Guest post by Bob Tisdale:
NOTE: This post contains 5 .gif animations that total 10MB. (below the continue reading line) Have patience. They may take a while to load.
This post is a follow-up to the recent post October to December 2010 NODC Ocean Heat Content (0-700Meters) Update and Comments. I wanted to discuss the ARGO-based period separately.
For those new to ARGO, under the heading of “What is Argo?”, the University of California, San Diego Argo webpage describes Argo as a “global array of 3,000 free-drifting profiling floats that measures the temperature and salinity of the upper 2000 m of the ocean.” The UCSD Argo website provides much more information, including an argo.avi video.
Much of the data in this post is supplied by ARGO for the upper 700 meters.
THE ARGO ERA (2003 TO PRESENT)
The NOAA NCEP webapge that presents the Global Ocean Data Assimilation System (GODAS) Input data distributions (1979-present) (Plots) allows users to plot the number of Temperature profiles at different depths for the globe, or for the Atlantic, Indian, and Pacific Oceans. An example of Global data for depths of 250 to 500 meters is shown in Figure 1. According to it, ARGO floats have been in use since the early 1990s, but they had very limited use until the late 1990s. ARGO use began to rise then, and in 2003, ARGO-based temperature readings at depth became dominant. Based on that, I’ll use January 2003 as the start month for the “ARGO-era” in this post.
http://i56.tinypic.com/1448yo3.jpg
Figure 1
Note the significant drop in samples in 2010. I have not found an explanation for this.
The NCEP GODAS Input data (Plots) webpage also allows visitors to create maps of temperature profile locations. Animation 1 is a gif animation that shows the annual data locations from 1979 to 2004. The measurements made with Expendable Bathythermographs (XBTs) are shown in red (x), the moored buoys that are parts of the TAO/ TRITON (Pacific) and PIRATA (Atlantic) projects are shown in green (+), and the blue (o) are ARGO-based measurements. Note how sparse the data is in the Southern Hemisphere prior to the early 2000s, especially south of 30S.
http://i55.tinypic.com/14ikdxs.jpg
Animation 1
Unfortunately, GODAS switched map formats in 2005 and again in 2006, so an animation that included the three map formats would be difficult to watch. The format used in 2005 is unlike those in use before or after, so I’ve excluded it in both animations. Animation 2 shows the Monthly temperature profile locations from January 2006 to December 2010. Note the decline in sampling in 2009/10, especially in the Indian Ocean. Why? Dunno.
http://i55.tinypic.com/2copg03.jpg
Animation 2
ARGO-ERA TREND VERSUS GISS PROJECTION
In past posts, when I’ve compared the NODC Global Ocean Heat Content to GISS projections, I’ve used the rate of 0.98*10^22 Joules per year for the GISS projection. This value was based on Roger Pielke Sr’s February 2009 post Update On A Comparison Of Upper Ocean Heat Content Changes With The GISS Model Predictions. The recent RealClimate posts Updates to model-data comparisons and 2010 updates to model-data comparisons have presented the projections based on Gavin Schmidt extending a linear trend of the GISS Model-ER simulations past 2003. The linear trends in both graphs are approximately 0.7*10^22 Joules per year. I’ll use this value in the comparison, but first a few more notes.
Gavin writes in the 2009 post, “Unfortunately, I don’t have the post-2003 model output handy, but the comparison between the 3-monthly data (to the end of Sep) and annual data versus the model output is still useful,” and he continues, “I have linearly extended the ensemble mean model values for the post 2003 period (using a regression from 1993-2002) to get a rough sense of where those runs could have gone.”
The only paper that I’m aware of in which GISS presented their simulations of Ocean Heat Content was Hansen et al (2005) “Earth’s energy imbalance: Confirmation and implications”. Science, 308, 1431-1435, doi:10.1126/science.1110252 (PDF). In it, they only presented their data from 1993 to 2003. Refer to their Figure 2 (not illustrated in this post).
