Guest post by Bob Tisdale
In this post, I divide the globe (60S-60N) into two subsets and remove the linear effects of ENSO and volcanic eruptions from GISS Land-Ocean Temperature Index data since 1982. This is done using common methods. I further adjust the data to account for secondary ENSO-related processes. The Sea Surface Temperature subsets used for these adjustments are identified. The processes are briefly discussed, supported by links to past posts, and the data are presented that support the existence of these secondary effects. An additional volcanic aerosol refinement that increases the global trend is made. The bottom line is, the GISS LOTI and Reynolds OI.v2 SST data indicates that natural variables could be responsible for approximately 85% of the rise in global surface temperature since 1982. I’ll be the first to point out that I qualified my last sentence with the word “could”. This post illustrates a story presented by the data, nothing more. But this basic evaluation indicates these secondary effects of ENSO require further research.
This post continues with the two-year series of posts that basically illustrate that the effects of El Niño-Southern Oscillation (ENSO) cannot be accounted for using a single index like a commonly used SST-based dataset such as NINO3.4, or CTI, or MEI. These indices represent only the sea surface temperature of the central and eastern equatorial Pacific (that’s modified in the case of the MEI). They do not represent the process of ENSO. They do not account for the warm water that is returned to the western Pacific and redistributed during the La Niña. This post provides further evidence of those effects.
This post is long but I elected not to divide it in two. It’s 6,000 words or 13 single-spaced pages in length. It includes 32 Figures, a gif animation, and a video. So there’s a lot to digest. I tried to anticipate questions and answer them.
REMOVING THE LINEAR EFFECTS OF ENSO AND VOLCANIC AEROSOLS HELP TO SHOW THE TIMING OF THE WARMING
Many papers and blog posts that attempt to prove the existence of anthropogenic global warming remove the obvious linear effects of El Niño-Southern Oscillation (ENSO) events and of stratospheric aerosols discharged by explosive volcanic eruptions. An example is Thompson et al (2009) “Identifying Signatures of Natural Climate Variability in Time Series of Global-Mean Surface Temperature: Methodology and Insights”… http://www.atmos.colostate.edu/ao/ThompsonPapers/ThompsonWallaceJonesKennedy_JClimate2009.pdf
…and its companion paper Fyfe et al (2010), “Comparing Variability and Trends in Observed and Modelled Global-Mean Surface Temperature.”
http://www.atmos.colostate.edu/ao/ThompsonPapers/FyfeGillettThompson_GRL2010.pdf
Let’s run through the process using GISS Land-Ocean Temperature (LOTI) data. That’s their global temperature anomaly dataset with the 1200km radius smoothing. A known problem with that dataset is that GISS Deletes Arctic And Southern Ocean Sea Surface Temperature (SST) Data. Since that creates a bias, we’ll delete the GISS LOTI data where they extend land surface data (with its higher variability) out over the oceans. That is, we’ll confine the data used in this post to 60S-60N.
Someone is bound to complain that I’ve deleted the Arctic data from the GISS LOTI data and that the Arctic is warming much faster than lower latitudes. Keep in mind that the Arctic is amplifying the effects of the rise in temperature at lower latitudes. This is the basis of the concept of polar amplification. If the vast majority of the change in temperature at the lower latitudes is natural, the same would hold true for the Arctic. Regardless, these latitudes were also chosen because the effects I want to illustrate with this post are relatively easy to display using them.
Back to the data: since GISS switches sources for their Sea Surface Temperature data from HADISST to Reynolds OI.v2 data in December 1981, we’ll look at the LOTI data starting in 1982. Smith and Reynolds (2004) Improved Extended Reconstruction of SST (1854-1997)] states the following about the OI.v2 SST data: “Although the NOAA OI analysis contains some noise due to its use of different data types and bias corrections for satellite data, it is dominated by satellite data and gives a good estimate of the truth.”
The truth is a nice place to start.
And we’ll smooth the monthly data with a 13-month running-average filter to lessen noise and season variations.
Figure 1 shows monthly GISS LOTI data (60S-60N), from March 1982 to November 2010, compared to NINO3.4 SST anomalies. The NINO3.4 data represent the Sea Surface Temperature of a region in the central equatorial Pacific bound by the coordinates of 5S-5N, 170W-120W. NINO3.4 SST anomalies are a commonly used proxy for the strength and frequency of El Niño and La Niña events, also known as ENSO. (And for those new to ENSO, refer to An Introduction To ENSO, AMO, and PDO – Part 1.) Note also that the NINO3.4 data has been scaled (multiplied by a factor of 0.16) so that the rises of the two datasets are about the same during the evolution of the 1997/98 El Niño. The NINO3.4 SST anomalies have also been shifted down 0.01 deg C and moved back in time by 3 months (lagged) to align the leading edges of the two datasets at that time. (The data in the graph starts in March 1982 because of the 3-month lag in the NINO3.4 data.) I chose the 1997/98 El Niño because that event wasn’t opposed by a volcanic eruption and it was large enough to overwhelm the background noise. As you can see, the wiggles of lesser El Niño events after 2000 don’t match as well.
http://i53.tinypic.com/deuxdz.jpg
Figure 1
Many of the large year-to-year changes in global temperatures are removed when we subtract the scaled NINO3.4 data from the GISS Global (60S-60N) LOTI data. Refer to the “olive drab” curve in Figure 2. Since the NINO3.4 data has a negative trend since 1982, we increase the trend in the GISS LOTI data by subtracting it. Also note how the ENSO-adjusted GISS LOTI data has “flattened” after 1998. Without the volcano-related dip and rebound starting in 1991, the period from 1988 to 1998 would also be relatively flat. It appears as though the ENSO-adjusted GISS LOTI rose in two steps since 1982. Let’s remove the cooling effects of the El Chichon and Mount Pinatubo eruptions to see if that holds true. We’ll use a GISS dataset that represents Stratospheric Aerosols. (ASCII data) Like the ENSO Proxy, we’ll scale the data and lag it. The estimated range of the impact of Mount Pinatubo on Global Temperatures varies from 0.2 to 0.5 deg C, depending on the study, so we’ll use approximately 0.35 deg C to account for its effect. Visually, that scaling appears right.
http://i51.tinypic.com/8vyd54.jpg
Figure 2
Figure 3 illustrates the GISS Land-Ocean Temperature Index (LOTI) anomaly data with the linear effects of ENSO events and the effects of large volcanic eruptions removed. Also illustrated is the linear trend. I’ve included the linear trend line to illustrate the effect the straight line has on the appearance of the data. The trend line gives the misleading impression that there has been a constant but noisy rise in global temperatures.
http://i55.tinypic.com/zur19i.jpg
Figure 3
During the discussion of Figure 2, I noted that the data appeared to flatten after 1998. The upward steps in the data can be illustrated if we present the period average temperature anomalies for the three periods of 1982 to 1987, 1988 to 1997, and 1998 to 2010.
http://i55.tinypic.com/m96sly.jpg
Figure 4
WHAT COINCIDES WITH THE UPWARD STEPS?
The timings of those upward steps coincide with the transitions from the large El Niño to La Niña events that took place in 1988 and 1998. This can be seen in Figure 5, which includes the adjusted GISS LOTI data. The other dataset is scaled NINO3.4 SST anomalies that have been inverted (multiplied by a negative number). Figure 5 is a gif animation, and in it, the NINO3.4 data shifts up and down. That was done to show how precisely the upward steps in the adjusted GISS data coincide with ENSO transitions. The adjusted GISS data trails the NINO3.4 data by a month or two. And the scales are correct for both upward steps.
http://i54.tinypic.com/2i8b02b.jpg
Figure 5
But how can ENSO be impacting global temperature data if we’ve subtracted the scaled NINO3.4 anomaly data from the Global (60S-60N) GISS LOTI anomalies?
LA NIÑA EVENTS ARE NOT THE OPPOSITE OF EL NIÑO EVENTS
The assumption made when we removed the linear effects of ENSO (discussion of Figure 1) was that La Niña events were the opposite of El Niño events. But they are not. (This is the same incorrect assumption made by papers like Thompson et al 2009). This post is very long and to adequately describe how La Niña events are not the opposite of El Niño events would make it much longer. So it will be best to provide links to earlier detailed discussions on this topic.
Refer to:
More Detail On The Multiyear Aftereffects Of ENSO – Part 1 – El Nino Events Warm The Oceans
And:
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.
