UPDATE (April 23, 2011): I’ve added an addition discussion of the delayed rebound effect at the end of the post.
Judith Curry’s post Earth’s Energy Imbalance is an introduction to the Hansen et al (2011) paper “Earth’s Energy Imbalance and Implications”. Anthony Watts also prepared a post that discussed the sea level rise suggested by Hansen et al (2011). Refer to the Watts Up With That? post NASA’s Hansen thinks sea level rise will be accelerating – I think not, offering a new paper and updated story on Hansen to show why.
The following are comments about the presentation of NODC Ocean Heat Content (Levitus et al 2009) data in Hansen et al (2011). The signal that causes the decadal variations in the Global Ocean Heat “Uptake” data appears to originate in the Tropical Pacific. The post also discusses their proposed “delayed rebound effect” in Ocean Heat Uptake, their failure to include Sea Level Pressure as a variable that impacts Ocean Heat Content, and their use of outdated Total Solar Irradiance data as a forcing.
Hansen et al used NODC Ocean Content data as one of their primary data sources to illustrate Earth’s energy imbalance. 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.
In their Figure 13, Hansen et al present what they called Ocean Heat Uptake. Their Figure 13 is Figure 1 in this post. We’ll concentrate on the green curve in the right-hand cell, the one identified as the “Updates of Levitus et al (2009)” dataset.
I’ve reproduced the Ocean Heat Uptake curve in Figure 2, using the NODC OHC data. The KNMI Climate Explorer is the source of the data, so it’s presented in Gigajoules per square meter (GJ/m^2). The curves are the same in Figures 1 and 2, but the units are different. Hansen et al also appear to use half years (1980.5, 1981.5, etc.) in their graph. Basically, what Hansen et al are illustrating with their Ocean Heat Uptake data are 6-year running trends of the Global Ocean Heat Content data. To start the data, the linear trend of the global ocean heat content for the period of 1978 to 1983 is determined and its value plotted at 1980. The 1981 data point would include the linear trend of the OHC data for 1979 to 1984, and so on, until the last point in 2007, which includes the linear trend for the period of 2005 to 2010. The graph shows that global OHC, based on 6-year trends, was dropping from 1980 to 1982/83. It rose from 1982/83 to 1990, increasing fastest in 1988. There was very little rise in Ocean Heat Content for the 6-year period centered on 1990. Then from 1990 to 2005 OHC rose at varying rates, and basically stopped rising for the 6-year period centered on 2005.
Hansen et al selected 1980 as a start year for their Ocean Heat Uptake graph, but the NODC OHC dataset begins in 1955. If we look at the Ocean Heat Uptake data using the full term of the data, Figure 3, we can see that 1980 was well chosen. The six-year period centered on 1980 had greatest drop in OHC since the period centered on 1965. Choosing 1980 for the start year (Figures 1 and 2) makes the increasing changes through the early 2000s seem significant. But with the entire dataset presented, the early period of positive trends from the late 1970s to the late 1980s suppresses the appearance of the recent wiggles. Highlighting “zero” also changes the perspective.
For those interested, Figure 4 is a comparison of annual NODC Global Ocean Heat Content data versus the Ocean Heat Uptake data, from 1955 to 2010. The OHC data has been scaled by a factor of 0.1 to help visual comparisons. Keep in mind the Ocean Heat Uptake (running 6-year trend) data is “centered” on the 3rdof 6 years, so it skews the dataset slightly.
ENSO DRIVES TROPICAL PACIFIC OCEAN HEAT CONTENT
Back in 2009, my first post on the NODC Ocean Heat Content data after it was included in the KNMI Climate Explorer was a discussion of the impacts of the El Niño-Southern Oscillation (ENSO) on Ocean Heat Content. Refer to the post ENSO Dominates NODC Ocean Heat Content (0-700 Meters) Data. And for those who are not familiar with ENSO, refer to the post An Introduction To ENSO, AMO, and PDO – Part 1.
Figure 5 compares Tropical Pacific (24S-24N, 120E-90W) Ocean Heat Content to scaled NINO3.4 Sea Surface Temperature (SST) anomalies. The NINO3.4 SST anomalies are a commonly used proxy for the frequency and magnitude of El Niño and La Niña events. This comparison illustrates the impact of ENSO on Tropical Pacific Ocean Heat Content.
