Media Reports of +40% Adjustment in Ocean Warming Were Greatly Exaggerated

From Dr Roy Spencer’s Blog

January 16th, 2019

Summary: The recently reported upward adjustment in the 1971-2010 Ocean Heat Content (OHC) increase compared to the last official estimate from the IPCC is actually 11%, not 40%. The 40% increase turns out to be relative to the average of various OHC estimates the IPCC addressed in their 2013 report, most of which were rejected. Curiously, the new estimate is almost identical to the average of 33 CMIP climate models, yet the models themselves range over a factor of 8 in their rates of ocean warming. Also curious is the warmth-enhancing nature of temperature adjustments over the years from surface thermometers, radiosondes, satellites, and now ocean heat content, with virtually all data adjustments leading to more warming rather than less.

I’ve been trying to make sense out of the recent Science paper by Cheng et al. entitled How Fast are the Oceans Warming? The news headlines I saw which jumped out at me (and several others who asked me about them) were:

World’s Oceans Warming 40% Faster than Previously Thought(,

The oceans are heating up 40% faster than scientists realized which means we should prepare for more disastrous flooding and storms (

For those who read the paper, let me warn you: The paper itself does not have enough information to figure out what the authors did, but the Supplementary Materials for the paper provide some of what is needed. I suspect this is due to editorial requirements by Science to make articles interesting without excessive fact mongering.

One of the conclusions of the paper is that Ocean Heat Content (OHC) has been rising more rapidly in the last couple decades than in previous decades, but this is not a new finding, and I will not discuss it further here.

Of more concern is the implication that this paper introduces some new OHC dataset that significantly increases our previous estimates of how much the oceans have been warming.

As far as I can tell, this is not the case.

Dazed and Confused

Most of the paper deals with just how much the global oceans from the surface to 2,000 m depth warmed during the period 1971-2010 (40 years) which was also a key period in the IPCC 5th Assessment Report (AR5).

And here’s where things get confusing, and I wasted hours figuring out how they got their numbers because the authors did not provide sufficient information.

Part of the confusion comes from the insistence of the climate community on reporting ocean warming in energy content units of zettajoules (a zettajoule is 1,000,000,000,000,000,000,000 Joules, which is a billion trillion Joules… also a sextillion Joules, but male authors fear calling it that), rather than in what is actually measured (degrees). This leads to confusion because almost nowhere is it ever stated what assumed area of ocean was used in the computation of OHC (which is proportional to both temperature change and the volume of seawater involved in that temperature change). I’ve looked in this paper and other papers (including Levitus), and only in the 2013 IPCC report (AR5) did I find the value 3.6 x 10^14 square meters given for ocean area. (Just because we know the area of the global oceans doesn’t mean that is what is monitored, or what was used in the computation of OHC).

Causing still further confusion is that Cheng et al. then (apparently) take the ocean area, and normalize it by the entire area of the Earth, scaling all of their computed heat fluxes by 0.7. I have no idea why, since their paper does not deal with the small increase in heat content of the land areas. This is just plain sloppy, because it complicates and adds uncertainty when others try to replicate their work.

It also raises the question of why energy content? We don’t do that for the atmosphere. Instead, we use what is measured — degrees. The only reason I can think of is that the ocean temperature changes involved are exceedingly tiny, either hundredths or thousandths of a degree C, depending upon what ocean layer is involved and over what time period. Such tiny changes would not generate the alarm that a billion-trillion Joules would (or the even scarier Hiroshima bomb-equivalents).

But I digress.

The Results

I think I finally figured out what Cheng et al. did (thanks mostly to finding the supporting data posted at Cheng’s website).

The “40%” headlines derive from this portion of the single figure in their paper, where I have added in red information which is either contained in the Supplementary Materials (3-letter dataset IDs from the authors’ names) or are my own annotations:

The five different estimates of 40-year average ocean heating rates from the AR5 report (gray bars) are around 40% below the newer estimates (blue bars), but the AR5 report did not actually use these five in their estimation — they ended up using only the highest of these (Domingues et al., 2008). As Cheng mentions, the pertiment section of the IPCC report is the “Observations: Oceans” section of Working Group 1, specifically Box 3.1 which contains the numerical facts one can factmonger with.