And:
I provide a relatively brief description in the following section.
WHY ENSO INDICES LIKE NINO3.4 SST ANOMALIES DO NOT ACCOUNT FOR THE PROCESS OF ENSO
ENSO is a process, and ENSO indices such as NINO3.4 SST anomalies, the Cold Tongue Index (CTI), or the Multivariate ENSO Index (MEI) do not account for that process.
El Niño description: A reduction in the strength of the Pacific trade winds triggers an El Niño. A number of interrelated events then take place. Huge amounts of warm water from the surface and, more importantly, from below the surface of the western tropical Pacific (the Pacific Warm Pool) slosh east during an El Niño and are spread across the surface of the central and eastern equatorial Pacific. The increased area of warm water on the surface allows the tropical Pacific Ocean to discharge more heat than normal into the atmosphere through evaporation. That, combined with the change in location of the convection, cause drastic changes in global atmospheric circulation patterns. As a result, global temperatures vary. And most parts of the globe outside of the central and eastern tropical Pacific warm during an El Niño. The changes in atmospheric circulation work their way eastward–over the Americas, the Atlantic, Europe and Africa, the Indian Ocean and Asia. Eventually, the changes reach the western Pacific, but by that time, the El Niño is transitioning to a La Niña.
Refer again to the NINO3.4 SST anomaly data in Figure 1. A La Niña event, based on the temperature values on a graph, appears to be an El Niño of the opposite sign, and for some regional responses in temperature and precipitation that is true. But as noted before the use of NINO3.4 SST anomalies or other ENSO indices does not capture the fact that ENSO is a process. Those indices fail to account for the relocation and redistribution of huge amounts of warm water.
In the description of the El Niño, I noted that huge amounts of warm water from the surface and below the surface of the West Pacific Warm Pool had sloshed east during an El Niño. What happens to all of that warm water from below the surface of the Pacific Warm Pool that had been spread across the surface of the central and eastern tropical Pacific during the El Niño? Before the El Niño, it was below the surface and not included in the measured global surface temperature anomalies. During the El Niño, some of the warm water that had been below the surface is now on the surface of the central and eastern tropical Pacific and included in the measured global temperature. In response, surface temperatures there rose. The ENSO index captures that part of the process and only that part.
But during the La Niña, what happens to the warm water? It wasn’t all “used up” by the El Niño. And what happens to all of the subsurface warm water that had shifted east during the El Niño and had remained below the surface. It doesn’t simply disappear during the La Niña. Answering those questions explains why La Niña events are not the opposite of El Niño events, and why an ENSO index does not capture the aftereffects of an ENSO event.
The leftover warm water returns to the western Pacific. This is accomplished in a few ways. One is through a phenomenon called a slow-moving Rossby Wave. This can be seen in Video 1. It illustrates global Sea Level Residuals from January 1998 to June 2001 and captures the 1998/99/00/01 La Niña in its entirety. The video was taken from the JPL video “tpglobal.mpeg”. The slow moving Rossby wave is shown as the westward moving band of elevated sea level at about 10N. Watch the effect it has on western Pacific Sea Level Residuals when it reaches there.
The second way that the leftover warm water is carried to the Western Pacific is through a strengthening of the trade winds. During a La Niña event, trade winds strengthen above their “normal” levels and the ocean currents carry the warm water back to the west and then poleward.
Animation 1 is taken from the videos in the post La Niña Is Not The Opposite Of El Niño – The Videos. It presents the 1997/98 El Niño followed by the 1998 through 2001 La Niña. Each map represents the average Sea Surface Temperature anomalies for a 12-month period and is followed by the next 12-month period in sequence. Using 12-month averages eliminates the seasonal and weather noise. The effect is similar to smoothing data in a time-series graph with a 12-month running-average filter.
There are a number of things to note in Animation 1. First, the El Niño and La Niña events cause changes in the sea surface temperatures in the central and eastern tropical Pacific. The NINO3.4 SST anomalies used in this post are a measure of that variation in the central equatorial Pacific, and only that variation. Second, during the El Niño, note how the sea surface temperatures warm first in the Atlantic, then in the Indian Ocean, and then in the western Pacific. The warming is caused by changes in atmospheric circulation. And by the time these changes in atmospheric circulation make their way east to the western Pacific and it starts to warm there, the El Niño is transitioning to La Niña. Third, note how the sea surface temperature anomalies in the Western Pacific (and East Indian Ocean) continue to rise as the La Niña event strengthens. Fourth, note how the SST anomalies remain elevated in the East Indian and West Pacific Oceans during the entire term of the 1998/99/00/01 La Niña.
http://i53.tinypic.com/etb58j.jpg
Animation 1
The Sea Surface Temperatures of the East Indian and West Pacific Oceans remain elevated during the La Nina because the stronger trade winds reduce cloud cover. The reduction in cloud cover allows more Shortwave Radiation (visible light) to provide additional warming of the tropical Pacific waters east of the Pacific Warm Pool. The ocean currents carry this sunlight-warmed water to the west and then poleward.
DIVIDING THE GLOBE IN TWO HELPS IDENTIFY THE REASONS FOR THE UPWARD STEPS IN THE GISS LOTI DATA
To help illustrate the reasons for the upward shifts in the ENSO- and Volcano-adjusted GISS LOTI data (Figure 4), let’s divide the data into two subsets split at 20N. Refer to Figure 6.
http://i52.tinypic.com/jjpl5c.jpg
Figure 6
First we’ll look at the Northern Hemisphere GISS LOTI anomaly data, north of 20N. It has a relatively high linear trend since 1982, about 2.8 deg C/Century. Part is due to the additional variability of the North Atlantic. To compound that, these latitudes have a relatively high land surface area, and land surface temperatures vary much more than sea surface temperatures. The land surface area of the Northern Hemisphere latitudes of 20N-60N is about 45% of the total surface area, but the land surface in the tropical and Southern Hemisphere latitudes of 60S-20N is only 17%.
http://i53.tinypic.com/20mdjc.jpg
Figure 7
The dataset shown in Figure 7 has not been adjusted for ENSO or volcanic eruptions. Let’s correct first for ENSO, then for the volcanic eruptions, using the same methods we did for the Global (60S-60N) data. Figures 8 and 9 illustrate the interim steps and the required scaling factors, and Figure 10 illustrates the result.
http://i53.tinypic.com/15odzkh.jpg
Figure 8
http://i52.tinypic.com/34jdp94.jpg
Figure 9
http://i55.tinypic.com/vmx18x.jpg
Figure 10
The Northern Hemisphere data still has a relatively high trend, approximately 2.2 deg C/Century. But what causes the additional variability if we’ve removed the effects of ENSO and volcanic eruptions? The additional variations are often described as noise, but they have sources.
THE KUROSHIO-OYASHIO EXTENSION HOLDS THE ANSWER
There is a strong ENSO-related warming of the Kuroshio-Oyashio Extension that occurs during La Niña events. This was discussed and illustrated in my recent post The ENSO-Related Variations In Kuroshio-Oyashio Extension (KOE) SST Anomalies And Their Impact On Northern Hemisphere Temperatures. That secondary warming can be used to explain a major portion of the year-to-year variability in Northern Hemisphere land and sea surface temperature. And, along with ENSO, it helps to explain nearly all of the variations in the Northern Hemisphere (20N-60N) GISS LOTI data, including the rising trend. Figure 11 illustrates the location of the KOE dataset used in this post (30N-45N, 150E-150W).
http://i52.tinypic.com/14twvox.jpg
Figure 11
The GISS LOTI anomalies for much of the Northern Hemisphere warm (cool) when the Kuroshio-Oyashio Extension SST anomalies warm (cool). This can be seen in the correlation map of annual (January to December) Kuroshio-Oyashio Extension SST anomalies and annual Northern Hemisphere (0-90N) GISS LOTI data, Figure 12. Also note the correlation with the North Atlantic.
http://i54.tinypic.com/303llxg.jpg
Figure 12
As mentioned above, the secondary warming of the Kuroshio-Oyashio Extension was discussed in detail in my recent post The ENSO-Related Variations In Kuroshio-Oyashio Extension (KOE) SST Anomalies And Their Impact On Northern Hemisphere Temperatures. A quick description of the process: During a La Niña event, leftover warm water from the El Niño is returned to the Western Pacific and spun poleward by the North and South Pacific gyres. Much of that warm water finds its way to the Kuroshio-Oyashio Extension, where it apparently impacts atmospheric circulation.
