluded that their coverage represents global ocean temperatures:
“[W]e assume that our records represent the World Ocean and thus are comparable in volume with the current estimates (Levitus et al., 2012).”
Perhaps because of the lone positive coverage from Mann and other activists that it afforded them the last time they mentioned it, Rosenthal and co-authors once again claim in this paper that their cherry-picked short-term (55 years) warming rate (now 0.033°C per decade [0-700 m] for 1955-2010 instead of 0.015°C per decade for 1971-2010 as indicated by the IPCC) greatly exceeds the overall 8,000-year trend (from 10,000 years ago to 2,000 years ago) of “0.002°C per decade”. They also compare the same 55-year trend to a 400-year trend (1200 to 1600 A.D.) Just as directly comparing a -1.2°C temperature drop over an 8 month period to the overall 3+ decade satellite temperature trend would be an abominable misrepresentation, comparing a 55-year trend to a 400-year trend or an 8,000-year trend is, to put it bluntly, statistical malpractice.
“Levitus et al. (2012) report a mean ocean warming of the 0-700 m ocean layer of 0.18°C between 1955 and 2010, corresponding to ~0.033°C per decade. To obtain a first order comparison, we assume that our records represent the World Ocean and thus are comparable in volume with the current estimates (Levitus et al., 2012). Assuming the intermediate depth ocean (0-700 m) cooled between 10 and 2 Ka by ~1.5 °C we calculate a cooling rate 0.002°C per decade. Similarly, considering the intermediate depth ocean (0-700 m) cooled by ~0.5 °C between 1200 and 1600 CE we calculate a temperature change of -0.013 °C per decade. In both cases these rates are smaller than the modern rates even when applying the observed IWT changes to the whole ocean (as opposed to just the Pacific as was done in Rosenthal et al. (2013).”
‘Global’ Sea Surface Temperature Reconstructions
Let’s consider the sea surface temperature reconstructions. Notice the annotations indicating when humans presumably began influencing temperatures (after 1950) with their CO2 emissions. As shown, all or nearly all the warming in the last few centuries took place prior to the era of presumed anthropogenic influence. Instead of CO2 emissions, some other forcing mechanism needed to have caused the large magnitude and rapidity of ocean temperature changes.
Holocene Ocean Heat Reconstructions
The new (or reconditioned) graphs of ocean heat content ( 0-700 m) are no better, and perhaps worse, than the sea surface temperature reconstructions when it comes to depicting the negligibility of the anthropogenic influence.
The Alleged Anthropogenic Influence Effectively Begins Only After 1950
The reason why we only include the post-1950 period as the era of the presumed anthropogenic influence on ocean temperatures is simple: anthropogenic CO2 emissions were effectively flat and negligible prior to 1950 … after which they exploded.
In fact, the IPCC claims that the net radiative forcing from all combined anthropogenic sources was “more than 1.5 W m-2” between 1951-2011.
“Over the period 1951–2011 the trend in anthropogenic forcing is almost 0.3 W m–2 per decade and thus anthropogenic forcing over this period is more than 1.5 W m–2.” IPCC AR5 Chapter 8, page 699
As Rosenthal et al. (2017) note, the total net radiative forcing from all anthropogenic influences since 1750 has been 1.6 W m-2, which leaves almost no room for humans to have had more than a modest (~0.1 W m-2) impact on climate prior to 1950.
“It is generally assumed that the effects of direct forcing through the last millennium were relatively uniform in both hemispheres, much like the effects of the recent increase in GHG though with substantially smaller impact compared with the estimated total anthropogenic radiative forcing of 1.6 ± 0.8 W/m2.”
So if nearly all the radiative influence on ocean temperatures from human activity took place after 1950, and if nearly all the abrupt and far more pronounced ocean temperature changes (warmings and coolings) took place prior to 1950, this would imply that there are other factors driving warming and cooling trends besides anthropogenic forcing (CO2). Rosenthal et al. seem to reluctantly admit in the paper that the models of radiative forcing don’t seem fully capable of explaining the Holocene’s rapid temperature changes.
“Here we review proxy records of intermediate water temperatures from sediment cores and corals in the equatorial Pacific and northeastern Atlantic Oceans, spanning 10,000 years beyond the instrumental record. These records suggests that intermediate waters [0-700 m] were 1.5-2°C warmer during the Holocene Thermal Maximum than in the last century. Intermediate water masses cooled by 0.9°C from the Medieval Climate Anomaly to the Little Ice Age. These changes are significantly larger than the temperature anomalies documented in the instrumental record. The implied large perturbations in OHC and Earth’s energy budget are at odds with very small radiative forcing anomalies throughout the Holocene and Common Era. … The records suggest that dynamic processes provide an efficient mechanism to amplify small changes in insolation [surface solar radiation] into relatively large changes in OHC.”
However, they do not admit that the glaring lack of a post-1950 leap in temperature change depicted in their graphs seriously calls into question the dominance of anthropogenic forcing on ocean temperatures relative to natural variation and unforced net ocean heat changes.