From the discussion in Box 3.1, one can compute that the AR5-estimated energy accumulation rate in the 0-2000 m ocean layer (NOT adjusted for total area of the Earth) during 1971-2010 corresponds to an energy flux of 0.50 Watts per sq. meter. This can then be compared to newer estimates computed from Cheng’s website data (which is stated to be the data used in the Science study) of 0.52 W/m2 (DOM), 0.51 W/m2 (ISH), and 0.555 W/m2 (CHG).

Significantly, even if we use the highest of these estimates (Cheng’s own dataset) we only get an 11% increase above what the IPCC claimed in 2013 — not 40%.

Agreement Between Models and Observations

Cheng’s website also contains the yearly 0-2000m OHC data from 33 CMIP5 models, from which I calculated the average warming rate, getting 0.549 W/m2 (again, not scaled by 0.7 to get a whole-Earth value). This is amazingly close to Cheng’s 0.555 W/m2 he gets from reanalysis of the deep-ocean temperature data.

This is pointed to as evidence that observations support the climate models which, in turn, are of course the basis for proposed energy policy changes and CO2 emissions reduction.

How good is that multi-model warming rate? Let me quote the Science article (again, these number are scaled by 0.7):

“The ensemble average of the models has a linear ocean warming trend of 0.39 +/- 0.07 W/m2 for the upper 2000 m from 1971-2010 compared with recent observations ranging from 0.36 to 0.39 W/m2.”

See that +/- 0.07 error bar on the model warming rate? That is not a confidence interval on the warming rate. It’s the estimated error in the fit of a regression line to the 33-model average warming trace during 1971-2010. It says nothing about how confident we are in the warming rate, or even the range of warming rates BETWEEN models.

And that variation between the models is where things REALLY get interesting. Here’s what those 33 models’ OHC warming profiles look like, relative to the beginning of the period (1971), which shows they range over a factor of 8X (from 0.11 W/m2 to 0.92 W/m2) for the period 1971-2010!


What do we make of a near-perfect level of agreement (between Cheng’s reanalysis of OHC warming from observational data, and the average of 33 climate models), when those models themselves disagree with each other by up to a factor of 8 (700%)?

That is a remarkable stroke of luck.

It’s Always Worse than We Thought

It is also remarkable how virtually every observational dataset — whether (1) surface temperature from thermometers, (2) deep-ocean temperature measurements, atmospheric temperature from (3) satellites, and from (4) radiosondes, when reanalyzed for the same period, always ends up with more (not less) warming? What are the chances of this? It’s like flipping a coin and almost always getting heads.

Again, a remarkable stroke of luck.

Posted in Blog Article | 8 Comments »

Chuck Todd Devotes an Hour to Attacking a Strawman

January 3rd, 2019

or, All Credentialed Journalists are Sex Abusers

Meet the Depressed host Chuck Todd, sans brain.

Chuck Todd, on a recent episode of Meet the Press, highlighted the issue of global warming and climate change. He unapologetically made it clear that he wasn’t interested in hearing from people on the opposing side of the scientific issue, stating:

“We’re not going to debate climate change, the existence of it. The Earth is getting hotter. And human activity is a major cause, period. We’re not going to give time to climate deniers. The science is settled, even if political opinion is not.”

This is what’s called a “strawman” argument, where you argue against something your opponent never even claimed.

I cannot think of a single credentialed, published skeptical climate scientist who doesn’t believe in the “existence” of climate change, or that “the Earth is getting hotter”, or even that human activity is likely a “major cause”. Pat Michaels, Richard Lindzen, Judith Curry, John Christy, and myself (to name a few) all believe these things. That journalists continue to characterize us as having extremist views shows just how far journalism has fallen as a (somewhat) respectable profession.

What if I claimed that all journalists are sex abusers? Of course, no reasonable person would believe that. Yet, I would wager that up to half of the U.S. population has been led to believe that climate change skeptics are “deniers” (as in, Holocaust deniers), about whom journalist Ellen Goodman said 12 years ago,

“Let’s just say that global warming deniers are now on a par with Holocaust deniers”

At least my hypothetical claim that “journalists are sex abusers” is statistically more accurate than journalists’ claims that we skeptical scientists “deny” this, that, and the other thing (for those allegations, see Mark Halperin, Matt Lauer, Tom Brokaw, Charlie Rose, Tavis Smiley, Michael Oreskes, and others).