El Niño events discharge heat from the tropical Pacific, and the tropical Pacific OHC drops in response. In most instances, La Niña events recharge only part of heat released during the El Niño, and the tropical Pacific OHC rebounds but not fully. The La Niña events accomplish this through the strengthening of the Pacific trade winds, which reduces cloud cover over the tropical Pacific, which in turn increases Downward Shortwave Radiations (visible light), which warms the Tropical Pacific. And some of that water that was warmed by the sun collects in the West Pacific Warm Pool until the next El Niño. Then there are the significant La Niña events that can cause the tropical Pacific to gain more heat than was released by El Niño before it. Refer again to Figure 5. The initial portion of the 1973/74/75/76 La Niña not only recharged the heat released during the major 1972/73 El Niño, it also caused an upward shift in the amount of warm waters available for upcoming El Niño events. Then Tropical Pacific Ocean Heat Content decreased from 1980 to 1995, varying in response to the individual El Niño and La Niña events. The 1995/96 La Niña had a significant effect on the tropical Pacific OHC. Note the substantial rise in OHC at that time. The explanation can be found in McPhaden (1999) “Genesis and Evolution of the 1997-98 El Niño.McPhaden writes, “For at least a year before the onset of the 1997–98 El Niño, there was a buildup of heat content in the western equatorial Pacific due to stronger than normal trade winds associated with a weak La Niña in 1995–96.” That weak La Niña caused another upward shift in Tropical Pacific OHC, which served as the fuel for the 1997/98 El Niño. The 1998/99/00/01 La Niña then recharged the Ocean Heat Content, and tropical Pacific OHC has been fluctuating at its new elevated level since then.
THE RELATIONSHIP BETWEEN TROPICAL PACIFIC AND GLOBAL OCEAN HEAT UPTAKE
If we compare Global Ocean Heat Uptake to Tropical Pacific Ocean Heat Uptake, Figure 6, we can see that the Tropical Pacific Ocean Heat Uptake is a major component of Global Ocean Heat Uptake. The two dataset correlate quite well. (For those interested, the correlation coefficient is approximately 0.81.)
If we present the OHC data in monthly format and use 73-month trends instead of 6-year trends, Figure 7, it appears the variations in Global Ocean Heat Uptake are actually responses to Tropical Pacific Ocean Heat Uptake. And since ENSO is the driver of Tropical Pacific Ocean Heat Content , would it then seem logical that variations in Global Ocean Heat Uptake are responses to ENSO? Again, refer to the post ENSO Dominates NODC Ocean Heat Content (0-700 Meters) Data. Since ENSO has a substantial footprint in most climate variables, I’m surprised Hansen et al (2011) didn’t look for the relationship between Tropical Pacific and Global Ocean Heat Uptake, or if they did, why they didn’t note the relationship.
A DELAYED REBOUND EFFECT FROM MOUNT PINATUBO?
Hansen et al (2011) write in the abstract, “A recent decrease in ocean heat uptake was caused by a delayed rebound effect from Mount Pinatubo aerosols and a deep prolonged solar minimum.”
First, it’s difficult to find the impact of Mount Pinatubo on the Ocean Heat Content data for many of the ocean basins. Refer once again to the post ENSO Dominates NODC Ocean Heat Content (0-700 Meters) Data. I included GISS Aerosol Optical Thickness Data (a proxy for the timing and magnitude of explosive volcanic eruptions) in many of the graphs. Only two of the ocean basins, the South Atlantic and South Indian Oceans , show clear signs of the effects from Mount Pinatubo in 1991, but they rebounded quickly, as one would expect. Refer to Figures 9 and 11 in that post. Since the Ocean Heat Content data for the basins show few signs of being impacted by Mount Pinatubo, the claim of a delayed rebound is surprising.
Second, as shown in the recent post ARGO-Era NODC Ocean Heat Content Data (0-700 Meters) Through December 2010, the recent flattening of Global OHC data is caused primarily by the significant drops in the OHC of the North Atlantic and South Pacific, Figure 8. But these are countered by the significant rises in the Indian and South Atlantic Oceans. Are Hansen et al suggesting that the “delayed rebound effect” is only impacting the North Atlantic and South Pacific basins? If Hansen et al are proposing a mechanism for the “delayed rebound effect”, I did not find it. I would have to believe it would be Meridional Overturning Circulation (MOC), and would be based on the assumptions that the cooler waters created by the Mount Pinatubo eruption would be circulated to depths below the 700 meter reach of the NODC OHC dataset and then reemerge 20-plus years later. If so, then the MOC circulation time would have be the similar in the North Atlantic and South Pacific, but not the other ocean basins. And if they are proposing MOC now in their Energy Balance Models, they should also consider that warm and cool waters distributed by ENSO would also be subducted to depths below 700 meters and then reemerge “x” decades later, with more warm water than cool being redistributed during epochs when El Niño events dominate, and vice versa when La Niña events dominate.