For those who might be concerned that extending the linear trend does not represent the actual model simulations, refer to Page 8 of the .pdf file GISS ModelE: MAP Objectives and Results. The graph there presents two GISS OHC Model E simulations, one with the Russell Ocean model, the other with the HYCOM Ocean model. The simulations run to 2010 for both models. Do they extend further into the future? And for those who want to attempt to duplicate that comparison of the Model-ER and Model-EH versus the early NODC OHC data, the NODC OHC data (older version) was based on the 2005 Levitus paper “The Warming Of The World Ocean: 1955 to 2003” (Manuscript). Link for the 0 – 700 meters data.
Back to the comparison of the ARGO-era OHC data and the GISS Projection: The most recent version of the NODC OHC data is linked here for 0 – 700 meters. I’ve compared it for the period of 2003-2010 to the GISS projection in Figure 2. Note that I’ve shifted the data down so that it starts at zero in 2003. The GISS projection of 0.7*10^22 Joules per year dwarfs the linear trend of the ARGO-era NODC OHC data. No surprise there.
http://i53.tinypic.com/vh5gtd.jpg
Figure 2
NOTE ABOUT THE DATA
The remainder of the data in this post was downloaded from the KNMI Climate Explorer Monthly observations webpage. The NODC OHC data there is presented in Gigajoules per square meter (GJ/m^2), not the units (10^22 Joules) provided by NODC. That’s why the scale and trends in Figures 2 and 3 are different. The NODC also provides their OHC data on a quarterly basis, but KNMI presents it as monthly data, thus allowing for comparisons to other monthly datasets. This is why the OHC data appears in 3-month tiers in Figures 3, 4 and 5.
GLOBAL AND OCEAN BASIN TRENDS
Figure 3 shows the Global NODC OHC data for the period of January 2003 to December 2010. Comparing its linear trend (0.19 GJ/m^2 per Century) to the trend of the long-term data from 1955 to 2002 shown in Figure 4 (0.52 GJ/m^2 per Century), there has been a significant flattening of the Global OHC data in recent years. And this flattening was not anticipated by the GISS models, which show a continuous rise through 2010.
http://i56.tinypic.com/315mmg6.jpg
Figure 3
####################
http://i54.tinypic.com/sxhqio.jpg
Figure 4
Of course, the oceans are not warming uniformly. Refer to Figure 5. The trends for the North Pacific and the Southern Oceans are basically flat. The only two ocean basins with major increases in OHC during the ARGO era are the South Atlantic and the Indian Oceans, while the North Atlantic, Arctic, and South Pacific Oceans show significant declines in OHC.
http://i56.tinypic.com/28qvtqu.jpg
Figure 5
Note: The coordinates for the ocean basins are:
North Atlantic = 0-75N, 78W-10E
South Atlantic = 60S-0, 70W-20E
Indian = 60S-30N, 20E-120E
North Pacific = 0-65N, 120E-90W
South Pacific = 60S-0, 120E-70W
Arctic = 65N-90N
Southern = 90S-60S
ARGO-ERA CHANGES IN NODC OHC
Figure 6 is a map that displays the change in ARGO-era OHC, from 2003 to 2010. It was created by using 2003 as the base year for anomalies, and plotting the annual OHC values for 2010. Much of the cooling in the North Atlantic has taken place at mid and lower latitudes. In the South Pacific, there was also a decline in the lower latitudes, but there appears to also have been a drop there at higher latitudes along the Antarctic Circumpolar Current (ACC).
http://i51.tinypic.com/21crset.jpg
Figure 6
Animations 3, 4 and 5 present the ARGO-era OHC data, using 12-month averages. The first cells are the average OHC from January to December 2003. These are followed by cells that show the period of February 2003 to January 2004 and so on, until the final cell that captures the average OHC from January to December 2010. The 12-month average reduces the noise and any seasonal component in the data. I’ve also included a graph of NINO3.4 SST anomalies (smoothed with a 12-month filter, and centered on the 6th month) since the effects of ENSO dominate the OHC data. The NINO3.4 SST anomaly graph infills with time. Animation 3 presents global maps.