The fact is that even if humans are, say, 60% responsible for the warming of the global ocean and atmosphere over the last 60 years (which would be consistent with both the UN IPCC’s and Todd’s phrasing), the lastest analyses (Lewis & Curry, 2018) of what this would mean leads to an eventual warming of only 1 deg. C from a doubling of atmospheric CO2 (we are currently about halfway to that doubling). That’s only 1/3 of what the IPCC claims is going to happen, and an even smaller fraction of what the ratings-boosting extremists who journalists like to trot out will claim.

A Nuance Chuck Todd is Ill-Prepared to Discuss

Journalists are notoriously under-informed on science issues. For example, let’s look at the claim that recent warming has been human-caused. It is easy to show that such attribution is more faith-based than science-based.

Between 2005 and 2017, the global network of thousands of Argo floats have measured an average temperature increase of the upper half of the ocean of 0.04 deg. C. That’s less than 0.004 C/year, an inconceiveably small number.

Significantly, it represents an imbalance in energy flows in and out of the climate system of only 1 part in 260. That’s less than 0.5%, and climate science does not know any of the NATURAL flows of energy to that level of accuracy. The tiny energy imbalance causing the warming is simply ASSUMED to be the fault of humans and not part of some natural cycle in the climate system. Climate models are adjusted in a rather ad hoc manner until their natural energy flows balance, then increasing CO2 from fossil fuels is used as the forcing (imposed energy imbalance) causing warming.

That’s circular reasoning. Or, some might say, garbage in, garbage out.

The belief in human-caused warming exceeding a level that what would be relatively benign, and maybe even beneficial, is just that — a belief. It is not based upon known, established, and quantified scientific principles. It is based upon the assumption that natural climate change does not exist.

So, journalists do a lot of talking about things of which they know nothing. As Scarecrow from the Wizard of Oz said in 1939,



38 thoughts on “Media Reports of +40% Adjustment in Ocean Warming Were Greatly Exaggerated

  1. The figure clearly shows the data problem — from 1993 onwards Agros data the temperature shows steep rise [models] and prior to that just slow growth. Do we need to give importance to such data???

    Dr. S. Jeevananda Reddy

    • Josh Willis made several upward adjustments to OHC since ARGO was implemented, much of it by throwing out “cold” data he didn’t like. I used to follow the discussion closely between him and Dr. Roger Pielke Sr.

      In fact, if one were to look back, virtually any statistically significant warming by any metric was a result of adjustment of the raw data.

      Since when have clouds been well understood, or even the sun? Someone explain how the observed upside down greenhouse effect agrees with AGW “theory”. I mean, how does reflected IR warm 300 ft down into the ocean depths? Oh that’s right, overturning of the ocean waves, the heat is driven downward to rapidly increase the heat content of the ocean. I’ll remember that the next time I stir my cup of coffee.

  2. Bravo Dr. Spencer. Personally after 30 years of following the doom of the planet since reading what was predicted back in ~1990 in a Popular Science magazine at the doctor’s office, I think any warming globally caused by humans, if it could actually be measured, would be indistinguishable from 0.

  3. Regarding Ellen Goodman: “…she married Boston Globe journalist Robert Levey in 1982. Her stepson, Gregory Levey, died by self-immolation in 1991 protesting the First Gulf War.” (Wikipedia)

    Wow, that is investing some serious negative values in your loved ones.

  4. Most journalist are “objective truth” abusers.
    They are prostituting their integrity, by regurgitating “shock and awe” climate porn, without bothering to give it a microsecond of logical thought.
    They have lost their way. The lack of support by the data, just makes them more convinced that rational analysis needs to be silenced, in favor of ramming an agenda through the resistance of those of us required to think for a living.

  5. “It also raises the question of why energy content? We don’t do that for the atmosphere. Instead, we use what is measured — degrees. The only reason I can think of…”
    There is a very simple reason. There is a 3D field of temperatures measured. There is a lot of temperature change near the surface, and very little way down. The only way to get a summary number is to do a volume integration of temperature. That is the heat content (after taking account of specific heat capacity, which also varies somewhat). The only sensible single temperature you could then quote would be to divide that by the heat capacity of the ocean. Or possibly the volume. But the heat content is the primary figure calculated.