The agreement between the variations in KOE SST anomalies and the adjusted Northern Hemisphere GISS LOTI anomalies is shown in Figure 13. I find that match quite remarkable. The additional spike (highlighted in blue) in the KOE data that starts in 1990 is out of place. It will make itself known later in this post. The other thing to note is the scaling factor required to align the two datasets in Figure 13. The scaling factor of 0.7 is very high. We’ll discuss this later in the post.
http://i51.tinypic.com/2j15udv.jpg
Figure 13
Some might think the agreement between those datasets is a lucky coincidence. Of course, the agreement between the adjusted LOTI data and the unadjusted KOE data in Figure 13 is based solely on the lags and scaling factors I used. But the scaling and lags were established logically. Eyeballing the data, the scaling factors appear to be correct. And as we shall see, using the same methods, the results are very similar for the data that covers the Tropics and Southern Hemisphere.
LET’S LOOK AT THE TROPICS AND SOUTHERN HEMISPHERE
The Southern Hemisphere and Tropics dataset includes the GISS LOTI data from 60S-20N, Figure 14. This subset has a relatively low trend, approximately 1 deg C/Century. Some of this is related to the amount of continental land mass. For these latitudes, land represents only about 17% of the surface area. The Southern Ocean (90S-60S), which is outside of the latitudes portrayed in the post, also impacts the Southern Hemisphere data. And since the Southern Ocean SST anomaly trend over this period is negative, its interaction with the Southern Hemisphere oceans lowers the trend of the dataset.
http://i54.tinypic.com/eiqtsy.jpg
Figure 14
And again, using the same methods, we’ll adjust for ENSO, then volcanic eruptions, Figures 15 and 16, and present the results, Figure 17. Refer to Figures 15 and 16 for the scaling factors.
http://i56.tinypic.com/e5nxg1.jpg
Figure 15
http://i56.tinypic.com/2vx28tt.jpg
Figure 16
http://i53.tinypic.com/2cqy0s3.jpg
Figure 17
As shown in Figure 17, removing the effects of the volcanoes has once again lowered the trend, and removing the ENSO data reduced the year-to-year variations.
Now we need a dataset for these latitudes to illustrate the secondary warming due to the leftover warm water from El Niño events and use it to account for the adjusted GISS LOTI data for the latitudes of 60S-20N.
THE SOUTH PACFIC CONVERGENCE ZONE (SPCZ) EXTENSION SST ANOMALY DATA AND CORRELATION MAP ARE REVEALING
The KOE was used in the discussion of the Northern Hemisphere data, so it seems logical that a similar area exists in the South Pacific. And for this discussion, we’ll designate that area as the South Pacific Convergence Zone (SPCZ) Extension. The SPCZ Extension data will be the SST anomalies of the area east of Australia (35S-20S, 160E-150W). As shown in Figure 18, it had a relatively high SST anomaly at the peak of the 1998/99 portion of the 1998 through 2001 La Niña.
http://i56.tinypic.com/2zhpb92.jpg
Figure 18
The SST anomalies for SPCZ Extension are shown in Figure 19.
http://i51.tinypic.com/2dinmgw.jpg
Figure 19
Like the KOE Extension data, the SST anomalies of the SPCZ Extension warm greatly during transitions from El Niño to La Niña events and appear to shift upward at those times. Refer to Figure 20.
http://i54.tinypic.com/35mf4v8.jpg
Figure 20
Creating the correlation map of annual (January to December) SPCZ Extension SST anomalies and annual Tropical and Southern Hemisphere (90S-20N) GISS LOTI data was eye-opening. It appears the SPCZ data is a good proxy for those areas in the western tropical Pacific and southwest Pacific that warm during La Niña events. It would also appear to show the effects those western Pacific areas have on the rest of the globe. As we can see in Figure 21, when the SPCZ Extension warms (cools) many areas throughout the tropics and Southern Hemisphere warm (cool). But as illustrated in Figure 20, the warming that occurs during La Niña events is not counteracted by the cooling during El Niño events. This causes the data to rise in steps during the La Niña events.
http://i56.tinypic.com/30d9xt0.jpg
Figure 21
Does the correlation map indicate that the upward shifts in the SPCZ Extension data also exist in the tropical and Southern Hemisphere GISS LOTI data? My understanding of correlation maps is that they emphasize the larger events in the data, and if we refer again to Figure 20, the larger events are those that occur during these upward shifts. We can also confirm this by comparing the respective time-series graphs.
Figure 22 illustrates the adjusted GISS LOTI data for the Tropics and Southern Hemisphere north of 60S. Also shown are scaled (0.25) SPCZ Extension SST anomalies. There are minor divergences from time to time, but in general the two curves agree surprisingly well.
http://i53.tinypic.com/6tmj9y.jpg
Figure 22
WHAT’S THE BOTTOM LINE?
What do the curves and linear trends of the adjusted GISS LOTI data look like if the KOE and SPCZ Extension data are removed? And what happens when you combine the two results to form a global dataset with all of the adjustments? Let’s take a look. The Northern Hemisphere GISS LOTI data (20N-60N) that’s been adjusted for ENSO and volcanic aerosols and the KOE SST anomalies is shown in Figure 23. Recall the divergence circled in blue in Figure 13; that’s the cause of the significant additional dip in 1990. Other than that, this was not a bad first attempt with scaling factors. But notice how small the trend is, 0.13 deg C/Century. If that dip was removed, the trend would be even lower.
http://i51.tinypic.com/jze2o9.jpg
Figure 23
The Tropical and Southern Hemisphere GISS LOTI data (60S-20N) with the ENSO, Volcano, and SPCZ Extension adjustments is shown in Figure 24. The trend is basically flat. This dataset appears noisy, but look at the temperature scale. The range is only one-quarter of one used in Figure 23.
http://i53.tinypic.com/67i7iu.jpg
Figure 24
We can combine the Northern Hemisphere data (20N-60N) with the Tropical and Southern Hemisphere data (60S-20N) using a weighted average. (The latitudes of 20N-60N represent approximately 29% of the surface area between 60S-60N.) Figure 25 shows the result. The linear trend is basically flat at 0.06 deg C/Century. The saw-tooth pattern is interesting, but…
http://i53.tinypic.com/29cw9dc.jpg
Figure 25
Due to the timing, the saw-tooth pattern appears to indicate that there was a lagged (repeated) volcano signal in the data. Refer to Figure 26. The reason I say repeated is that originally when the volcanic signal was removed, the Aerosol Optical Depth data was lagged 3 months and the leading edges of the data aligned well in Figures 9 and 16. The volcano signals in Figures 25 and 26, assuming those spikes are volcano signals, are lagged 9 months. The additional signal may also simply mean the Sato Mean Optical Thickness data doesn’t account perfectly for the decay of the volcano signal and that an additional adjustment is required.
http://i51.tinypic.com/6rjxpg.jpg
Figure 26
So let’s make the secondary volcano correction, refer to Figure 27. That will raise the linear trend of the adjusted GISS LOTI data.
http://i54.tinypic.com/2a8s204.jpg
Figure 27
After all of the adjustments are made, there is a small trend, about 0.24 deg C/Century. Compared to the original, unadjusted data, Figure 28, the trend of the adjusted data is only about 15% of the original GISS LOTI data for 60S-60N.
http://i56.tinypic.com/wnxa9.jpg
Figure 28
This makes perfect sense since there is little to no evidence of an anthropogenic global warming effect on global Ocean Heat Content (OHC) data. All one needs to do is divide the global oceans into tropical and extratropical subsets per ocean basin. Then it’s relatively easy to determine that ENSO, changes in Sea Level Pressure, and AMO/AMOC are responsible for that vast majority of the rise in OHC since 1955. Refer to:
A. ENSO Dominates NODC Ocean Heat Content (0-700 Meters) Data
B. North Pacific Ocean Heat Content Shift In The Late 1980s
C. North Atlantic Ocean Heat Content (0-700 Meters) Is Governed By Natural Variables
SHOULDN’T THE KUROSHIO-OYASHIO EXTENSION AND SPCZ EXTENSION DATA BE DETRENDED?
In this post and in The ENSO-Related Variations In Kuroshio-Oyashio Extension (KOE) SST Anomalies And Their Impact On Northern Hemisphere Temperatures, we illustrated that the Kuroshio-Oyashio Extension and South Pacific Convergence Zone Extension SST anomalies rise in steps during La Niña events. Since those upward steps are clearly responses to ENSO, there should be no need to detrend those datasets.