http://i54.tinypic.com/eu4pzq.jpg
Animation 3
Animation 4 is the North Pole stereographic view. Note the warming of the western tropical North Pacific during the 2007/08 La Niña. It’s tough to miss. There also appears to be a lagged decline in the North Atlantic OHC in response to the 2007/08 La Niña. Will we see a lagged increase there next year?
http://i53.tinypic.com/2mo8fuq.jpg
Animation 4
And Animation 5 is the South Pole stereographic view. Note the persistence of the warm and cool anomalies moving southward from the equatorial Pacific in waves, and also into the South Indian Ocean. I believe those would be classified as oceanic Rossby waves.
http://i54.tinypic.com/2v9rqy0.jpg
Animation 5
CLOSING
Watching the animations, it is very obvious that ENSO and the distribution of warm and cool waters caused by ENSO are major components of Global Ocean Heat Content. Refer to ENSO Dominates NODC Ocean Heat Content (0-700 Meters) Data for further discussion and illustrations. OHC studies such as Hansen et al (2005), however, do not include ENSO in their models. They assume that Anthropogenic Greenhouse Gases have a measurable impact on Ocean Heat Content. The impacts of the failure of GISS to include ENSO and other natural variables in their analysis was illustrated and discussed in detail in Why Are OHC Observations (0-700m) Diverging From GISS Projections?
Refer also to North Pacific Ocean Heat Content Shift In The Late 1980s and North Atlantic Ocean Heat Content (0-700 Meters) Is Governed By Natural Variables.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
This information and analysis is most appreciated. It seems difficult to argue for a large climate sensitivity value unless ocean heat content is rapidly increasing. I look forward to further posts from Bob Tisdale on this topic.
LOL. Can someone mail figure #2 to Gavin?
Beautiful animations. I love data. Animation 5 is astounding.
Fig. 1: The Y-axis is not identified, a must in even elementary math classes.
It is not intuitive what has units up to 20,000.
IanM
Thanks, Anthony.
I am getting very tired of all this talk about models and what they show. In my high school physics class, I learned a central definition of science: it makes successful predictions. For example, Aristotelian physics, produced by “intelligent thought” predicted that a lighter weight would fall at a slower rate than a heavier one. The famous Pisa experiment debunked that one. Newtonian physics predicts that a body in motion will remain in motion unless acted upon by an outside force–such as friction. The measurements on this one can be carried out to three decimal places or more and produce a perfect graph.
The models are notorious for failed predictions, and not just in the new century. They are unscientific.
Calling it “climate science” does not make it so. Only successful predictions make science. It is the skeptics who have that track record. We are the scientists.
A very informative set of animations. Congrats.
“They assume that Anthropogenic Greenhouse Gases have a measurable impact on Ocean Heat Content”
And the mechanism for this is….?
The SST and energy content of the oceans seems to be disconnected, which I think means that as the surface warms, the heat permeates down, increasing the heat content but not “allowing” the surface to get warmer. The CAGW theory has A-CO2 as the reason the heating is happening at all, by warming the air at the air-sea contact. Here is where I see the problem vis-a-vis the CAGW theory:
In order for the sea at depth to get warmer by heat loss from the air, a temperature gradient must continually exist between the air and the sea surface. The tropics have only a 0.24C difference over the last 30 years. The increased power at this stage has got to be about 0.15 W/m2. This is insufficient for the top 700m of high density ocean water to heating up at an accelerating rate as measured. If increased insolation is the cause, however, the 70% surface of the globe that the sea covers, and the ability of sunlight to penetrate the sea surface to >50 m, does give the level of energy the ocean heat content increase represents. AT 1.5% decrease in cloud cover, about 1.4 W/m2 of insolation is added worldwide. A local change in cloud cover of 4.5% over one-third the globe does the same thing. The variation of insolation through the year timed to slight changes in global albedo will do the same with even smaller areas (as Jan-July insolation varies by about 18 W/m2 due to orbital eccentricity alone). Insolation, not atmospheric CO2, seems to have the power to change the oceanic heat content as measured.
