Skip to content

It is indisputable that the world has warmed in the past 150 years but how that warming relates to man’s activities is debatable given the irregularity of that warming and the influence of natural cycles. The post 2000 global temperature standstill is particularly problematic. It is said by some that a more robust demonstration of global warming will come from the oceans with increasing heat content and sea- levels rise (though see a previous post for a discussion on sea-level changes.)

It is difficult to measure how the heat content of the ocean changes. The problems are everywhere; only in the past few years have we had anything approaching adequate data and coverage, different systems measure different things in different ways, how to select data, how to process it and indentify sources of errors. In reality the scientists concerned select ‘valid’ data and then decide what cyclical periods should be removed so as to smooth out annual variations, how to map data taking into account under and oversampled regions of the ocean and how to harmonise data from different observing systems. It is not surprising then that different teams working on ocean heat changes have produced very different results.

Lyman et al seek to overcome these difficulties by looking at a range of results about the ocean heat content since about 1994. They find, once individual absolute results are removed, that the general shape of the post 1994 ocean heat curves are roughly similar with no change up to 2001, then an increase for two years, followed by another stable period until 2009. See fig 1 and note that the errors are large. Click on image to enlarge.


Lyman et al suggest that from 1997 to 1998 – the time of a strong El Nino – that some of the curves show cooling. In fact, only one of them does and that is a very marginal effect given the error bars and the large interannual variability. Lyman also suggests some curves show warming over this period, which is also difficult to justify looking at the data. Overall it is difficult to make any significant conclusions about differences among these curves.

Curiously, Lyman et al point out that ocean temperatures have been statistically constant since about 2000 even though one of the most important features in fig 1 is the 2001 – 2003 rise. They note that during this standstill sea levels have continued to rise which they attribute to melting ice however they comment that it takes less energy to melt ice than to warm the ocean for an equivalent rise in sea level. No great conclusions there.

It is interesting that the flattening of the ocean heat content curve occurs around 2004 which was the time that the Argo array of ocean sensors became the main source of ocean heat data. This is a situation reminiscent to the introduction of satellite-based sea level measurements introduced in the early 1990’s that also resulted in a (as yet unexplained) discontinuity of the data with the previously used tidal gauges. The Argos data is universally regarded as heralding a revolution in monitoring the state of the oceans, but the possibility remains that this abrupt change is due to an as yet unrecognized problem causing a bias in the observing system.

Overall, Lyman et al look at the various data sets available post 1994 which are from a variety of platforms, instruments, and analysis techniques and broadly conclude that they show the same thing which they describe as a decadal warming with a flattening after 2003, although the pre-2003 errors are large. Fitting a least squares line to the data they obtain an oceanic warming rate of 0.63 +/- 0.28 W/m2 for 1993-2003. The fit to the entire 16-year record yields what they term as a “robust” warming rate of 0.64 +/- 0.11 W/m2. Click on image to enlarge.


Kevin Trenberth discusses this result in a Nature commentary. His diagram of the findings, fig, 3 shows the average warming trend (red line). To my mind, although the data shows that the ocean heat content is greater now than it was in 1994 it is too much to conclude that there is a linear warming trend or indeed conclude that the warming trend is likely to continue in the future. In fact the graph show, and even this is speculation, that the ocean heat content has jumped from one stable period to another. Click on image to enlarge.



This might prove to be the start of a phenomenon seen often in scientific literature, especially in climate change literature. An interesting result with substantial caveats and gross uncertainties is presented. It is then described and repeated in stronger terms with the caveats played down. It thus enters the debate simplified, amplified, and out of proportion to its scientific validity becoming a sort of myth.

Ocean measurements are in their infancy and sixteen years is too short a time period to draw any conclusions about the nature of the changes seen. They could be cyclic and unrelated to any greenhouse gas accumulation. It has been said by some commentators that although the global average temperature has not increased for at least 10 years there is more energy going into the Earth system as a result of the greenhouse effect and that the oceans are storing this energy. This might be the case but given the data we have at the moment does not make that a firm conclusion. This will be the subject of further posts in the near future.