A NOTE ABOUT THE ATLANTIC MULTIDECADAL OSCILLATION
There is a natural variable I did not account for in this post, and it is the Atlantic Multidecadal Oscillation, or AMO. I did not remove its impacts on the Northern Hemisphere data. For those new to the AMO, refer to An Introduction To ENSO, AMO, and PDO — Part 2.
As noted in that post, RealClimate defines the Atlantic Multidecadal Oscillation (“AMO”) as, “A multidecadal (50-80 year timescale) pattern of North Atlantic ocean-atmosphere variability whose existence has been argued for based on statistical analyses of observational and proxy climate data, and coupled Atmosphere-Ocean General Circulation Model (“AOGCM”) simulations. This pattern is believed to describe some of the observed early 20th century (1920s-1930s) high-latitude Northern Hemisphere warming and some, but not all, of the high-latitude warming observed in the late 20th century. The term was introduced in a summary by Kerr (2000) of a study by Delworth and Mann (2000).”
I could have accounted for the AMO before removing the impacts of ENSO and the volcanic eruptions. But I chose to leave it in so that I could include the impact of the KOE on the North Atlantic.
As shown in Figure 29, the trend of the North Atlantic SST anomalies between 20N-60N is 70% higher than the North Pacific SST anomalies trend. By accounting for that additional “some, but not all” trend from the AMO, the scaling factor required to align the KOE dataset with the North Hemisphere data would drop.
http://i52.tinypic.com/x4lx5t.jpg
Figure 29
THE KOE SCALING IS TOO HIGH
The scaling factor for the Kuroshio-Oyashio Extension data in Figure 13 was 0.7. To some, it would not seem likely that the secondary warming of the KOE could raise temperatures for the Northern Hemisphere (20N-60N) that high, especially when one considers the multiplier for the SPCZ Extension was 0.25 in Figure 22.
First: Let’s consider the known effects of an El Niño event. When surface temperatures around the globe warm in response to an El Niño, most of those areas warm due to changes in atmospheric circulation. That is, they do not rise because the heat released into the atmosphere is warming the land and sea surfaces. The following is an example I often use. During an El Niño, the tropical North Atlantic warms even though it is separated from the Pacific by the Americas. The tropical North Atlantic warms during the El Niño because the El Niño causes a weakening of the North Atlantic trade winds. With the decrease in Atlantic trade wind strength there is less evaporation, and if there is less evaporation, sea surface temperatures rise. There is also less upwelling of cool water from below the surface when the trade winds weaken. This also causes sea surface temperatures to rise.
Therefore, it is through teleconnections or atmospheric bridges, not the direct transfer of heat, that the KOE would impact the areas of the Northern Hemisphere.
Second: There is a second western boundary current extension in the Northern Hemisphere, and it is the Gulf Stream Extension in the North Atlantic. For this quick discussion, we’ll define the Gulf Stream Extension by the coordinates of 35N-45N, 75W-30W. The map in Figure 30 is a correlation map and it shows that when the Gulf Stream Extension warms (cools) there are many parts of the Northern Hemisphere that warm (cool). And note that the eastern tropical Pacific is negatively correlated, indicating that these areas warm during La Niña events.
http://i52.tinypic.com/x1alur.jpg
Figure 30
Scroll back up to Animation 1. It also shows the parallel warming of the Gulf Stream Extension with the KOE.
But do the SST anomalies of the Gulf Stream Extension cool during El Niño events? As shown in Figure 31, the SST anomaly variations of the Gulf Stream Extension and the Kuroshio-Oyashio Extension are very similar. Both datasets can warm significantly during La Niña events but they do not drop proportionally during El Niño events. In an earlier linked post, I described the process that causes the KOE to warm, but I have not found a paper that describes the warming of the Gulf Stream Extension at those times. Why does the Gulf Stream Extension respond differently to El Niño and La Niña events? Like the KOE, is the warm water created during an El Niño also carried north by the Gulf Stream during the following La Niña? Do the changes in atmospheric circulation caused by the La Niña add to the warming? During the La Niña, does an increase in the strength of the North Atlantic trade winds also reduce cloud cover over the tropical North Atlantic? Does the warm water created by the decrease in cloud cover and resulting increase in sunlight then get transported to the Gulf Stream Extension? There are too many unanswered questions for me to use the Gulf Stream Extension data in this post.
http://i52.tinypic.com/4g6i9w.jpg
Figure 31
But, the parallel warming of the KOE and the Gulf Stream Extension during the transitions from El Niño to La Niña events would help to reduce the KOE scaling factor required to explain the step changes in the adjusted GISS LOTI data.
WHAT ABOUT SOLAR?
If we scale sunspot numbers so that the variations from solar minimum to maximum represent about a 0.1 deg change in temperature, and if we lag the sunspot data 6 years, it compares well visually with the adjusted GISS LOTI data. Refer to Figure 32. Someone with additional data processing tools could duplicate the steps taken in this post and confirm how well the two curves align.
http://i51.tinypic.com/23jsjo1.jpg
Figure 32
WHAT FUELS THE EL NIÑO EVENTS?
The warm water created during the previous La Niña(s) via the increase in Downward Shortwave Radiation (visible light) fuels El Niño events. This was discussed in 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.
CAN THE THIS TYPE OF EVALUATION BE EXTENDED BACK IN TIME?
I would not expect that what was presented in this post could be extended back in time. The Pacific climate shifted in 1976/77. In the abstract of Trenberth et al (2002), they write, “The 1976/1977 climate shift and the effects of two major volcanic eruptions in the past 2 decades are reflected in different evolution of ENSO events. At the surface, for 1979–1998 the warming in the central equatorial Pacific develops from the west and progresses eastward, while for 1950–1978 the anomalous warming begins along the coast of South America and spreads westward. The eastern Pacific south of the equator warms 4–8 months later for 1979–1998 but cools from 1950 to 1978.”
The way ENSO events interacted with the Kuroshio-Oyashsio Extension and the SPCZ Extension also appear different before and after 1979 in the correlation and regression analyses presented in that paper. Link to Trenberth et al (2002):
http://www.cgd.ucar.edu/cas/papers/2000JD000298.pdf
SOURCES
Most of the data used in this post are available through the KNMI Climate Explorer Monthly observations webpage. GISS LOTI is identified there in the second field under “Temperature” as “1880-now anomalies: GISS”, with the “1200km” radius smoothing. The Reynolds OI.v2 is listed under SST as “1982-now: 1° Reynolds OI v2 SST”. The coordinates used are identified in the text and/or on the graphs.
And if you want to attempt to duplicate my results but have never used the KNMI Climate Explorer, refer to the post Very Basic Introduction To The KNMI Climate Explorer for a place to start.
The dataset used to simulate the impacts of the volcanic eruptions is available through GISS:
http://data.giss.nasa.gov/modelforce/strataer/tau_line.txt
The Sunspot data is available through the KNMI Climate Explorer Monthly climate indices webpage. Refer to the Sunspots (1749-now, SIDC) field under the heading of “Sun”.
CLOSING REMARKS
This was a very basic attempt to approximate the effects of natural variables on global temperatures, using scaling and lags that were eye-balled. Sometimes basic things work well, and in this case, they appear to have done that. The similarities between the adjusted GISS LOTI datasets and the respective KOE and SPCZ Extension data were remarkable. While those similarities and the correlation maps do not prove the KOE and SPCZ Extension SST anomalies cause those addition rises in surface temperature, they imply that natural factors are causing the upward steps in global temperatures illustrated in Figure 4.
After some preliminary discussions, I divided the global (60S-60N) GISS LOTI data into two sections. The linear impacts of ENSO and volcanic eruptions were then removed from those subsets. The processes that cause the Sea Surface Temperatures in two parts of the Pacific to warm greatly during La Niña events were discussed. The unadjusted SST anomalies of the KOE and the SPCZ Extension were then compared to their respective adjusted GISS LOTI anomalies. The related curves were surprisingly similar. After removing the impacts of the KOE and the SPCZ Extension from the related GISS LOTI data, the linear trends dropped significantly. When the two GISS LOTI datasets were again combined, we had removed approximately 85% of what some consider to be the “anthropogenic global warming signal.”