So: it seems to me from the above arguments, that the changes in the OHC compared to the changes in atmospheric CO2 and atmospheric temperature demonstrate that insolation, not A-CO2 must be the cause of oceanic warming. The sea is warming the air, not the air warming the sea.
Is the increase in OHC relative to SST and other air temperatures not a fundamental bust wrt AGW?
In your first figure, I don’t understand why you would add the Argo, XBT and TAO data together (your black curve). Surely they were meant to be stand-alone measurements. Secondly, your OHC anomaly graph: with an ordinate in 0.05 Gj/m sqd graduations, it isn’t hard to see why there could be large swings on the graph. What kind of error bars would these figures have.
Perhaps the GISS models are wrong, (assuming that they do model ocean heat content to agree with the AGW theory). ARGO is observed data which shows the expected cyclic nature of heat content.
Ian:
The “y” axis label is the title at the top (i.e., number of temperature profiles / month).
Cheers,
Ian L. McQueen says: “Fig. 1: The Y-axis is not identified, a must in even elementary math classes. It is not intuitive what has units up to 20,000.”
Sorry. I downloaded the graph from the NOAA GODAS website, decreased its size and posted it. Based on the GODAS webpage linked in the post, the y-axis is “number of profiles accumulated monthly”.
Gary Pearse says: “In your first figure, I don’t understand why you would add the Argo, XBT and TAO data together (your black curve).”
Figure 1 is not “my” graph. I had nothing to do with it production. I simply copied it from the NOAA GODAS website:
http://www.cpc.ncep.noaa.gov/products/GODAS/data_distribution.shtml
Mr. Tisdale, in one of your earlier webpage http://bobtisdale.blogspot.com/2008_11_01_archive.html
you published graph(s) for the N. Atlantic Subpolar gyre SST (1854-2008), I would be grateful if you could direct me to the link for the relevant data file.
Thanks.
Hoser says: “Beautiful animations. I love data. Animation 5 is astounding.”
Thanks. Here’s a link to the longer-term version of Animation 3.
It starts in 1990 and runs through 2010. There’s no graph to the right, but the red dots in the upper right-hand corner tick off the years. Note: If you’re at work, watch the volume, because the music kicks in around the 2 minute mark.
Best discussion of the Argo data that is available anywhere.
One would have to ask – what changed in the climate in 2003. Nothing really, accurate data became available that is all. A few more La Nina’s? There is no long-term trend in the ENSO so there should be no long-term trend in OHC as a result of the ENSO. (The pre-2003 data was probably subject to the same math that is applied to the tree-ring reconstructions).
vukcevic says: “Mr. Tisdale, in one of your earlier webpage http://bobtisdale.blogspot.com/2008_11_01_archive.html
you published graph(s) for the N. Atlantic Subpolar gyre SST (1854-2008), I would be grateful if you could direct me to the link for the relevant data file.”
The NOMADS website no longer allows long-tern access the ERSST.v2 (obsolete) data. But it’s still available in through the KNMI Climate Explorer:
http://climexp.knmi.nl/selectfield_obs.cgi?someone@somewhere
And I did list the coordinates in the post…
http://bobtisdale.blogspot.com/2008/11/interesting-correlation-with-north.html
…for the North Atlantic Subpolar Gyre on the graphs as 45N-60N, 60W-30W.
Bob Tisdale writes:
“OHC studies such as Hansen et al (2005), however, do not include ENSO in their models. They assume that Anthropogenic Greenhouse Gases have a measurable impact on Ocean Heat Content.”