    • For the record the volume of oceans calculation currently stands at 1.332 billion cubic kilometers.
      It is always less than 4degree C at any depth beyond 2500m above that it depends on a mixture of local and global factors.

    • The only way to get a summary number is to do a volume integration of temperature.

      No protest from Nick when anomalies that mean one thing at -40 and a completely different thing at +30 are averaged together. No protest from Nick when models that are completely and obviously wrong and should be thrown out are averaged with all the other models. Apparently you can do all kinds of averaging of things that shouldn’t be averaged and it is OK with Nick. But the one metric where an average would produce an absurdly small number and suddenly Nick supports changing metrics.

    • Nick, everything you just said applies to temperature as well as it does to OHC. A “summary number” OHC corresponds to an average temperature change. In the atmosphere, it also varies in 3D, yet we can still express in in a change in Joules or delta-temperature, for any layer or volume or area or time period we want. For a given volume of ocean (or atmosphere), the two are interchangeable, with only a heat capacity of water (or air) needed to convert one into the other,

      • You write: I cannot think of a single credentialed, published skeptical climate scientist who doesn’t believe in the “existence” of climate change, or that “the Earth is getting hotter”, or even that human activity is likely a “major cause”. Pat Michaels, Richard Lindzen, Judith Curry, John Christy, and myself (to name a few) all believe these things.

        Perhaps that’s true of climate change and the earth getting hotter (though warming slightly would be a better term, in my mind), but human activity being a “major cause”? Really? You don’t know “a single credentialed, published skeptical climate scientist” who believes otherwise. How about, say, solar physicist Willie Soon? Or doesn’t he count as a “climate scientist”?

        I mean no offense and, on the contrary, am a genuine admirer of your work. I’m just, well, skeptical about your claim.

        • Fair question, but you neglected my qualifier of “credentialed”. I was being purposely selective of those with PhDs in Meteorology or Atmospheric Science, or have done fundamental research sufficient to make them professors in those fields. The necessary training and experience includes atmospheric dynamics, clouds microphysics and dynamics, atmospheric radiative transfer, climate energy budget studies, etc. These are complex fields of study and are difficult to pick up on your own. Willie has provided interesting and useful input to the field, and I don’t mean to minimize that contribution.

          • Thanks, Roy, much appreciated. A little slippery, I must say, but in any case, and in my humble, non-scientist opinion, I’ve kept a “Global Warming” file for maybe a decade now, my initial doubts about the issue surfacing as I watched not only non-scientist but POLITICIAN Al Gore rise heavenward with the MannMade hockey stick in the presentation that will one day render him the laughingstock of a world that wonder why in the ever-lovin’ world humanity saw fit to demonize that which freed us from the “Malthusian trap” of preindustrial destitution and premature death. (Answer:

            And for the recored, I had the pleasure of meeting you at the America First Energy Conference last August, which was perhaps the most enjoyable and productive conference I’ve ever attended, as I made several important connections on my company’s behalf, leading, among other things, to a DOE RFP submission this past Tuesday on “Coal-Based Power Plants of the Future,” our subtitle being “A Zero Emissions Design for All Coal-Combustion Residuals: Solids, Liquids, and Gases” (including the NUTRIENT gas known as CO2).

            If interested in learning more (if fact, if anyone else reading this is), let me know.

          • Dee, u place way too much faith in humanity:
            “hockey stick in the presentation that will one day render him the laughingstock of a world that wonder ”

            The general population still Lauds the likes of Lincoln and LBJ.

            No thanks. They are not capable of perspective or critical, independent thought.

            They will drag those of us who can down with them

      • Roy,
        “For a given volume of ocean (or atmosphere), the two are interchangeable, with only a heat capacity of water (or air) needed to convert one into the other,”
        Yes, they are. But often one is more useful to know that the other. It’s like voltage and charge/current in electricity. Nature has given us both voltmeters and ammeters; sometimes you want one, sometimes he other. Surface temperature is useful, and is what people commonly notice. It is the potential that determines whether heat flows in or out of your body. And the global average anomaly is useful, because it is reasonably evenly distributed, so if the global anomaly rises, it is likely that your local temperature will be up too. The heat content of the air is not relevant here; you may be uncomfortable losing heat to the atmosphere, but you won’t change the air significantly.

