Boreholed science: making giant leaps of logic in Antarctica

This press release is from the UNIVERSITY OF CALIFORNIA – BERKELEY

During last warming period, Antarctica heated up 2 to 3 times more than planet average

Amplification of warming at poles consistent with today’s climate change models

Following Earth’s last ice age, which peaked 20,000 years ago, the Antarctic warmed between two and three times the average temperature increase worldwide, according to a new study by a team of American geophysicists.

The disparity – Antarctica warmed about 11 degrees Celsius, nearly 20 degrees Fahrenheit, between about 20,000 and 10,000 years ago, while the average temperature worldwide rose only about 4 degrees Celsius, or 7 degrees Fahrenheit — highlights the fact that the poles, both the Arctic in the north and the Antarctic in the south, amplify the effects of a changing climate, whether it gets warmer or cooler.

The calculations are in line with estimates from most climate models, proving that these models do a good job of estimating past climatic conditions and, very likely, future conditions in an era of climate change and global warming.

“The result is not a surprise, but if you look at the global climate models that have been used to analyze what the planet looked like 20,000 years ago – the same models used to predict global warming in the future — they are doing, on average, a very good job reproducing how cold it was in Antarctica,” said first author Kurt Cuffey, a glaciologist at the University of California, Berkeley, and professor of geography and of earth and planetary sciences. “That is noteworthy and a confirmation that we know how the system works.”

These models currently predict that as a result of today’s global climate change, Antarctica will warm twice as much as the rest of the planet, though it won’t reach its peak for a couple of hundred years. While the most likely climate change scenario, given business-as-usual greenhouse gas emissions, is a global average increase of 3 degrees Celsius (5 degrees Fahrenheit) by 2100, the Antarctic is predicted to warm eventually by around 6 degrees Celsius (10 degrees Fahrenheit).

The new results, which are the first good calculation of Antarctica’s ice age temperature and the amount of warming since, do rule out a couple of climate models that do not include enough feedback to accurately reproduce the amplified temperature in the polar regions, Cuffey said.

Cuffey and his colleagues, including Gary Clow of the U.S. Geological Survey in Lakewood, Colorado, published their results online last week in the early edition of the Proceedings of the National Academy of Sciences.

Deglaciation in Antarctica

The analysis is based on the fact that as the world warmed following the coldest part of the last ice age 20,000 years ago, the ice deep inside the Antarctic glaciers warmed more slowly than Earth’s surface, just as a frozen turkey put into a hot oven will still be cold inside even after the surface has reached oven temperature. By measuring the remaining difference – the 20,000-year old ice deep in the West Antarctic ice sheet is about 1 degree Celsius cooler than the surface — the scientists were able to estimate the original temperature based on how fast pure ice warms up.

Clow measured twice, once in 2011 and again in 2014, the temperature in a 3.4-kilometer-deep (2-mile-deep) borehole from which the West Antarctic Sheet Divide ice core had been drilled during an eight-year project that ended in 2011. Ice at the bottom of the borehole was deposited about 70,000 years ago; ice about one-sixth of the way up about 50,000 years ago; and ice about one-third of the way to the surface 20,000 years ago.

Cuffey developed a technique to combine these temperature measurements, which are smoothed as a result of heat diffusion in the ice, with isotopic measurements of old ice to come up with an estimated temperature of 11.3 degrees, plus or minus 1.8 degrees Celsius, warming since the depths of the ice age.

Interestingly, the Antarctic temperature increased much more rapidly than did Arctic temperatures after the glacial maximum. By 15,000 years ago, Antarctica had warmed to about 75 percent of its temperature today. The Arctic took another 3,000-4,000 years to warm this much, primarily because of the fact that the Northern Hemisphere had huge ice sheets to buffer warming, and the fact that changes in ocean currents and Earth’s orbital configuration accelerated warming in the south.

Antarctica was also more sensitive to global carbon dioxide levels, Cuffey said, which increased as the global temperature increased because of changing ocean currents that caused upwelling of carbon-dioxide-rich waters from the depths of the ocean.

The situation today, with global warming driven primarily by human emissions of carbon dioxide from burning fossil fuels, is different from natural cycles, he said. The ability of the oceans to take up carbon dioxide cannot keep up with the rising levels of greenhouse gases in the atmosphere, which means carbon dioxide and global temperatures will continue to increase unless humans cut their carbon dioxide emissions.

###

Co-authors with Cuffey and Clow are Eric Steig, T.J. Fudge, Michelle Koutnik and Edwin Waddington of the University of Washington in Seattle, Christo Buizert of Oregon State University in Corvallis, Richard Alley of Pennsylvania State University in University Park and Jeffrey P. Severinghaus of the Scripps Institution of Oceanography at UC San Diego.

Anthony Watts commentary:

They didn’t include a link to the paper in the press release, so I fished it out:

Click to access 1609132113.full.pdf

I don’t believe much of the claims made in this paper, for several reasons.