You have just blown global warming/climate change/climate disruption/climate whatever out of the water. Brilliant work. Thanks so very much.
CO2 takes public transportation from large cities such as Chicago and Detroit to the beaches. Then it enters the water directly at the beaches. Yes, that is the mechanism. Hansen and Schmidt just haven’t published the paper yet. /sarc – I hope this isn’t necessary.
PS on my biologists diatribe above. I think a survey of biologists’ and biology students’ thoughts, beliefs and policy prescriptions they would like to see concerning CO2-GW would make a very fine post and maybe be a bit of a wakeup call for both biologists, and policymakers.
RE: Figure 3: When ever a simple dataset and its linear regression is shown, I think it is a glorious opportunity to also plot the 80% confidence band of the linear regression. I suspect in the case of Figure 3, that a slope of <=0 is within the 80% confidence.
Showing the confidence bands in Figure 5 would be completely impractical. However, having the 5 trends all pass through the same midpoint give a visual impression that the mean squared error of each sample is small. Maybe error bars at the crossing (or in the left or right margin) could denote uncertainty of the mean to scale.
Raw data placed in an easy-to-understand graphical format on a website that provides widespread distribution and discussion. Thank you gentlemen, you’re doing a real service for the advancement of science.
I had known from previous postings that fluid circulation was a major heat transfer mechanism, but the north and south polar views give a much better idea of just how constrained the flows between the oceanic basins are, and why minor changes in the amount of inflow into areas like the arctic could result in such large ice melt or accumulations.
Compare Figure 1 and Figure 3.
It seems to me, without doing the math, that Figure 3 is a difference of the sample mean at time t – mean of the samples over all time(t).
But from Figure 1, the number of samples N is increasing with time t.
So is the mean sq error of the sample at t=2004 larger than at t=2008 simply because 1/sqr(N(t=2004)) > 1/sqr(N(t=2008)).
Unless the data has been randomely resampled so that N is constant for all time, isn’t homoscedasticity violated? If N changes with time, then Uncertainty in the sampled mean should be plotted on Figure 3 as a trumpet curve opening to the left.
Bob:
Great post!
Has Josh Willis chimed in on your findings? If so and your at liberty to disclose, please do so. If not, will you invite his comments? Thanks.
Brian M. Flynn says: “Has Josh Willis chimed in on your findings?”
Nope. I receive comments from scientists on rare occassions.
This made me think about the recent thread about the lag in the climate system and how that relates to climate sensitivity. I wrote a comment then which argued that the lag due to greenhouse gas forcing need not be the same as the lag due to solar forcing (with the latter probably being much longer). This is because greenhouse gas forcing is in the long wave region of the spectrum which is absorbed within the top few millimetres of the ocean and must therefore leads to immediate surface warming and compensating radiation, evaporation and convection losses . The evidence of these plots is that this is indeed the case and that if the large increase in CO2 since 1955 had any effect it is very small and has already reached equilibrium. There can be no warming in the pipeline if there is no energy to supply it.
However my own view is that surface temperatures are far too closely correlated with the Milankovitch cycles and sun spots for there not to be a strong solar element in climate variation. The fact that TSI is roughly constant is a red herring. A change in solar forcing does not require a change in total energy emitted; a change in spectrum (towards the UV) at the surface could have the desired effect. My hope is that they will publish the energy change at various depths over a whole solar cycle because my hypothesis is that warming at around 100 metres due to variations in solar UV is the real source of the energy inbalance because this energy increase has no compensating energy loss at the surface. This is only corrected when the ocean currents bring the warm pools to the surface where the stored energy can be finally lost to space. The time constant of this is almost certainly much longer (at least decadal) and so I see nothing in these plots that could not be explained by the mid century solar peak.
I am not sayin this is the mechanism but I suggest such a solar forcing is plausible and varifiable. Can someone in the AGW camp please explain the mechanism by which CO2 forcing can produce these results.