This post differs from studies such as Thompson et al (2009). Thompson et al assumed that the ENSO proxy accounts for all of the processes within the Pacific that take place during ENSO events. In reality, NINO3.4 SST anomalies (or the CTI SST anomalies they used) can only account for the linear responses to the changes in equatorial Pacific SST anomalies. NINO3.4 SST anomalies cannot be assumed to account for the ENSO processes that take place within the Pacific or the aftereffects of those processes. What I presented in this post was a simple way to view those aftereffects within the Pacific and the global responses to them.
In short, I presented a story told by the GISS Land-Ocean Temperature Index and Reynolds OI.v2 SST data between the latitudes of 60S to 60N.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Bob,
Thanks for the interesting study. I’ve linked it to a local blog I participate in.
I’ve seen concern for the use of “eye-balled”. Can you explain or qualify this?
Bob from the UK says:
January 11, 2011 at 12:56 am
“Arctic ice has only been measured during the warm PDO, since the PDO has turned cold, the ice has now stabilised and since 2007 has grown.”
____
Uh, no, not quite. We’ve got the lowest January levels of Arctic sea ice ever on satellite record following the lowest December level ever. Saying the ice has “now stabilized” is wishful skeptical hyperbole at best…
Bob_FJ says:
January 10, 2011 at 10:35 pm
“..
3) However, what can be said is that what predominantly heats the oceans is sunlight. On the other hand, GHG induced back radiation has little effect because being long-wave, it is extinguished in the skin of the water, is able to re-radiate virtually instantly, and can only penetrate via rapid mechanical mixing…”
You have been misinformed about how down-welling radiation from the GHG’s operates to warm the oceans. Penetration of the down-welling radiation from the GHG’s, into the bulk of the ocean, is not required for the long wave radiation to have an effect on ocean temperatures.
The sun’s rays warm the top layer of the ocean down to 10’s of meters below the surface. The heat from the sun flows to the top of ocean by convection. The rate of flow is driven by the temperature difference between the cooler top surface and the bulk below, where the heat is absorbed.
The absorption of the down-welling radiation increases the temperature of the the very top surface of the ocean. This reduces the difference between the top surface and the bulk of the ocean which absorbs the sun’s heat, reducing the rate of convection of heat absorbed by the bulk of the ocean to the top surface. This is the mechanism by which the down-welling radiation from GHG’s make the oceans warmer. The downward radiation itself does not have to penetrate into the bulk of the ocean to be effective in making it warmer.
http://www.realclimate.org/index.php/archives/2006/09/why-greenhouse-gases-heat-the-ocean/
“R. Gates says:
January 11, 2011 at 10:31 am
We’ve got the lowest January levels of Arctic sea ice ever on satellite record following the lowest December level ever.”
At this time of year the range of ice cover from year to year is so small that that is insignificant.
Wake me up in March.
Re that word ‘evah’. Do you mean since 1979 on a planet 4 billion years old?
Bob Tisdale says:
January 11, 2011 at 5:51 am
“eadler replied, “It seem to me that you claim that a warmer ocean surface will make the heat content of the ocean increase even more. Since an increase in heat content would make the surface temperature likely to become even warmer.”
Not sure how you assumed that from my posts, eadler.
You continued, “In fact if a strong self sustaining unforced warming of the oceans as we have seen, around 0.8W/M^2 does exist, the climate of the earth would be in the process of an unstable runaway condition.”
An assumption on your part. What I have shown in those posts is that the rise in OHC from 1955 to present can be explained by natural factors that have resulted in upward shifts in OHC, not a monotonous rise. The only ocean basin where the rise might be misinterpreted as being monotonous WAS the North Atlantic, but that has shown a significant drop since 2005.
You continued, “Finally, in order to determine total ocean heat content, it is prudent to go as deeply as possible. There can be an interchange of heat content between the depths below 700 meters and the region above 700M that would affect the total heat content in the region between 700M and the surface.”
Unfortunately, there’s little long-term OHC data at depths below 700 meters. But as shown in the posts I linked for you, those varaitions can be explained with natural variables.”
Whether you look at 2000M or 700M, the earth’s oceans as a whole have been gaining heat since 1970. This heat gain varied from location to location, but despite the variation from tree to tree, the forest has gained heat.
You ascribe the increase in ocean heat content to natural factors. The only “natural factor” you mention is ocean surface temperature. The problem with this explanation is that ocean temperatures, to first order are a consequence of an increase in ocean heating, rather than a cause of ocean heating. If a hotter surface temperature due to El Nino or other ocean oscillations is the cause of ocean heating, due to a second order effect, such as reduced cloud cover allowing more sunlight to enter the ocean, or some other effect, you haven’t made a case for it in your article.
Since the primary source of ocean heat is the sun, and the sun has not gained in intensity since 1970, and volcanic action has not reduced markedly, these cannot explain the overall increase in ocean heating.
George E. Smith says:
“CO2 cannot (get involved in latent heat transaction) since it remains immutable in the atmosphere for thousands of years.”
Well actually.
It does get involved indirectly. It increases the energy in the system to increase evaporation and speed up or intensify the water cycle.
However the water cycle negates the effect of more CO2 by just accelerating the extra energy to space for a zero effect on the equilibrium temperature of the bulk ocean.
I recently provided elsewhere a clearer explanation as to why the extra downward IR from more CO2 is unable to affect the equilibrium temperature of the bulk ocean and I’ll repeat it here:
“I think it would be fruitful to look very closely at the interface between SST(int) and SST(skin).
For definitions see here:
http://ghrsst-pp.metoffice.com/pages/sst_definitions/
It is necessary to get a clear idea as to exactly why the higher temperature of SST(skin) fails to slow down the rate of energy flow from the subskin below.
A. The Default situation
i) Evaporation occurs primarily because of pressure and density differentials between water and air. Thus it will occur even if both water and air are at the same temperature. The process of evaporation is not dependent on any temperature differential. There are other influences that will increase or decrease the rate of evaporation but they need not concern us here.
ii) At Earth’s atmospheric pressure the energy required to provoke evaporation is always less than the energy taken from the local environment when evaporation occurs so we need to analyse exactly where the deficit can be provided from.
iii) In the absence of DLR it is taken from the water below because the water is generally warmer than the air hence the development of a layer of cool water 1mm deep and 0.3C cooler than the ocean bulk below.
B. When DLR is added to the mix.
i) DLR in itself does nothing. Before it can warm anything it must be absorbed by a water molecule.
ii) When DLR impacts the water surface some molecules will evaporate immediately and others will need to wait a moment to acquire enough additional energy.
iii) Those which are in the process of evaporating form SST(int). Those which are busily acquiring energy form SST(skin). The molecules in SST(skin) steadily gain more energy and move upward towards SST(int). In the process they gain more energy and become warmer with sensible energy that registers on our sensors.
C. The response to DLR once evaporation from DLR begins.
i) The molecules in SST(int) evaporate producing a local energy deficit. The energy most readily available is in the nearest molecules of SST(skin) so a flow of energy is set up from SST(skin) to SST(int)
ii) That energy flow is upward so the additional energy being supplied to the molecules in SST(skin) cannot flow downward to increase the temperature of the subskin.
iii) We then have both energy AND individual molecules moving upwards towards SST(int)
iv) The DLR cannot penetrate beyond SST(skin) so ALL the DLR gets absorbed by molecules in that region and ALL those molecules in due course find their way to SS(int). Thus there is no surplus energy from DLR left over to warm the subskin and even if there were it is flowing in the wrong direction.
v) Meanwhile remember that there is a net deficit to deal with when evaporation occurs. If ALL the DLR is now in molecules that are going to move upward and evaporate it can only be provided by a cascade of energy from molecule to molecule up through SST(skin).
vi) But at the bottom of that cascade where SST(skin) has it’s interface with the subskin there is still going to be that deficit. That remaining deficit must be accounted for and it already has been catered for by the pre-existing upward flow of energy from the ocean bulk below which is always present even in the absence of DLR
vii) Additionally that energy is already of the correct quantity to make up the deficit because the DLR is ALL accounted for in the process of accelerated evaporation leaving the background equilibrium undisturbed.
Thus DLR in any quantity or from any source cannot alter the equilibrium temperature of the oceans.
Now if there is a flaw in any of that then someone please tell me now.
eadler says: “You ascribe the increase in ocean heat content to natural factors. The only ‘natural factor’ you mention is ocean surface temperature. The problem with this explanation is that ocean temperatures, to first order are a consequence of an increase in ocean heating, rather than a cause of ocean heating.”
