I’ve moved to WordPress: http://bobtisdale.wordpress.com/

Sunday, March 22, 2009

The Latest Revisions to Ocean Heat Content Data

I’ve moved to WordPress.  This post can now be found at The Latest Revisions to Ocean Heat Content Data
The November 5, 2008, NASA Earth Observatory article “Correcting Ocean Cooling”… http://earthobservatory.nasa.gov/Features/OceanCooling/…contained the graph of the newly revised Ocean Heat Content on page 4 of the article. It compares the original OHC data to the newly revised data. Figure 1 is that graph.

Figure 1

Using the coordinate capabilities of MS Paint, I “duplicated” the graph of the revised data in Figure 2 so that I could run a few quick comparisons in my post “Revised Ocean Heat Content.”

In the Recent Ocean Heat and MLO CO2 Trends thread at WattsUpWithThat, blogger DJ provided a link to the NOAA National Oceanic Data Center’s upcoming Levitus et al paper on Ocean Heat Content to be published in Geophysical Research Letters.
Figure 3 is the OHC graph from that webpage. Note the Annual data since 2004.5. What happened to the significant drop in 2007 in Figure 1? Have trouble seeing the difference between Figures 1 and 3?
Figure 3

Following the “basin time series” links will bring you to the NODC’s listing of “WORLD” Yearly Ocean Heat Content from 1955 to 2007.

Figure 4 is a comparative graph of the data I created from the Earth Observatory article and the data used in the Levitus et al paper. The two datasets track well until 2007.5.
Figure 4


As illustrated in Figure 4, the depiction of Ocean Heat Content varies from month to month even from the same data supplier. But there are other recent papers that illustrate Ocean Heat Content. These are illustrated in the manuscript of the Levitus et al (2009) paper “Global Ocean Heat Content 1955-2008 in light of recently revealed instrumentation problems”. Refer to Figure 5, which is Figure S9 in the Levitus et al paper. [Note that the Levitus et al data (red curve) includes the 2008 data in this graph.] Levitus et al, Ishii and Kimoto, and Dominguez et al were all published within a year of one another. All three papers illustrate the same variable, but the data varies significantly between the three datasets. Note the divergence of the Levitus et al data (red curve) in 2003.
Figure 5

A decade from now, when researchers sort out the problems of measuring Ocean Heat Content, when they agree on the methodologies to be used to calculate it, it may serve as a worthwhile measure of climate change. At present it does not.

UPDATE – May 7, 2009

Let me clarify my closing comment about the OHC reconstructions.

Given: El Nino events redistribute heat from the tropical Pacific to the high latitudes so that it can be radiated into space more readily. Let’s say I wanted to analyze the 1997/98 El Nino to the determine how much of that heat was released to the atmosphere and how much was simply redistributed to the extratropical North and South Pacific and to other ocean subsets. Refer to the following graph. It’s the comparative graph of Levitus et al, Ishii and Kimoto, and Domingues et al OHC datasets, Figure S9 in the Levitus et al paper. I’ve highlighted 1997 and 1998. In 1997, the OHC in all three datasets increased, and in 1998, they all decreased. BUT look at the differences in the magnitudes of the changes in 1998. Which dataset depicts the changes correctly? Right now, I don’t have enough confidence in any of them to do the study I’ve suggested.



tallbloke said...

Hi Bob, I found the following exchange on sci.oceanography a couple of years ago and have a question about it.

1] Warming of the World Ocean. Levitus et al, Science vol 287 2000
2] Anthropogenic warming of Earth's Climate System. Levitus et al,
Science v 292 2001
"Has anyone else here used the data presented in these papers? A
colleague and I have, but we cannot reproduce the net heat gain of
18.2 x 10^22 J in the worlds' oceans for the period 1955-1996 which
was mentioned in [2].

According to [2], this number comes from a straight line fit to the
5-year averaged ocean data from 1957.5 to 1994.5 (the year index
refers to the mid-point of the 5 year averages), extrapolated out to
cover the original 41 years 1955-1996. Ie a trend of 0.44 x 10^22 J
per year. The data are presented in Fig 4 of [1], and available from
the authors.

