October to December 2010 NODC Ocean Heat Content (0-700Meters) Update and Comments
Guest post by Bob Tisdale
INTRODUCTION
The National Oceanographic Data Center’s Ocean Heat Content (OHC) data for the depths of 0-700 meters are available through the KNMI Climate Explorer Monthly observations webpage. The NODC OHC dataset is based on the Levitus et al (2009) paper “Global ocean heat content(1955-2008) in light of recent instrumentation problems”, Geophysical Research Letters. Refer to Manuscript. It was revised in 2010 as noted in the October 18, 2010 post Update And Changes To NODC Ocean Heat Content Data. As described in the NODC’s explanation of ocean heat content (OHC) data changes, the changes result from “data additions and data quality control,” from a switch in base climatology, and from revised Expendable Bathythermograph (XBT) bias calculations.
This update includes the data through the quarter of October to December 2010. There has been an upswing in the Indian Ocean OHC data. And in the tropical Pacific, there’s been a delayed response to ENSO or a downward shift. Other than those, there are no other major changes with the latest 3 months on which to report.
GLOBAL
The Global OHC data through December 2010 is shown in Figure 1. It continues to be remarkably flat, considering the rise that took place during the 1980s and 1990s.
http://i53.tinypic.com/jrsoc3.jpg
Figure 1
In an upcoming post, I’ll present only the post-2003 data, the era when ARGO floats dominated OHC data.
A CHANGE OF COORDINATES
I’ve changed the coordinates of the Indian Ocean and South Pacific data. The coordinates I was using for the Indian Ocean (60S-30N, 20E-145E) caused too much overlap with the North Pacific and Tropical Pacific data. So I’ve shifted the coordinates so that the Indian Ocean is now represented by 60S-30N, 20E-120E. This required that I shift the South Pacific; it’s coordinates are now 60S-0, 120E-90W.
TROPICAL PACIFIC
Figure 2 illustrates the Tropical Pacific OHC data (24S-24N, 120E-90W). The major variations in tropical Pacific OHC are related to the El Niño-Southern Oscillation (ENSO). Tropical Pacific OHC drops during El Niño events and rises during La Niña events.
http://i54.tinypic.com/2vrxw1i.jpg
Figure 2
At least it should. Figure 3 compares tropical Pacific OHC to NINO3.4 SST anomalies (a commonly used ENSO proxy) where the NINO3.4 SST anomalies have been scaled and inverted (multiplied by a scaling factor of -0.15) to help show the relationship. The drop in the tropical Pacific OHC during 2010 is unusual. It should be rising (recharging) during this period. It’s impossible to tell at this time if this is a delayed response or a downward shift.
http://i52.tinypic.com/24n2m2r.jpg
Figure 3
The equatorial Pacific, on the other hand, Figure 4, is responding as one would expect.
http://i53.tinypic.com/f51snm.jpg
Figure 4
We’ll have to keep an eye on the tropical Pacific OHC data.
INDIAN OCEAN
Figure 5 illustrates the Indian Ocean OHC data. Note the sudden upswing since 2006. It’s odd when we consider the trends for most of the other ocean basins since 2003 are flat or negative. (I’ll illustrate this in an upcoming post.)
http://i54.tinypic.com/1fwilz.jpg
Figure 5
The Tropical Pacific OHC dropped and the Indian Ocean OHC rose; one might think warm water has migrated from the Tropical Pacific to the Tropical Indian Ocean. If we combine the Tropical Indian and Pacific subsets and compare it to the Tropical Pacific, Figure 6, we can see the two datasets mimic one another and that the recent drop is suppressed. It’s possible (and likely) there has been some migration of warm water from one subset to the other (likely because the current known as the Indonesian Throughflow does flow between the tropical Pacific and Indian Oceans).
http://i52.tinypic.com/1vj2f.jpg
Figure 6
In fact, this transport appears to take place in the animation of NODC OHC from 2005 to 2010, Animation 1, which was taken from the video that’s included in the post The Electric Kool-Aid Ocean Heat Content Animation.
http://i53.tinypic.com/5dvryd.jpg
Animation 1
And here’s the YouTube video from that post. (The animation with music starts around the 2 minute mark, so check your volume setting if you’re at work.)