The three posts I linked in the text of this post, and that I linked a second time for you in my January 10, 2011 at 5:23 pm reply to you, were Ocean Heat Content posts. They were not about “ocean surface temperature” . The units are Gigjoules per sqaure meter (GJ/m^2) not in deg C. The NINO3.4 SST anomalies you see in the graphs (as explained in those posts) are there as references for timing, nothing more.
Let’s try again. Here are the links:
http://bobtisdale.blogspot.com/2009/09/enso-dominates-nodc-ocean-heat-content.html
AND:
http://bobtisdale.blogspot.com/2009/12/north-pacific-ocean-heat-content-shift.html
AND:
http://bobtisdale.blogspot.com/2009/10/north-atlantic-ocean-heat-content-0-700.html
DCA engineer says: “I’ve seen concern for the use of “eye-balled”. Can you explain or qualify this?”
The scaling and lag were established by appearance of the two variables, using the larger(est) event (the 1997/98 El Nino, the 1991 Mount Pinatubo eruption, etc.) as reference, not by statistical methods using all events. The data indicates that the larger events (such as the 1997/98 El Nino) are strong enough to overcome the noise that can mask the global response to lesser events. In other words, I used the larger events as reference because the response to them was clearest.
eadler says:
January 11, 2011 at 11:08 am
“Since the primary source of ocean heat is the sun, and the sun has not gained in intensity since 1970, and volcanic action has not reduced markedly, these cannot explain the overall increase in ocean heating.”
Strawman argument that is as bad as saying the tiny atmospheric amount of CO2 has no contribution on the greenhouse effect. The sun doesn’t need to gain intensity because it already has in the recent past, just needs to maintain it. This is the problem with people failing to understand the difference between temperature and heat. The temperature can stay the same, but the energy increases or decreases. The oceans rely on SWR at the surface, what occurs measured in the upper troposphere has little bearing compared with how much is reflected back to space from clouds and snow/ice. Albedo from medium/low level clouds warms or cools the ocean surface by increasing or decreasing over time across the global surface. This albedo has been changing over recent decades and is responsible for most ocean warming with the increased maintained higher solar levels.
Lets consider LWR (long wave radiation) and SWR (short wave radiation) with LWR not being able to warm a small volume of water and especially not the deep ocean when compared to SWR. The difference in the role between LWR and SWR on a volume of the ocean is like comparing a teaspoon of water with a swimming pool.
The swimming pool can change temperatures up to 2c and back down again, how many teaspoons of water at 0.1c warmer added are going to require to increase the swimming pool 1c? There are so many needed that when this amount is achieved, far more energy will have already been lost to the atmosphere with the natural day and night cycle of the sun with evaporative cooling. Now we have a comparable estimate of how LWR is compared to SWR in climate. The oceans don’t freeze because the SWR controls the energy in it and LWR has no measurable affect compared with it’s many orders times greater brother.
Chris Brown says: “Have you not just correlated temperature with temperature?”
I have adjusted hemispheric surface temperature datasets based on volcanic aerosols, and on a temperature-based ENSO index. Both volcanoes and ENSO events are known to cause variations in global temperatures. I then argued and illustrated that there are secondary effects of ENSO that are not accounted for using the temperature-based ENSO index. I compared the ENSO and volcano-adjusted data to the SST anomalies of two parts of the western Pacific (the Kuroshio-Oyashio Extension and the SPCZ Extension) where the secondary effects of ENSO appear strongest. Originally, that was where I ended the post. The agreements between the adjusted hemispheric temperature data and the respective KOE and SPCZ datasets were very strong and would have made for a great post. (I would have ended with Figures 13 and 22 and those were good matches between datasets.)
What has some people concerned (including yourself apparently) is I elected to take the post a few steps farther. I took the hemispheric surface temperature data that were adjusted for ENSO and volcanoes and further adjusted them with the datasets that were intended to illustrate the secondary effects of ENSO. It is the secondary effects that are presently not accounted for in climate studies like the one referenced in the post, Thompson et al. These secondary effects are the unknowns, and this is why the first paragraph of the post includes the following sentence: “But this basic evaluation indicates these secondary effects of ENSO require further research.”
Bob Tisdale
You wrote January 11, 2011 at 2:36 am :
“…while I will agree that the corrections GISS makes always seem to contribute to the trend, you also have to keep in mind that they do document the changes…”
Documentation of changes does no necessarily mean correctness. Bear in mind that you limited your own study to latitudes +/- 60 degrees, apparently because you had greater confidence in GISS for this region.
What sticks out like dog’s balls, is that there is a distinct disparity between GIStemp, and the other big 3 for the widely described 1998 “super El Nino“. Whilst I don’t have great confidence in Hadcrut3, it’s 1998 does line-up rather well with UAH & RSS. The satellite data picks-up stratospherically, what is a surface driven anomaly rather markedly.
The big 4 time-series plus NCDC global records are conveniently overlaid here with common baselines:
http://www.climate4you.com/images/AllCompared%20GlobalMonthlyTempSince1979.gif
The other big 3, all show a distinct plateau commencing at, I would say; 1998, variously described as a cooling phase or lack of warming. Even RealClimate recognised this in an article some time back. (although it was rapidly closed after 3 days of comments… whoops).
If you like to draw step-changes, then one could argue a step change at 1998, on the basis of the combination graph cited above.
The effect of GISS depressing the 1998 value (some time after 1999, and raising 2005 etc), is to show a lessening of the plateau, or if you like a more remorseless global warming.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Bob, did you see my January 10, 2011 at 10:35 pm ?
It proposes a less complicated and/or supportive argument.
I see that eadler has some objections to it, and has even quoted the oracular mount of RC. I may have time to respond today
Bob Tisdale
You wrote January 11, 2011 at 2:36 am :
“…while I will agree that the corrections GISS makes always seem to contribute to the trend, you also have to keep in mind that they do document the changes…”
Documentation of changes does no necessarily mean correctness. Bear in mind that you limited your own study to latitudes +/- 60 degrees, apparently because you had greater confidence in GISS for this region.
What sticks out like dog’s balls, is that there is a distinct disparity between GIStemp, and the other big 3 for the widely described 1998 “super El Nino“. Whilst I don’t have great confidence in Hadcrut3, it’s 1998 does line-up rather well with UAH & RSS. The satellite data picks-up stratospherically, what is a surface driven anomaly rather markedly.
The big 4 time-series plus NCDC global records are conveniently overlaid here with common baselines:
http://www.climate4you.com/images/AllCompared%20GlobalMonthlyTempSince1979.gif
The other big 3, all show a distinct plateau commencing at, I would say; 1998, variously described as a cooling phase or lack of warming. Even RealClimate recognised this in an article some time back. (although it was rapidly closed after 3 days of comments… whoops).
If you like to draw step-changes, then one could argue a step change at 1998, on the basis of the combination graph cited above.
The effect of GISS depressing the 1998 value (some time after 1999, and raising 2005 etc), is to show a lessening of the plateau, or if you like a more remorseless global warming.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Bob, did you see my January 10, 2011 at 10:35 pm ?
It proposes a less complicated and/or supportive argument.
I see that eadler has some objections to it, and has even quoted the oracular mount of RC. I may have time to respond today
eadler says:
January 11, 2011 at 10:45 am
LWR has to be millions of times greater then demostrated to beable to complete with any tiny variable change in the solar output. It has a miscule effect and was used by realclimate as a get out of jail clause because they knew the oceans dominate atmospheric global temperatures. They have not provided evidence the LWR is changing the water below the surface one bit. SWR easily explains the changes when this penetrates down to 100m depth compared with 0.5mm while cloud albedo had decreased.
Lets look at this energy wise, even if LWR and SWR were equal (which they are far from not) SWR would still be around 200,000 times greater. Vary the LWR from CO2 another 50 percent higher and this energy contribution is tiny (orders smaller) to just a 0.01 percent change from SWR. I know the skin idea is wrong because it is demonstrated with numbers being too big to make any difference. The change is so small compared to SWR that any tiny build up over many years is easily lost to space during the same period. The energy lost to space is much bigger, but we have stable temperatures because this compares well with SWR. If you still don’t believe this then try this experiment at home and compare the same volumes of water at same temperatures to begin with, one exposed to the sun for the day and one only exposed to LWR shaded from the sun. Then measure the temperature changes at the end and record any differences.