1. Look at the co-authors, they have a history of bias towards climate alarm.

Steig tried to prove warming in Antarctica a few years ago by applying Mann’s flawed PCA techniques, and was firmly rebuked with a skeptic written paper that showed their entire technique was a train wreck. Alley is all over the map, and has trouble connecting cause and effect, like the time he compared Antarctica, Penguins, and drunk drivers. I don’t trust ether of these two, because they are blinded by ideology.

2. The leaps of logic in the press release are stunning.

For example:

The calculations are in line with estimates from most climate models, proving that these models do a good job of estimating past climatic conditions and, very likely, future conditions in an era of climate change and global warming.

That’s not proof, it’s a correlation. And as we know, correlation is not causation. 

3. I don’t believe boreholes have any significant resolution at the scale of time they are talking about, and that’s not just my opinion.

The definitive book on the subject, Borehole Climatology: a new method how to reconstruct climate By Louise Bodri, Vladimir Cermak says so:

The book says 1-2 kilometers, yet the authors of this paper say:

We measured temperatures in the 3.4-km-deep WDC borehole (Materials and Methods). The temperature profile reveals a direct thermal remnant of the deglacial transition and subsequent Holocene temperature changes (Fig. 1).

So, they are in uncharted territory. Is the 3.4km borehole used in this paper able to resolve the temperature back that far? Is it in the sweet spot of “favorable conditions” alluded to in the book? Do “favorable conditions” allow for the resolution claimed in the paper past 2 km? Can they really determine the temperature accurate from 20,000 years ago with this technique? I have my doubts. It looks like the flimsiest of science to me because there is really [no] way to check for certain, they try to using an isotopic record to calibrate it, but it too is a proxy.

Cuffey developed a technique to combine these temperature measurements, which are smoothed as a result of heat diffusion in the ice, with isotopic measurements of old ice to come up with an estimated temperature of 11.3 degrees, plus or minus 1.8 degrees Celsius, warming since the depths of the ice age.

Using one proxy to calibrate another is like having one two estimated auto repair bills checking each other for accuracy, but you still don’t know what the final bill is until the car is actually worked on, a diagnosis made, and repairs completed. Hands-on trumps estimates every time, and in this case, all we have are two  proxy estimates of temperature with uncertainty, but no actual temperature record.

Look at the blue line in the left panel (which they don’t clearly define in the caption). The temperature at 3400m (or 100m above the bed of the hole) shoots up. They claim this is a climate signal. But, what if they are simply getting closer to a geothermal heat source? They have no way of knowing. From LiveScience in 2014:

Antarctic heat

Researchers have long known that volcanoes lurk under the ice of West Antarctica. This is a seismically active region, where East and West Antarctica are rifting apart. In 2013, a team of scientists even found a new volcano beneath the West Antarctic Ice Sheet.

We covered it at WUWT:  New paper finds West Antarctic glacier likely melting from geothermal heat below

A paper published today in Earth and Planetary Science Letters finds evidence that one of the largest glaciers in West Antarctica, the Thwaites Glacier, is primarily melting from below due to geothermal heat flux from volcanoes located along the West Antarctic Volcanic Rift System, i.e. not due to man-made CO2.

The authors of this new West Antarctic Ice Sheet borehole paper don’t mention volcanoes or volcanic activity of the area at all in the paper as a possible bias, and they make no mention of geothermal heat in the same context. The only appearance of the word geothermal comes here:

Optimization. Singular value decomposition was used to find parameter values that minimized the squared mismatch of modeled and measured temperatures. Every such optimization involved free parameters related to surface temperature variations plus three additional free parameters: the modern mean surface temperature, the present ice thickness (known to be in the range 3,450–3,470 m), and the rate of basal melt. The latter accounts for the geothermal heat flux, which is not an independent parameter. The number of simultaneous free parameters in all optimizations (Eqs. 1 and 2) remains constant (six).

Six free parameters? No wonder their observed and modeled results fit so well! I’m reminded of this famous quote:

With four parameters I can fit an elephant, and with five I can make him wiggle his trunk.

Attributed to von Neumann by Enrico Fermi, as quoted by Freeman Dyson in “A meeting with Enrico Fermi” in Nature 427 (22 January 2004) p. 297

4. The part of the paper is a cause for serious concern:

Those paleoclimate studies using Antarctic data have likely underestimated the temperature change in response to the well-constrained change in CO2 and thus underestimated climate sensitivity. In particular, the analysis producing the lowest estimated sensitivity in the compilation (36, 37) (a bit larger than 2.2 ◦C) predicted an Antarctic LGM cooling smaller than even estimates from the East Antarctic isotopic proxies. Our reconstruction affirms that this is an underprediction. Antarctica accounts for a small fraction of global area and hence does not contribute much directly to climate sensitivity estimates, but the discrepancy may indicate too-weak feedbacks in some of the models used to assimilate proxy data.

To me, that looks more like editorializing than science. They are trying to lecture others on CO2 sensitivity thousands of years ago without having any accurate record of CO2 changes in Antarctica…because, there simply isn’t any. In their figure 3b (below) they present GHG and albedo forcings, but these are model outputs, not measurements.

To me, this paper looks like just another attempt by Steig and peers to find that missing Antarctic warming they got burned on a few years ago. I don’t consider it particularly compelling nor useful.