Climate Science Glossary

Term Lookup

Enter a term in the search box to find its definition.


Use the controls in the far right panel to increase or decrease the number of terms automatically displayed (or to completely turn that feature off).

Term Lookup


All IPCC definitions taken from Climate Change 2007: The Physical Science Basis. Working Group I Contribution to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, Annex I, Glossary, pp. 941-954. Cambridge University Press.

Home Arguments Software Resources Comments The Consensus Project Translations About Support

Twitter Facebook YouTube Mastodon MeWe

RSS Posts RSS Comments Email Subscribe

Climate's changed before
It's the sun
It's not bad
There is no consensus
It's cooling
Models are unreliable
Temp record is unreliable
Animals and plants can adapt
It hasn't warmed since 1998
Antarctica is gaining ice
View All Arguments...

New? Register here
Forgot your password?

Latest Posts


Does ocean cooling disprove global warming?

Posted on 18 May 2009 by John Cook

Last week's post on sea level rise yielded some interesting comments on ocean heat including some new papers I reference below (h/t to Chris). However, I confess my interest waned when the discussion turned to the relative merits of Energy and Environment. While the validity of editor reviewed literature versus peer review is an important issue, such topics yield no actual understanding on the latest developments with ocean heat content. So what does the science say?

The discussion revolves around a paper Cooling of the global ocean since 2003 (Loehles 2009) which looks at ocean heat content as measured by Argo. Argo is a network of over 3000 floats scattered across the globe that measure temperature and salinity of the upper ocean. Loehles finds a cooling trend from 2003 to 2008.

Figure 1: Heat content smoothed with 1-2-1 filter and overlaid with linear trend portion of best-fit model (slope = -0.35 x 1022 J/yr)

As oceans contain around 80% of the climate's total energy, ocean heat is a good measure of what's happening with our climate. So recent ocean cooling has led some to conclude that global warming has stopped. Probably the most articulate article is The Global Warming Hypothesis and Ocean Heat by William DiPuccio. His logic is as follows:

  1. The anthropogenic global warming hypothesis says ocean heat should increase fairly steadily and uninterrupted (monotonic), barring any volcanic eruptions
  2. The ocean has been cooling since 2003
  3. Therefore, the anthropogenic global warming hypothesis is either false or fundamentally inadequate

I might point out that ocean heat should rise whether global warming is natural or anthropogenic (they single out anthropogenic). But that's a minor nitpick. Let's examine points 1 and 2 in more detail.

1. Does ocean heat monotically increase from year to year?

If the climate is steadily accumulating heat, does this mean the ocean heat content will also show a monotonic steady trend? To answer this, we need to view observations of ocean heat content over the past 40 years.

Figure 2: global ocean heat from 1955 to 2008. Blue line is yearly ocean heat content for the 0–700 m layer (Levitus 2008). Red line is the global mean stratospheric optical depth, indicating the timing of major volcanic eruptions (NASA GISS, data ends in 1999).

The longer record reveals short term variability amidst the long term warming trend. Volcanic eruptions (indicated in red by the stratospheric optical depth) impose a short term cooling influence of several years. But there is also variability due to cyclic effects such as El Nino. It's not unusual for the warming trend to slow or even show cooling over several years. It's also worth noting that the Levitus reconstruction doesn't show cooling in recent years - instead a slight warming trend (albeit less than the long term trend). Which leads us to the next point.

2. Has the ocean been cooling since 2003?

Ocean heat is directly measured by buoys that sink through the ocean, measuring water temperature at different depths. The most comprehensive system is the Argo network which was gradually deployed from 2003 through to 2007, with 3388 floats now spread throughout the globe.

There have been early difficulties in measuring ocean heat. Expendable bathythermographs, or XBT's, measured ocean temperatures before the Argo network was deployed. XBT's have been found to introduce a warming bias so when the warmer XBT data was combined with the later Argo data, the most recent trend showed exagerated cooling (more on that here). In addition, some Argo floats have had pressure sensor issues which impose a further cooling bias.

Loehles 2009 uses a reconstruction of Argo data by Josh Willis (Willis 2008). Another analysis of the same raw Argo data was performed by Leuliette 2009 - a comparison of Willis 2008 and Leuliette 2009 can be found in Figure 3:

Figure 3: Monthly variations in global mean steric sea level computed by Willis 2008 (gray line) and Leuliette 2009 (black line).