Bob_FJ says: “Bob, did you see my January 10, 2011 at 10:35 pm ?”
Yup. I already replied to the first part of it, when I replied to the correction.
You continued, “It proposes a less complicated and/or supportive argument.”
I’m not sure how they make the same arguments I made in this post. My primary discussions were that NINO3.4 SST anomalies do not capture the process of ENSO and that there are secondary effects of ENSO that could explain much of the rise in surface temperature since 1982.
Matt G says:
January 11, 2011 at 2:04 pm
eadler says:
January 11, 2011 at 11:08 am
“Since the primary source of ocean heat is the sun, and the sun has not gained in intensity since 1970, and volcanic action has not reduced markedly, these cannot explain the overall increase in ocean heating.”
Strawman argument that is as bad as saying the tiny atmospheric amount of CO2 has no contribution on the greenhouse effect. The sun doesn’t need to gain intensity because it already has in the recent past, just needs to maintain it. This is the problem with people failing to understand the difference between temperature and heat. The temperature can stay the same, but the energy increases or decreases. The oceans rely on SWR at the surface, what occurs measured in the upper troposphere has little bearing compared with how much is reflected back to space from clouds and snow/ice. Albedo from medium/low level clouds warms or cools the ocean surface by increasing or decreasing over time across the global surface. This albedo has been changing over recent decades and is responsible for most ocean warming with the increased maintained higher solar levels.”
I will accept the verdict of climate scientists who are experts on the sun. They say that the sun is not responsible for the warming we have seen since 1970.
In fact solar radiation has declined since 1960 while global warming has accelerated. Check out the graph on the following web page.
http://www.skepticalscience.com/solar-activity-sunspots-global-warming-intermediate.htm
Eighteen peer reviewed papers are quoted on the above page, concluding that solar radiation was not a significant contributor to the warming observed since 1970.
You continue,
“Lets consider LWR (long wave radiation) and SWR (short wave radiation) with LWR not being able to warm a small volume of water and especially not the deep ocean when compared to SWR. The difference in the role between LWR and SWR on a volume of the ocean is like comparing a teaspoon of water with a swimming pool.
The swimming pool can change temperatures up to 2c and back down again, how many teaspoons of water at 0.1c warmer added are going to require to increase the swimming pool 1c? There are so many needed that when this amount is achieved, far more energy will have already been lost to the atmosphere with the natural day and night cycle of the sun with evaporative cooling. Now we have a comparable estimate of how LWR is compared to SWR in climate. The oceans don’t freeze because the SWR controls the energy in it and LWR has no measurable affect compared with it’s many orders times greater brother.”
Sorry but your analogy is wrong. On average the LWR from the atmosphere incident on the ocean surface has a flux of about 324W/M^2, and is TWICE the energy flux of the incoming solar radiation which is absorbed by the ocean , only168W/M^2. Without this LWR, the earth’s average surface temperature would be 33 degrees cooler.
http://www.google.com/imgres?imgurl=http://www.windows2universe.org/earth/Atmosphere/images/earth_rad_budget_kiehl_trenberth_1997_big.gif&imgrefurl=http://www.windows2universe.org/earth/climate/warming_clouds_albedo_feedback.html&h=456&w=664&sz=40&tbnid=khMQlXF5TEG-vM:&tbnh=95&tbnw=138&prev=/images%3Fq%3Dtrenberth%2Bearths%2Benergy%2Bflux%2Bdiagram&zoom=1&q=trenberth+earths+energy+flux+diagram&hl=en&usg=__a9nDZ3klcoNaBuYYmmKnFmQuE0w=&sa=X&ei=5vAsTYCRD4K8lQeu3fC6Cg&ved=0CBsQ9QEwAA
The LWR warms the ocean’s surface skin, and slows the escape of the solar radiation absorbed by the ocean bulk, by reducing the convection of heat to the ocean surface.
The earth’s surface is cooled by the upward LWR, evaporation and convection.
Bob Tisdale says:
January 11, 2011 at 12:51 pm
eadler says: “You ascribe the increase in ocean heat content to natural factors. The only ‘natural factor’ you mention is ocean surface temperature. The problem with this explanation is that ocean temperatures, to first order are a consequence of an increase in ocean heating, rather than a cause of ocean heating.”
The three posts I linked in the text of this post, and that I linked a second time for you in my January 10, 2011 at 5:23 pm reply to you, were Ocean Heat Content posts. They were not about “ocean surface temperature” . The units are Gigjoules per sqaure meter (GJ/m^2) not in deg C. The NINO3.4 SST anomalies you see in the graphs (as explained in those posts) are there as references for timing, nothing more.
Let’s try again. Here are the links:
http://bobtisdale.blogspot.com/2009/09/enso-dominates-nodc-ocean-heat-content.html
AND:
http://bobtisdale.blogspot.com/2009/12/north-pacific-ocean-heat-content-shift.html
AND:
http://bobtisdale.blogspot.com/2009/10/north-atlantic-ocean-heat-content-0-700.html”
I read those links. They show graphs regarding variation in the heating of different regions of the earth’s oceans. They say nothing about the source of the heating although they correlate it with El Nino and other ocean oscillatory phenomena, which are related to periodic sea surface temperature variations. As I have mentioned, high ocean surface temperatures are a ,result of ocean heating, not a source of ocean heating.
The oceans are heating either because they are gaining additional heat from the sun, downward radiation from the atmosphere, or are losing heat more slowly through radiation, evaporation or convection to the air. You haven’t presented any evidence that bears on these sources of ocean heating in the 3 links you have provided. Breaking the ocean up in sections doesn’t shed any light on the sources of ocean heating.
eadler says:
January 11, 2011 at 4:56 pm
Sorry what you mentioned doesn’t say this is wrong and not read the post properly either for the solar explanation or ocean one. Why have you decided to ignore albedo and SWR at the surface and gone straight to solar levels measured above the cloud levels? The figures you have given of LWR and SWR are not comparable because one doesn’t penetrate the surface and the other penetrates 100m deep. With you’re same flawed logic you are admitting the greenhouse effect of 33c has nearly twice more energy then for the entire source of the sun. The sun is responsible for 255c not around 16c and therefore the planet is around 288k not 49k. I’m am in disbelief that you think LWR warms the planet more than the sun.
Stephen Wilde says:
January 11, 2011 at 11:10 am
“George E. Smith says:
“CO2 cannot (get involved in latent heat transaction) since it remains immutable in the atmosphere for thousands of years.”
Well actually.
It does get involved indirectly. It increases the energy in the system to increase evaporation and speed up or intensify the water cycle.
However the water cycle negates the effect of more CO2 by just accelerating the extra energy to space for a zero effect on the equilibrium temperature of the bulk ocean.
I recently provided elsewhere a clearer explanation as to why the extra downward IR from more CO2 is unable to affect the equilibrium temperature of the bulk ocean and I’ll repeat it here:
“I think it would be fruitful to look very closely at the interface between SST(int) and SST(skin).
For definitions see here:
http://ghrsst-pp.metoffice.com/pages/sst_definitions/
It is necessary to get a clear idea as to exactly why the higher temperature of SST(skin) fails to slow down the rate of energy flow from the subskin below.
A. The Default situation
i) Evaporation occurs primarily because of pressure and density differentials between water and air. Thus it will occur even if both water and air are at the same temperature. The process of evaporation is not dependent on any temperature differential. There are other influences that will increase or decrease the rate of evaporation but they need not concern us here.
ii) At Earth’s atmospheric pressure the energy required to provoke evaporation is always less than the energy taken from the local environment when evaporation occurs so we need to analyse exactly where the deficit can be provided from.
iii) In the absence of DLR it is taken from the water below because the water is generally warmer than the air hence the development of a layer of cool water 1mm deep and 0.3C cooler than the ocean bulk below.
B. When DLR is added to the mix.
i) DLR in itself does nothing. Before it can warm anything it must be absorbed by a water molecule.
ii) When DLR impacts the water surface some molecules will evaporate immediately and others will need to wait a moment to acquire enough additional energy.
iii) Those which are in the process of evaporating form SST(int). Those which are busily acquiring energy form SST(skin). The molecules in SST(skin) steadily gain more energy and move upward towards SST(int). In the process they gain more energy and become warmer with sensible energy that registers on our sensors.