        The average temperature of the ocean with depth does not have such a role. It does not determine the heat flow into any critical environment (unlike SST). But heat content of the ocean is useful for accounting. When the surface warms, heat will flow from the environment we care most about (surface) into the depths. OHC is a measure of that.

    • Can you tell me what the differences between the atmospheric layers and the differences between the ocean layers are that justify the use of zettajoules and degrees c Nick ? Because your description of the oceans layers and their behaviour can be equally applied to the atmosphere.

      I am afraid i go with Roy’s suspicion that the only way to get a scary number is to use zettajoules.

  6. Good analysis. And don’t forget, major bias and other corrections are constantly being made to the bathythermometers that are supposedly measuring the ocean temperature to thousanths of a degree C, for example for a summary.

    And that’s besides the issue that each bathythermometer is supposed to be measuring the temperature change of 188,167 km3 of ocean (equivalent to 38 Lake Michigans). Does anyone think one thermometer can accurately measure the temperature of even one Lake Michigan with an accuracy of 0.001°C?

  7. Dr. Spencer, I think you’ve gotten the gist of your analyis correct. But a number of points are still not clear, even in the supplementary materials. The paper itself refers to Resplandy 2018 ( although it’s not listed in the supplementary methods citations. A co-author told me they did not use the Resplandy calculations found in their paper itself (as they’ve been proved wrong, but rather they used revised calculations published in an obscure website without citing it (

  8. “It also raises the question of why energy content? We don’t do that for the atmosphere. Instead, we use what is measured — degrees.”

    I actually bought up that exact issue in a comment a few articles ago. They are trying to use temperature as a proxy for energy to create a forcing. You have automated weather stations that do temp and wind readings every 60 seconds so an area under the graph of both is the classical physics equivalent of energy at that location for the entire day. Sum them up for each second of AWS around the world and you have probably the best guess value of energy in the atmosphere.

    Nick did like he has above deflected the comment into some other side issue.

    • Very typical of Nick, I find. He usually dives for the deep cover of the most meaningless minutiae in avoidance of the obvious flaws in his “big picture”.

    • Temperature is a measure of internal energy, U. A measure of enthalpy is more meaningful if wanting a measure of energy.

  9. If you have a real-world item you are trying to measure and you have 8 analyses measuring it which vary so much that the highest is double the lowest (CHG vs SMT), it is pretty clear that the analysts, as a group, have absolutely no clue of what they are doing.

  10. The silly thing is these guys have convinced themselves that they are working from physical first principles in attributing all the warming to humans. In their minds human culpability is the same as gravity. This allows them to dismiss and cull contradictory evidence as wrong.

    A while back Mr. Mosher was kind enough to direct me to some BE work claiming that temperature change could be explained by CO2 and volcanism. They got a pretty good fit after fiddling with log bases and using TWO independent variables or “parameters”.

    What struck me was the sense whoever did this (not Mr. Mosher) felt these curve fitting parameters were constants of nature, like Planck’s or Boltzmann’s constants. It was like they were predicting the existance of an unknown planet by the disturbance of the orbit of a known one.

    Unfortunately, this lad was predicting the planet Vulcan.

  11. One day there really will be a wolf at the door.
    When these headline grabbing, sloppy, unprofessional, probably unethical papers come out… yawn… with so many of these paper confirmation bias runs rampant, after awhile who cares, it’s wrong. “it’s worse than we thought” syndrome.
    They take great pains to have a mental illustration of the hard working unassuming scientist, completely oblivious toiling away, and the look of worry that ‘ it’s worse than we thought’. How many years has AGW been using that picture gram. Or the look of worried scientist leaving the US for France when Trump was elected. Did we ever get a number on the scientists that left?
    Somewhere a google plex of monkeys are banging at keyboards. Somewhere in there is the complete works of AGW

  12. Why does this article contain 2 articles treating different topics in the Roy Spencer’s blog? Looks like a wrong copy-paste.