Willis 2008 shows a cooling trend since 2004, while Leuliette shows a warming trend. The primary difference between the two is found early in the Argo record, when there were fewer Argo buoys deployed. Leuliette 2009 suggests the discrepancy between the two seems to be due to poor sampling and differences in how the data was handled. But which dataset is more accurate?

When confronted with two papers offering different results, a useful referree is an independently determined dataset. As well as using Argo data, Cazenave 2009 creates two independent estimates of ocean heat. Sea level rise is comprised of two components: mass change due to melting ice and steric sea level rise due to changes in ocean density. Thermal expansion is the main driver of steric changes (salinity is also a minor factor) so steric sea level rise is another measure of total ocean heat.

The first reconstruction uses satellite gravity measurements to calculate the change in ocean mass. They then subtract ocean mass sea level rise from total sea level rise to calculate the steric sea level rise. The second reconstruction uses satellite gravity measurements to calculate the change in mass of land ice and land water. The sea level rise from this contribution is subtracted from the total sea level rise to obtain another estimate of steric sea level rise. Both reconstructions show a statistically significant warming trend.

Argo offer two data streams - real time where the data is available almost instantaneously and delayed which undergoes more rigorous checks. Cazenave uses only measurements with the highest quality control settings (an approach the folk at Surfacestations would surely approve of). The Argo trend closely matches the other two reconstructions.

Figure 4: Three reconstructions of steric sea level, with seasonal element removed. Blue curve estimated from the difference between altimetry and GRACE-based ocean mass. Green curve estimated from the difference between satellite altimetry and total land ice plus land waters contribution. Red curve: ARGO-based estimate (Cazenave 2009).

In climate discussions, the most common error is focusing on a single piece of the puzzle while ignoring the big picture. The ocean cooling meme commits this error twofold. Firstly, it scrutinises 6 years worth of data while ignoring the last 40 years of ocean warming. Secondly, it hangs its hat on one particular reconstruction that shows cooling, while other results and independent analyses indicate slight warming.

The bottom line is there is still uncertainty over the reconstruction of ocean heat. Generally, the various reconstructions show the same long term trends but don't always agree when it comes to inter-decadal variability. The uncertainty means one cannot conclude with confidence that the ocean is cooling. Independent analysis seem to indicate that over last half dozen years, the ocean has shown less warming than the long term trend but nevertheless, a statistically significant warming trend.

UPDATE 20 May: John Cross (note that: Cross, not Cook) has emailed me an overview of the Argo system (PDF, 5Mb). It was published on Feb 2006 and answers many of the questions floating around in the comments - worth a read!

0 0

Printable Version  |  Link to this page


Prev  1  2  

Comments 51 to 55 out of 55:

  1. Posted in the wrong spot - Doh! Never mind, it is still relevant. Just on your Svensmark and Friis-Christensen comment - the paper is not peer reviewed (but you bring it up). I had a quick glance. You are intepreting the discussion in a radically different way from what they themselves say. 'In a recent paper (ref. [1]) Mike Lockwood and Claus Frohlich have argued that recent trends in solar climate forcing have been in the wrong direction to account for "the observed rapid rise in global mean temperatures". These authors accept that "there is considerable evidence for solar influence on Earth's pre-industrial climate and the Sun may well have been a factor in post-industrial climate change in the first half of the last century." But they argue that this historical link between the Sun and climate came to an end about 20 years ago. Here we rebut their argument comprehensively.' You may not agree but you have no right to make an assertion that is diametrically opposed to what is said in very clear terms and then attribute that to the authors.
    0 0
  2. On nett climate effects from clouds - it is indeterminate - but simply to discount the SW effect by assuming no nett effects is less than scientific. There is a beautiful new site at: It is worth checking the entire site. It provides nice pictures and graphs on all sorts of things - including low cloud and high clouds from the ISCCP. It shows tropical high level cloud increasing since 1999 and low level cloud decreasing. This should be a warming from IR trapping by clouds which offsets the increase in shortwave radiation at the surface. Nonetheless, it is obvious to blind Freddy that the planet is not warming all that quickly and that there are changes in the radiative budget caused by cloud changes that need to be accounted for. The position of the IPCC in regarding clouds as a climate feedback rather than a forcing is untenable.
    0 0
  3. re #51 It's the data that Svensmark and Friis-Christensen present, Rob, and not their interpretation. One should address the data in a paper before considering the reliability of what the authors say. They show very clearly in their Figure 2B their consideration of the CRF since 1958. There is virtually zero secular trend over this period (a slight cooling contribution, if anything). Are you suggesting they don't believe their own data presented in the figures they prepare and present? If the stongest advocates of the CRF-cloud-climate connection show there has been zero CRF contribution to any cloud-climate secular variation (outwith the solar cycle), then one can hardly try to sneak a CRF-cloud-climate connection for the very large late 20th century and contemporary warming in by the back door! The CRF data in the Jasper Kirby article is more or less the same. I don't think there is anything controversial about that. Incidentally, the Svensmark-Friis-Christensen is another of those examples of dismal science from a small sub set of the proponents of the CRF-climate "link". They totally misinterpret the Lockwood-Frolich paper, which nowhere stated nor inferred that the "historical link between the sun and the climate came to an end about 20 years ago". Lockwood and Frolich (proper solar scientists) showed that all of the measurable solar parameters were in the wrong direction for warming since around the mid-late 1980's. The point is that the CRF simply hasn't undergone any trend during the period in which it has been monitored in detail, that could have made any significant contribution to the very marked warming of the last 30-odd years (a secular trend of only a few % according to Jasper Kirby in the web paper you brought to our attention). This is a severe problem for those who assert a strong CRF-cloud-radiative forcing connection, since during the solar cycle the CRF can vary by up to 25% (see Svnsmark-Friis-Christensen Figure 2). Clearly if virtually zero long term change in the CRF can influence the cloud coverage in a way that you seem to be inferring (and Jason Kirby makes a bit of an inference in that direction too, without coming right out and stating this - see page 7/8 of his article), the CRF-cloud link must be exquisitely sensitive. Why then if we inspect the albedo-cloud-moonshine data in the Palle and Goode articles we have been discussing, do we see no cyclical variation of cloud/albedo/moonshine through the solar cycle with its very large changes in CRF? Even your blind Freddy (poor chap!) would consider that a rather problematic flaw in the dodgy CRF-cloud-climate notion.
    0 0
  4. Look - you can have a go at me but leave Freddy out of this. Svensmark-Friis-Christensen have obviously interpreted the data they present differently to you. But this was a paper that you originally introduced. Let's not worry about that - but concentrate on the peer reviewed work. To quote Jasper Kirkby – ‘The data for the period 1957–2001 show the solar cycle modulation and the effect of geomagnetic shielding, which leads to reduce fluxes and modulation amplitudes at lower geomagnetic latitudes (Fig. 7). Although the GCR reduction occurred mainly in the first half of the twentieth century (Fig. 6), the cosmic ray measurements shown in Fig. 7 suggest a continuing decreasing trend in the second half of the century, by a few per cent in the lower stratosphere and upper troposphere… On the other hand, there has been a substantial increase of solar magnetic activity since the Little Ice Age, and a corresponding reduction of the cosmic ray intensity. This suggests that the possibility of an indirect solar mechanism due to cosmic-ray forcing of the climate should be seriously considered.’ There is a cloud and cosmic ray correlation here – not sure how reliable it is. I don’t think it works like that – there is a lag between cosmic ray intensity and global cloud cover. The basis of the idea is in the record of cosmogenic isotopes preserved in ice cores. A well known correlation of isotopes of beryllium and carbon and global temperature reconstruction over more than a thousand years. Hence the link between heliospheric modulated cosmic rays and climate. The best correlation is with a 10 year lag. The most likely connection is through ionisation of aerosols and subsequent growth of cloud condensation nuclei. The core of the science is pretty solid and widely accepted as the dominant cause of climate change prior to 1975. We are only quibbling about a supposed divergence of trends post 1975. Neutron counts peaked in 1991. Usoskin calculated a cosmic ray modulation parameter which peaked strongly in 1991. Cloud cover peaked in 1998. Global surface temperature peaked in 1998. Ocean temperatures are falling or at least steady in the ARGO data. Usoskin et al - Heliospheric modulation of cosmic rays: Monthly reconstruction for 1951–2004 I have referenced CERN, the Hadley Centre, the Max Planck Institute and the Sodankyla Geophysical Observatory. There is nothing dubious about the science. The only question is by how much cosmic rays influenced climate in the late 20th century.
    0 0
  5. The ocean T is flattening/cooling because there is no energy imbalance, and solar heat lag effects are now subsiding. Human c02 is having very little effect on surface or ocean T, with the sun and its heat lag on the oceans now waning.
    0 0

Prev  1  2  

You need to be logged in to post a comment. Login via the left margin or if you're new, register here.

The Consensus Project Website


(free to republish)

© Copyright 2023 John Cook
Home | Translations | About Us | Privacy | Contact Us