YouTube Link:http://www.youtube.com/watch?v=PUONorBCcxU
But the recent rise in Indian Ocean OHC is not limited to the tropics. Figure 7 compares Indian Ocean OHC to the OHC of the Indian Ocean South of 24S. The OHC of the mid-to-high latitudes also has the sudden surge.
http://i56.tinypic.com/9693zl.jpg
Figure 7
And yes, that rise and fall in the OHC of the Indian Ocean South of 24S during the late 1990s does look odd. In fact, if we smooth those two datasets, Figure 8, we can see how unusual that spike appears.
http://i53.tinypic.com/2pqkrpv.jpg
Figure 8
THE HEMISPHERES AND THE REST OF THE BASINS
http://i56.tinypic.com/2j49chc.jpg
(9) Northern Hemisphere
#####
http://i54.tinypic.com/2w67vbn.jpg
(10) Southern Hemisphere
#####
http://i56.tinypic.com/10cqgl5.jpg
(11) North Atlantic (0 to 75N, 78W to 10E)
#####
http://i53.tinypic.com/2vkfehv.jpg
(12) South Atlantic (0 to 60S, 70W to 20E)
#####
http://i55.tinypic.com/r02xrl.jpg
(13) North Pacific (0 to 65N, 100 to 270E, where 270E=90W)
#####
http://i52.tinypic.com/2w1y0dj.jpg
(14) South Pacific (0 to 60S, 120E to 290E, where 290E=70W)
#####
http://i51.tinypic.com/2eb5t39.jpg
(15) Arctic Ocean (65 to 90N)
#####
http://i53.tinypic.com/nvcw0k.jpg
(16) Southern Ocean (60 to 90S)
SOURCE
All data used in this post is available through the KNMI Climate Explorer:
http://climexp.knmi.nl/selectfield_obs.cgi?someone@somewhere
How we can say that calculations are right?
the changes result from “data additions and data quality control,” from a switch in base climatology, and from revised Expendable Bathythermograph (XBT) bias calculations.
If data won’t follow climatemodels, we make changes to data.
Yep that do look like the AMO.
The ocean heat content is a frustrating one. There is a persistent belief that the warm insulating layer of water somehow pushing heat down into the depths of the ocean. They do this without considering the buoyant forces involved. Most of the energy (warmth) is in the top 200 m of the oceans. Even in the tropics.
It would be interesting to see the breakdown in the energy increase for the top 200 m and then the lower 500 m. The top 200m is very effective and preventing energy from going even lower. Much like sea ice is effective at keeping cooling from going deeper. The warm and cold insulative effects are why the Earth is a nice place to live.
The flip side is that a warmer ocean CAUSES the Earth to lose more energy to the atmosphere and then to space. That alone would stop the magical CO2 warming from happening.
John Kehr
Interesting that Arctic Ocean heat content peaked in 2007 and since then is falling. The same year Arctic ice September extent hit a minimum.
Actually, eyeballing the graph in Figure 1, the 1980s was pretty flat (though not as flat as the recent years). The rises were in the 1970s and the 1990s, not the 1980s.
What it does show, though, is that we’ll need to wait another five years or so to see if this is just another pause in the growth, like in the 1980s, or if we’re at a peak.
Looking at the graphs of the North Pacific & Atlantic, the drop in temperature since 2003 ish seems to me to be about .15°, but the NH graph shows only half or less of that amount. Is the Arctic basin that much warmer to offset the difference?
“revised Expendable Bathythermograph (XBT) bias calculations” = ARGO fudges?
Brian H says: “‘revised Expendable Bathythermograph (XBT) bias calculations’ = ARGO fudges?”
Nope. The XBTs are different sensors. They were used well before ARGO came along. The XBT adjustments impacted most of the dataset, and I believe the new adjustments returned a small part of the 1970s-80s hump that had been removed with the adjustments described in the 2009 Levitus et al paper.
Steve Keohane says: “Looking at the graphs of the North Pacific & Atlantic, the drop in temperature since 2003 ish seems to me to be about .15°, but the NH graph shows only half or less of that amount. Is the Arctic basin that much warmer to offset the difference?”
I’m working on another post of the ARGO-era OHC data from 2003-2010. The trend for the North Pacific is slightly postive, and due to its area, it suppresses the North Atlantic’s impact on the Northern Hemisphere.
Also, the units in the OHC graphs are gigajoules per square meter, not deg C.
Question for Bob Tisdale — how reliable is the OHC record prior to the Argo time period? Or looking at it the other way, how reliable is the post 2003 OHC record?
It seems strange that once Argo started being the data source that the overall trend changed dramatically.
So it seems like the three choices are:
pre 2003 data is suspect (and I mean in a major way, not just tweaks); or
post 2003 data is suspect; or
a change in trends just happened to take place at the time of a major shift in instrumentation.
I know there isn’t any solid scientific evidence, but I’m curious as to your opinion or guess.
Dr. Tisdale,
It would help me if all of the graphs had the same vertical scale, or at least if related graphs did.