  13. I think I finally figured out what Cheng et al. did (thanks mostly to finding the supporting data posted at Cheng’s website).

    Is this what’s known as replicability in science?

    Just as well Cheng isn’t in the pharmaceutical business “I think Cyanide is what he meant by Paracetamol”.

    The data can’t be very clear if a Professor still has lingering doubts about another’s work, can it?

  14. I also thought it would be good to have a term to measure very small changes in danger levels.

    The Brazillion: a very close shave.

    • Except how can anyone say the models match the observations when the models vary by a factor of 8 for the metric in question (0-2000m OHC change, 1971-2010)?

  15. “World’s Oceans Warming 40% Faster than Previously Thought”
    “…we only get an 11% increase above what the IPCC claimed in 2013 — not 40%.”

    To be perfectly fair, this is not necessarily inconsistent if the IPCC claimed a rate of warming that was around 30% higher than they actually believed was happening.

  16. This is how scientists and journalists try to scare people with statistics. The actual amount of insolation absorbed by the surface is in the range of 160 to 170 w/m2. What they are talking about here is a tiny change of about 0.1 w/m2. Which is actually a tiny change of less than 1/10th of a percent. And well below what they think that actual range of energy absorbed. Which varies by about 10 w/m2.

  17. “why energy content”

    Well, you answered your own question. Its obvious that they haven’t been able to scare enough people (most of whom revel in their innumeracy) with mere tenths and hundredths of degrees over decades.

    As Nigel Tufnel would have it, these go to zetta.

    Of course, they will now run into trying to explain to millennial what Hiroshima means. No doubt most of them think its Japanese animation.

  18. “For example, let’s look at the claim that recent warming has been human-caused…”

    Everyone seems hung up on the issue of whether or not warming trends are human-caused. I suppose that’s because people have a “no trash left behind at the campsite” mentality, and the idea that we can interrupt a natural global climate cycle is too obviously idiotic to support efforts to raise taxes. But I will say it as many times the topic comes up:


    Sorry for yelling there.

    IF warming is potentially catastrophic, AND we can do something about it, SUCH THAT the actions we take have an expected benefit that exceeds the expected costs, we should take action. The source of the warming is nowhere in the calculus. If we can’t do anything appreciable about it, then we’re along for the ride whether or not we’re at fault, or if it’s due to natural phenomenon. If we are at fault, and can do something about it, but the expected benefits of a proposed action are outweighed by the expected costs, then we should not take that proposed action. If we are not at fault, but can take beneficial action against the natural climate cycle, then why wouldn’t we do so?

    The important questions are:
    Is the warming trend a net negative?
    Can we do anything about it?
    For any action, do the expected benefits outweigh the expected costs?

    Pose a hypothetical question asking if we know enough about the climate mechanisms to actively intervene in the global climate cycle IF we hypothetically knew also that the warming trend was natural. Do we know enough about the global climate mechanisms that one would feel comfortable supporting active intervention in the natural cycles? THAT will tell you the extent to which people actually believe the science is settled.

  19. Roy are the changes significant compared to those over the Holocene and Pleistocene ? Over the last 12,000 years there have been the drying up of N Africa 5900 years ago, The Minoan, Roman and Medieval warm periods. The Medieval period increased food production and helped to bring Europe out of the Dark Ages.There have been many collapses of civilisations associated with barbarian invasions- 1200 BC, 410 AD and 1200 AD Mongols. I would suggest that cooling, producing a reduction offood in settled areas and reduced growth of grass causing nomadic people to move to warmer climates, are far larger problems.

    We may be overdue for another ice age. Any reduction in temperature and precipitation in the grasslands of N America, the Steppes,Argentina, China, Australia and New Zealand would greatly reduce food production.

    The temperature in England rose at rate of about 3.4C /hundred years from about 1690-1740 it helped to produce the Agricutural Revolutions.

    Are there any accurate measurements of heat and CO2 output from undergound spreading plates and volcanos?
    1. What changes in climate are detrimental to humans?
    2. What are all the natural variations be it from the Sun, movement of the Earth and outpouring from interior?
    3. What is the influence of humans? The Athenians cut down trees in about 600 BC causing Athens to warm and the arabs destroyed irrigation ( IBn Khaldun ) but these were local effects.

Comments are closed.