We get a much lower answer of 13.5 x 10^22 J, ie 0.33 x 10^22 J per
year. It's only a least squares fit, so I don't see what we could have
done wrong. But our number is a long way off the published value, and
also a long way short of the model result (which was 19.7 x 10^22 J).


> Have you contacted the authors?

Yes, I got the data from one of them in the first place, and he
explained how they had calculated the figure (the description in the
paper isn't brilliant). But as soon as I pointed out the error, he
stopped replying."

The Levitus graph you show has a difference in ocean heat content of around 6x10^22J between 1955 and 1996.

This seems to be a drastic reduction from his 2000 paper quoted above. What is the explanation for this?

Bob Tisdale said...

Thanks, tallbloke. I found the original exchange here.

James Annan appears to be a member of the Global Warming Research Program at Frontier Research Centre for Global Change, Japan.

You asked, “This seems to be a drastic reduction from his 2000 paper quoted above. What is the explanation for this?”

I believe it’s attributed to bias corrections for the XBT data. Refer to the discussion starting in the last paragraph on page 1 of Levitus et al (2009):

But note how they claim the trends are similar but use a starting date that doesn’t catch the entire term. But they explain that away as, “The starting year is chosen because data coverage improved after the mid 1960s when XBT measurements of the upper ocean began.”

If I haven’t done it already, I’ll have to do a trend comparison for the entire period that the two datasets overlap.


Bob Webster said...

In the last chart, is the difference between the high and low estimates ("Present paper" and "Ishii & Kimoto", respectively) for ocean surface heat content in 2007 (~5 x 10^22 J) sufficient to produce a detectible difference in sea level (within the limits of the best sea level measurement devices)?

Since the analyses producing the high and low estimates in 2007 appear to yield identical heat content estimates for 2002, if the heat content differences in 2007 are sufficient to produce a measurable difference in sea level, why not just compare the sea level change over the five years from 2002 to 2007 to provide an indication which estimate is likely the more accurate?

It is understood that such a comparison would not necessarily prove either was absolutely accurate.

Bob Tisdale said...

Bob Webster: I haven't yet attempted to isolate the sea level rise due to thermal expansion. It seems like it would be a lot of work to remove the other sources of sea level variations.

tallbloke said...

Hi Bob, thanks for that, seems like a heck of a warm bias. Linking through to Bob's question, I wonder if someone's sky high sea level rise influenced the xbt data interpretation back before the latest satllite telemetry came online.

As it happens I decided to use the telemetry data to back calculate the average depth of the thermocline base required to get a thermal expansion consitent with IPCC estimates. They say half the 1993-2003 sea level rise was thermal expansion, around 16mm and this equates to a thermocline depth of an average 1000m, which vindicates Spencer vs Raymond Pierrehumbert.

I'm wondering that since the thermocline in equatorial and tropical waters is much shallower, whether a lot of the heat has been stored in the exratropical north atlantic, as this might explain the high anomaly?

Bob Tisdale said...

tallbloke: It's a tough question to answer since the OHC data in recent papers is only being presented since the mid-1950s. As North Atlantic SSTs decline over the next 25-30 years, the AMOC flow rate should also increase. Will the corresponding drop in the North Atlantic OHC bring it back into line with the others?

tallbloke said...

Bob, I think you meant the AMOC flow rate should decrease as SST's fall? Anyway, interesting question. I think it will, because the coming downturn in temperatures will affect the north Atlantic more than the tropical and equatorial Indian and Pacific oceans, and the heat transfer rates of warm waters from those oceans to the extra-tropical regions will diminish too.

Bob Tisdale said...

tallbloke: Nope. I meant AMOC flow would increase as North Atlantic SST anomalies decrease. Refer to my post “Atlantic Meridional Overturning Circulation Data”

As SST’s rose, AMOC flow rate appears to have decreased from the mid-70s to 2005, reaching its lowest level in 1999, as an apparent result of 1997/98 El Nino.

tallbloke said...

Ok, now I'm confused. The reason I thought you'd made a typo was due to this response you gave to carl Wolk last November:

Bob Tisdale said...

Carl: On decadal scales, an increase in North Atlantic SST should result in an increase in AMOC flow rate, and vice versa. Refer to:

Is this a mix up that has arisen from the inversion of the AMOC graph you used to find the correlation with ENSO?