The first two are scaled -0.3 to + 0.4 and -o.6 to +0.4 — similar but not identical and makes the rise in the first graph appear steeper. The next two are -o.6 to +0.4 and -1.5 to +1.5 (a three times difference in vertical spread).
I suspect that this is a result of automatic chart creation from spreadsheets, which imposes its own scale, but it does not allow easy visual comparison of the resultant graphs. IMHO, all the ‘NODC Ocean Heat Content Anomalies (0-700 meters)’, graphs, for example, should have the exact same scales so that we can compare them at a glance, one to another.
Thank you
John Kehr says: “…The flip side is that a warmer ocean CAUSES the Earth to lose more energy to the atmosphere and then to space. That alone would stop the magical CO2 warming from happening.”
Most alleged evidence of AGW is predicated on the same logic that thinks you can estimate the total number of people in the YMCA pool by counting those heading for the shower.
Kip Hansen says: “It would help me if all of the graphs had the same vertical scale, or at least if related graphs did.”
The problem is that the rise in the OHC of the North Atlantic is so much higher than the other basins that you’d never be able to visually digest the variations in the other basins. Here’s a trend comparison graph of the ocean basins from an earlier OHC post as reference:
http://i49.tinypic.com/11wbm3a.jpg
And that earlier post was this one:
http://bobtisdale.blogspot.com/2010/06/january-to-march-2010-nodc-ocean-heat.html
BTW, it’s just Bob. There’s no Dr. in front of my name.
(The animation with music starts around the 2 minute mark, so check your volume setting if you’re at work.)*
*unless your initials are GS.
Charlie A says: “Question for Bob Tisdale — how reliable is the OHC record prior to the Argo time period?”
Before ARGO, the measurements are primarily in the Northern Hemisphere. As late as 2003, there were very few readings south of 30S. The following link is to the NODC webpage that allows you to select a year or quarter of a given year and select a depth. The website then produces of a map of the locations of the temperature data used in the OHC data:
http://www.nodc.noaa.gov/cgi-bin/OC5/3M_HEAT/showfig.pl?action=start
Select 1955, a depth (400 meters for example), and any 3-month period, and then click on “Show Figure.” You won’t find a lot of data contributing to the dataset then.
Skip ahead a decade at a time and you’ll note how sparse the data is for this dataset until the last few years. ARGO wasn’t fully in place until 2007.
You asked, “Or looking at it the other way, how reliable is the post 2003 OHC record?”
ARGO data has to be corrected.
I love these movies of globs of heat going here and there, disappearing, then reappearing. Why? Because they outright falsify the idea that CO2 warms oceans. The calculated heat producing wavelength of AGW longwave radiation at the ocean surface is no where near being able to drive up or down, this swirling mass of heat. Not even when the color changes from neutral to ever so slightly beigy white warm.
The relative lack of variability of heat content in the Southern Ocean and South Pacific, and the relative stability of sea surface temperature in the Southern Ocean is interesting in terms of the temperature trends in West Antarctica. I think there is some support for the O’Donnell et al position here. A relatively small air temperature trend in West Antarctica seems to be in accord with relatively small trends in the adjacent Southern Ocean. This is simplistic comparison to be sure but sometimes thoses are the best ones.
Good post Bob. Lots of effort involved here.
I think we can ignore all the data before the Argo floats.
All the previous data has been adjusted to fit a pre-conceived viewpoint of what should have happened rather than what did happen.
As soon as the Argo data comes on stream and starts to dominate the data, the trends reverse. There are coincidences. And then there are coincidences.
@ Bob Tisdale — Thanks for the http://www.nodc.noaa.gov/cgi-bin/OC5/3M_HEAT/showfig.pl?action=start link to data distribution figures. The change in sampling of the South Pacific between 2000 and 2005 was pretty dramatic.
I love your “gobs of graphics” posts. They sometimes give me the feeling that the key to understanding what is going on with global warming is right there in front of me, just a little bit beyond my comprehension of what is going on with the OHC.
I’m surprised at the relative attention paid to the GISS/CRU/NOAA global temperature time series as compared to the attention paid to the OHC information. Those oceans are a convenient integrator of energy and give use the closest thing to real time feedback on the overall radiative imbalance of the earth.
Hi Bob, thanks for another excellent post. I don’t know how Josh Willis goes about correcting ARGO data but I suspect there’still a warm bias trying to keep up appearances. Or maybe heat is rising from deeper in the ocean and that will continue for a while.