The recently negative trends in the North Atlantic and South Pacific OHC data can also be seen in the Ocean Heat Uptake data for the individual ocean basins. That is, for those ocean basins, the Ocean Heat Uptake data have dropped into negative values recently. The North Atlantic is illustrated in Figure 9. Note how the North Atlantic shows escalating trends from 1980 to 2005 but they drop very quickly to negative trends in recent years. The South Pacific is shown in Figure 13. Note how its Ocean Heat Uptake data cycles back and forth from positive to negative, indicating that a negative trend is not unusual for the South Pacific. I’ve also included the global Ocean Heat Uptake data in the graphs of the individual ocean basins as a reference.
SEA LEVEL PRESSURE
I find no mention of Sea Level Pressure in Hansen et al (2011), yet changes in Sea Level Pressure can cause significant changes in Ocean Heat Content. Refer to the posts North Atlantic Ocean Heat Content (0-700 Meters) Is Governed By Natural Variables and North Pacific Ocean Heat Content Shift In The Late 1980s.
SOLAR RADIATION FORCING
Figure 14 shows the Solar Forcing data and impacts presented by Hansen et al (2011). The Total Solar Irradiance (TSI) data in the left-hand cell appears to be based on an early reconstruction by Judith Lean, including background effects, with current TSI data spliced onto the end. The Lean TSI dataset is outdated. Note the early rise in solar minimums from the 1880s to the 1940s.The current understanding of TSI variability is that the minimum TSI in 1880 should be in the vicinity of the recent solar minimums. In other words, that rise from the 1880s to the 1940s does not exist. The use of outdated TSI was discussed in the post IPCC 20th Century Simulations Get a Boost from Outdated Solar Forcings, which was also cross posted at WattsUpWithThat: IPCC 20th Century Simulations Get a Boost from Outdated Solar Forcings. Refer to the comments there by Leif Svalgaard, Solar Physicist from Stanford University. The red curve in the middle cell of Figure 14 shows the impact of the variations in TSI on Global Surface Temperatures. For the last three solar cycles, the temperature variations from solar minimum to maximum should be on the order of 0.07 to 0.1 deg C. Hansen et al (2011) are showing considerably less impact from solar variability.
Hopefully, someday, researchers investigating the causes of Ocean Heat Content variability will treat ENSO as a process instead of noise. El Niño events release heat from the Tropical Pacific, La Niña events recharge it, sometimes overcharge it, and ENSO redistributes heat within the oceans.
I’ll end the post with two animations of Ocean Heat Content from the post ARGO-Era NODC Ocean Heat Content Data (0-700 Meters) Through December 2010. The NINO3.4 SST anomalies to the right serves as a timeline and as an indicator of ENSO phase.
Refer also to the post More Detail On The Multiyear Aftereffects Of ENSO – Part 2 – La Nina Events Recharge The Heat Released By El Nino Events AND…During Major Traditional ENSO Events, Warm Water Is Redistributed Via Ocean Currents. While it is primarily a discussion of Sea Surface Temperature, there are also discussions of Ocean Heat Content.
The Ocean Heat Content and Sea Surface Temperature data presented in this post are available through the KNMI Climate Explorer:
UPDATE (April 23,2011) – ON THE REBOUND
Hansen et al describe the rebound as:
“Fig. 22e shows the effect of volcanic aerosols. Volcanoes cause a negative planetary energy imbalance during the 1-2 years that the aerosols are present in the stratosphere, followed by a rebound to a positive planetary energy imbalance. This rebound is most clearly defined after the Pinatubo eruption, being noticeable for more than a decade, because of the absence of other volcanoes in that period.”
I’ve cropped Cells e from their Figure 22 and included it as Figure 15 of this post. The “rebound” is the “overcorrection” in the Energy Imbalance data in the right-hand cell.
Hansen et al continue:
“The physical origin of the rebound is simple. Solar heating of Earth returns to its pre-volcano level as aerosols exit the stratosphere. However, thermal emission to space is reduced for a longer period because the ocean was cooled by the volcanic aerosols. In calculations via the response function, using equations (1) and (2), the volcanic aerosols introduce a dF/dt of one sign and within a few years a dF/dt of opposite sign. The integrated (cumulative) dF/dt due to the volcano is zero but the negative dF/dt occurred earlier, so its effect on temperature, defined by the climate response function, is greater. The effect of the temporal spacing between the negative and positive changes of F decreases as time advances subsequent to the eruption.”
The problem with the hypothesis: As discussed earlier, the Ocean Heat Content data shows little impact from Mount Pinatubo. The Global Ocean Heat Content would need to drop appreciably in order for the proposed “overcorrection” to exist.