I'm not calling you out here, just wondering which way round it goes.

Bob Tisdale said...

tallbloke: Figure 3 in the “3.2 Dynamics and Predictability of North Atlantic / European Climate Variability” clearly shows that North Atlantic SST anomalies and THC flow are in sync during the highlighted periods. Looking closer at Figure 3, their model runs also show there can be multidecadal periods when the two are out of phase. Does the ECMWF data show that we were recently in one of those out-of-phase periods? Dunno.

The paper that accompanies the ECMWF AMOC reconstruction (page 7) notes that their data contradicts some climate models, but that is with respect to the subpolar gyre.

They also write this: “At 50N, the MOC and strength of the subpolar gyre are correlated at interannual time scales, but show opposite secular trends. Heat transport variability is highly correlated with the MOC but shows a smaller trend due to the warming of the upper ocean, which partially compensates for the weakening of the circulation.”

Hmm. “The warming of the upper ocean…partially compensates for the weakening of the circulation.” Does that refer to the warming of the upper North Atlantic or just the SPG?

I’ll have to check to see if SPG SST anomalies and North Atlantic SST anomalies correlate. I would assume they do but I really need to check. And I’ll need to read some other references. Looks like something for another post.

tallbloke said...

Hi Bob, just returning to Bob Webster's comment about sea levels, and my observations regarding levitus et al and OHC, I just posted this at WUWT:

I just scaled off from the graph presented by the NOAA at their website which is pretty much the one calculated by Levitus et al 2009. It shows an increase in ocean heat content of around 5.5×10^22J between 1993 and 2003. During this time, the mean world sea level rose around 33mm according to satellite altimetry, and around half of that was due to thermal expansion according to IPCC estimates. (levitus is a lead IPCC author).

By my calculation, this amount of heat is only around half that required to get that expansion.

1) The altimetry is wrong.
2) Only a 1/4 of the sea level rise was due to expansion and the ice caps melted more than estimated. (Unlikely)
3) Levitus et al have underestimated the amount of heat stored in the oceans by a factor of two.

Of interest is that Levitus et al only use the ocean temperature data to a depth of 720m. If (3) is correct I conclude that they have missed a lot of the extra heat stored at a deeper level in the ocean while the sun was on it’s strong run of cycles.

Also of interest is that Levitus et al 2000 had the rise in heat content for 1955 – 1994 around 13.7×10^22J, but this seems to have been halved in the most recent effort. Bob Tisdale informs me that this was due to a ‘warm bias correction in the XBT data’. If that older data was in fact correct and the 1993 – 2003 data should be adjusted up by a directly proportional amount, the books would balance and be consistent with an increase of around 0.3C in sea surface temperature over the same period.

tallbloke said...

Bob, do you have any citations or links to other OHC data prior to 2000? I'm interested in discovering the spread of estimates and the methodologies of other scientists to compare with Levitus et al.


Bob Tisdale said...

tallbloke: You asked,"Bob, do you have any citations or links to other OHC data prior to 2000?"

I went through my OHC files and did not find one. Sorry.

tallbloke said...

Bob, I turned your OLR graph upside down and compared it to the last three solar cycles.
The 4W/m^2 swings in OLR match my ocean heat content calculation. Levitus is wrong, I'm certain of it.
The 14x10^22J I get is equivalent to 4W/m^2 and matches the swings in OLR. The ocean emits when the sun gets quiet. Thanks for your input, we'll crack this conundrum.

Bob Tisdale said...

tallbloke: Thanks for the look. Keep at it.



Tips are now being accepted.

Comment Policy, SST Posts, and Notes

Comments that are political in nature or that have nothing to do with the post will be deleted.
The Smith and Reynolds SST Posts DOES NOT LIST ALL SST POSTS. I stopped using ERSST.v2 data for SST when NOAA deleted it from NOMADS early in 2009.

Please use the search feature in the upper left-hand corner of the page for posts on specific subjects.
NOTE: I’ve discovered that some of the links to older posts provide blank pages. While it’s possible to access that post by scrolling through the history, that’s time consuming. There’s a quick fix for the problem, so if you run into an absent post, please advise me. Thanks.
If you use the graphs, please cite or link to the address of the blog post or this website.