C. The response to DLR once evaporation from DLR begins.
i) The molecules in SST(int) evaporate producing a local energy deficit. The energy most readily available is in the nearest molecules of SST(skin) so a flow of energy is set up from SST(skin) to SST(int)
ii) That energy flow is upward so the additional energy being supplied to the molecules in SST(skin) cannot flow downward to increase the temperature of the subskin.
iii) We then have both energy AND individual molecules moving upwards towards SST(int)
iv) The DLR cannot penetrate beyond SST(skin) so ALL the DLR gets absorbed by molecules in that region and ALL those molecules in due course find their way to SS(int). Thus there is no surplus energy from DLR left over to warm the subskin and even if there were it is flowing in the wrong direction.
v) Meanwhile remember that there is a net deficit to deal with when evaporation occurs. If ALL the DLR is now in molecules that are going to move upward and evaporate it can only be provided by a cascade of energy from molecule to molecule up through SST(skin).
vi) But at the bottom of that cascade where SST(skin) has it’s interface with the subskin there is still going to be that deficit. That remaining deficit must be accounted for and it already has been catered for by the pre-existing upward flow of energy from the ocean bulk below which is always present even in the absence of DLR
vii) Additionally that energy is already of the correct quantity to make up the deficit because the DLR is ALL accounted for in the process of accelerated evaporation leaving the background equilibrium undisturbed.
Thus DLR in any quantity or from any source cannot alter the equilibrium temperature of the oceans.
Now if there is a flaw in any of that then someone please tell me now.”
You must have missed the post I made above, where I showed the flaw in your thinking.
Because of the heating of the ocean depths by the sunlight , which penetrates a number of meters , the depths are warmer than the surface skin, because radiation, evaporation and conduction cause the surface skin to lose heat.
The DLR heats the surface skin. This suppresses the upward convection of the heat absorbed from the sun by the ocean depths below the surface. With reduced upward convection from below, more of the sun’s heat will remain in the ocean. If the DLR decreases, the temperature gradient between the surface skin and bulk increases, and more heat flows from the ocean depths to the surface where it is radiated away. If DLR increases, less heat flows from the depth to the surface. So DLR can have a significant effect on how much of the sun’s short waver energy is retained by the ocean.
http://www.realclimate.org/index.php/archives/2006/09/why-greenhouse-gases-heat-the-ocean/
Your explanation left out this important mechanism, so your conclusion that DLR cannot affect ocean heat is wrong.
Bob Tisdale, in response to your January 11, 2011 at 4:00 pm
You seem to have missed items 2),3),& 4) in my proposal, which in full was as follows:
1) The Hadcrut3, RSS, and UAH global data, all show that 1998 was the hottest year on record, globally. GIStemp also showed it as the hottest year back in 1999, before they since made highly questionable and big changes to the whole time-series.
(See previous WUWT http://wattsupwiththat.com/2010/05/18/gistemp-vs-hadcrut/)
2) There is little doubt that the 1997-1998 El Nino was the main driver for the 1998 global high average and that it was big. (That is without necessarily understanding why it is so global and rapid in effect.)
3) However, what can be said is that what predominantly heats the oceans is sunlight. On the other hand, GHG induced back radiation has little effect because being long-wave, it is extinguished in the skin of the water, is able to re-radiate virtually instantly, and can only penetrate via rapid mechanical mixing.
4) Thus the 1998 super El Nino induced global warming was a secondary effect of short-wave ocean heating, not necessarily recent, and had very little to do with GHG.
In a later comment, I wrote:
It proposes a less complicated and/or supportive argument.
This referred to the above, as perhaps being supportive of your study. (not to the ENSO 3.4 as published by the Oz BOM as you seem to have reacted.)
The point I was making was that this version of ENSO index has what I think is an astonishing correlation with Roy Spencer’s AMSR-E GLOBAL SST’s, as compared here:
http://farm6.static.flickr.com/5010/5328413916_5d953c3ded_z.jpg
I find it hard to wrap my head around a global lag of only 3 months, let alone the matching shape of the curves!
Bob says:As you can see in Figures 11 through 13, there were ENSO residuals left that were approximately the same magnitude as the reference ENSO signal.
Wow, you’re not kidding! Thanks for the explanation.
Bob, notably absent from your article was a direct comparison of KOE & GSE with -N34. In the following paper, note the lower correlation of NPGO with SPCZE (vs. with KOE) – see figure 4b:
Di Lorenzo, E.; Schneider, N.; Cobb, K.M.; Franks, P.J.S.; Chhak, K.; Miller, A.J.; McWilliams, J.C.; Bograd, S.J.; Arango, H.; Curchitser, E.; Powell, T.M.; & Riviere, P. (2008). North Pacific Gyre Oscillation links ocean climate and ecosystem change. Geophysical Research Letters 35, L08607. doi:10.1029/2007GL032838.
http://horizon.ucsd.edu/miller/download/NPGO/NPGO.pdf
NPGO might be a good place to start when exploring the mismatches (of KOE & GSE with -N34), which you’ve no doubt noted for the early 90s. Given your knowledge of El Nino Modoki, I’ve little doubt you’ll have pieced this all together by the time you finish reading this sentence (if you hadn’t already).
Thanks for drawing our attention to SPCZE & GSE. Your posts usually stimulate cross-referencing efforts that lead to dot connection. This time around has been no exception.
Best Regards.
—
Related:
Everyone should consider the possibility that for some indices, North Atlantic correlations (in correlation maps) are depressed by the North Atlantic’s sensitivity (being the smaller northern basin surrounded by a lot of land/ice, resulting “higher continentality”), which gives it a propensity towards high amplitude oscillations, including decadal-timescale nonlinear trends.
Nonlinearly removing the multidecadal component of North Atlantic SST and repeating analyses is one avenue towards multiscale insight. When nonlinearly detrending, I recommend against reliance on assumed functional forms; filters that allow the data to speak for themselves (such as repeat narrow-band smoothing with iterative end-correction) have superior utility for many purposes. An alternative (to nonlinear detrending) approach is working with derivatives.
–
Michael D Smith seems to be thinking about similar data exploration issues — interesting comments Michael – and a thought-provoking response from Bob on the superiority of the eyeballed adjustment in this particular case – very sensible rationale.
—
For those interested:
NPGO data:
http://www.o3d.org/npgo/data/NPGO.txt
Related Info:
http://www.ocean3d.org/npgo
Bob_FJ says: “4) Thus the 1998 super El Nino induced global warming was a secondary effect of short-wave ocean heating, not necessarily recent, and had very little to do with GHG.”
The 1997/98 El Nino was fueled by the warm waters created by the 1995/96 La Nina. Discussed that and provided a link to a McPhaden (1999) paper “Genesis and Evolution of the 1997-98 El Niño” in the following post:
http://bobtisdale.blogspot.com/2010/02/la-nina-underappreciated-portion-of.html
Paul Vaughan says: “Bob, notably absent from your article was a direct comparison of KOE & GSE with -N34.”
Sorry. Due to the length of this post, I didn’t include many comparisons. Refer to Figure 8…
http://i52.tinypic.com/wjvow.jpg
…from the earlier post on the KOE:
http://bobtisdale.blogspot.com/2010/12/enso-related-variations-in-kuroshio.html
That’s a couple more graphs for the follow-up post.
eadler says: “I read those links. They show graphs regarding variation in the heating of different regions of the earth’s oceans. They say nothing about the source of the heating although they correlate it with El Nino and other ocean oscillatory phenomena, which are related to periodic sea surface temperature variations. As I have mentioned, high ocean surface temperatures are a ,result of ocean heating, not a source of ocean heating.”
It appears I must write a post for those, like yourself, who are skeptical of skeptics, and provide a more detailed analysis of the rise in OHC.
And I will disagree with your final sentence, which read, “As I have mentioned, high ocean surface temperatures are a ,result of ocean heating, not a source of ocean heating.”
Sea surface temperature is a portion (component) of ocean heat content.
A couple of new charts.
The Daily UAH satellite temps over 2010 – Global, NH, SH, Tropics – (these are not the official numbers as there is some processing required but close enough).
Some variation but it has been a pretty consistent downtrend throughout the year. Let’s say maybe there is a NH bump in the late summer as the Kuroshio was peaking.
http://img831.imageshack.us/img831/4550/dailyuahtemps2010.png
And then the water vapour numbers for December were released (NCEP Reanalysis) and the levels are dropping rapidly now about as expected. More to come yet.
http://img251.imageshack.us/img251/1456/ensovstcwvdec10.png