I agree with Bill Illis that prior to ARGO, the OHC is pretty much a fairy story forced to fit the belief system of AGW proponents. This post on my blog illustrates the basic problem.
http://tallbloke.wordpress.com/2010/12/20/working-out-where-the-energy-goes-part-2-peter-berenyi/
The rise in your global graph from 2001 to 2004 is not plausible. We know albedo increased after 1998. I don’t believe the Levitus reconstruction is anywhere near the mark. There is a splice problem between XBT and ARGO. Whether the earlier XBT data has any global value taken on the basis that the southern hemisphere doesn’t change as much as the north remains to be seen. First we need to sort that splice out.
Bill Illis says: “All the previous data [before ARGO] has been adjusted to fit a pre-conceived viewpoint of what should have happened rather than what did happen. ”
It sounds as if you don’t agree with the XBT fall-rate adjustments. There’s an older OHC dataset on the KNMI Climate Explorer that has the 1970s-80s “bump”. It’s identified as “1958-2004: SODA 0-750m”. I divided it up the same was as I had in the “ENSO Dominates NODC Ocean Heat Content (0-700 Meters) Data” post and the hump is easily explained. Do you think there would be interest in a post about that dataset and the “bump”?
Thanks Bob, I always value your work – and to concur with others – the upper 200m is where most of that energy lies, so thanks for the data links where I can access this – one day I will get time to have a closer look.
Do you have any sense of why the North Atlantic has become such a heat store? Is it transfer of warm water from further south, or some change in insolation resulting from changes in cloud cover or wind patterns? And of course, the two may not be separable as wind patterns determine surface currents. Do we have any reliable data on how the heat content varies with the decadal oscillations such as AMO and AO?
RE: Do you think there would be interest in a post about that dataset and the “bump”?
Yes.
Does the unadjusted (in anyway) data not show the same trends as all/any adjusted data sets? Just ask’n . . .
-barn
The SODA group are still pretty active.
http://www.atmos.umd.edu/~ocean/
They put out a little pamphlet on one of their recent papers putting together 9 different estimates of OHC and most of them show the 1970s bump.
http://www.aviso.oceanobs.com/fileadmin/documents/OSTST/2008/carton_heat_content.pdf
They just published a new analysis of the XBT bias (I can’t find an on-line copy).
http://journals.ametsoc.org/doi/abs/10.1175/2010JCLI3534.1
On the XBT bias, once one reaches a conclusion that there is a bias, then actual experimental empirical studies need to be undertaken to quantify it and then adjust for it. It seems for the most part, they reached a conclusion and developed models for it (which then allows another “bias” to enter the analysis).
I prefer to see you do a post on the Argo data before the 1970s bump and XBT bias issue.
tallbloke: Thanks for the link.
Peter Taylor: I wrote a post on the North Atlantic OHC data right after KNMI added it to the Climate Explorer:
http://bobtisdale.blogspot.com/2009/10/north-atlantic-ocean-heat-content-0-700.html
In addition to that, if the hypothesis of THC/global conveyor belt is correct, the South Atlantic heat transport is from high latitudes to the equator, so the South Atlantic OHC would contribute to the North Atlantic, as you suggested. (though I’ve never been able to show it in an animation of any dataset.) The North Atlantic is also strongly impacted by ENSO through changes in atmospheric circulation, much more than the South Atlantic and Indian Oceans. AMOC can slow during an El Nino, and I would think that impacts the strength of the Gulf Stream as well. North Atlantic Sea Surface Temperatures are strongly impacted by the ENSO-caused changes in the strength of North Atlantic trade winds. There are papers on the impacts of Saharan aerosols on the North Atlantic.
So ENSO, changes in SLP, trade wind strength, AMOC, dust from the Sahara, feedback from ice cap melt, etc., all contribute to the variability.
And of course, if (big if) the North Atlantic OHC continues to drop at it’s present rate for 5-10-15 years, so that it’s long-term trend falls back in line with the other ocean basins, it may not look like such a heat sink.
barn E. rubble says: “Does the unadjusted (in anyway) data not show the same trends as all/any adjusted data sets? Just ask’n . . .”
There are no unadjusted OHC datasets.
If you’re asking whether the trends are different between the newer NODC dataset(s) without the hump and the older one with it, the answer is yes. The updates and corrections keep lowering the trend.
About the warming Indian ocean – is it not the case that La Nina cycles are characterised by strong east to west equatorial trade winds, which – as well as reinforcing Peruvian upwelling – heap up warm water along the western Pacific coasts? Since we have had 2 La Ninas since 2007, then there has been quite a lot of blowing of warm surface water toward the West Pacific since that time. Some of this warm surface water as Bob suggests is likely to leak into the Indian ocean. Thus could it follow that a warming Indian ocean (perhaps at least transiently) is a side-effect of an increased frequency of La Nina systems in the Pacific?