On Muller et al (2013) “Decadal variations in the global atmospheric land temperatures”

I received an email yesterday morning advising me that Muller et al (2013) had been published. (Thanks, Marc.) The title of the paper is “Decadal variations in the global atmospheric land temperatures”. The abstract is here and a preprint version of the paper is available from the Berkeley Earth Surface Temperature website here. The primary finding of the paper is that land surface temperature anomalies are more closely correlated with the Atlantic Multidecadal Oscillation (AMO) than they are to NINO3.4 sea surface temperature anomalies (as a proxy for El Niños and La Niñas or ENSO) and the Pacific Decadal Oscillation (PDO) index. We’ll discuss the very obvious reasons for this.

Muller et al also briefly discussed a couple sea level pressure-based indices. They will not be discussed in this post.

We’ll discuss why Muller should have included detrended North Pacific sea surface temperatures instead of the PDO in their comparisons with the AMO data, and we’ll use correlation maps to help show what the PDO represents—and what it doesn’t represent.

ON THE CORRELATION OF LAND SURFACE TEMPERATURES WITH THE AMO

The abstract reads:

Interannual to decadal variations in Earth global temperature estimates have often been identified with El Niño Southern Oscillation (ENSO) events. However, we show that variability on time scales of 2–15 years in mean annual global land surface temperature anomalies Tavg are more closely correlated with variability in sea surface temperatures in the North Atlantic. In particular, the cross-correlation of annually averaged values of Tavg with annual values of the Atlantic Multidecadal Oscillation (AMO) index is much stronger than that of Tavg with ENSO. The pattern of fluctuations in Tavg from 1950 to 2010 reflects true climate variability and is not an artifact of station sampling. A world map of temperature correlations shows that the association with AMO is broadly distributed and unidirectional. The effect of El Niño on temperature is locally stronger, but can be of either sign, leading to less impact on the global average. We identify one strong narrow spectral peak in the AMO at period 9.1±0.4 years and p value of 1.7% (confidence level, 98.3%). Variations in the flow of the Atlantic meridional overturning circulation may be responsible for some of the 2–15 year variability observed in global land temperatures.

My Figure 1 is Figure 3 from Muller et al (2013). We’ll concentrate on it for a few moments.

Figure 1 - Figure 3 from Muller et al 2013

Figure 1

The caption for their Figure 3 (from the preprint) reads:

Figure 3. Decadal fluctuations in surface land temperature estimates and in oceanic indices. The long-term variability was suppressed by removing the least-squares fit 5th order polynomial from each curve. (A) shows the 12-month smoothed land surface temperature estimates from the four groups. The decadal variations are very similar to each other. The Berkeley Earth data were derived from 2000 sites chosen randomly from a set of 30964 that did not include any of the sites from the other groups. (B) shows the AMO index compared to Tavg, the average of the four land estimates. (C) shows the ENSO index compared to the average of the four land estimates. (D) shows the AMO and ENSO directly. Note that the AMO agreement in (b) is qualitatively stronger than the ENSO agreement in (c).

When looking at their Figure 3, keep in mind that the data have been modified. Muller et al “suppressed” the “long-term variability” by subtracting the values of a 5th order polynomial curve from the data. See the example of the AMO data from the ESRL and the resulting 5th order polynomial curve here. They then smoothed the remainders with a 12-month filters for their Figure 3.

It should be very clear that El Niño and La Niña events (ENSO) are responsible for many of the year-to-year variations in the Average Land Surface Air Temperature data (Tave in cell C) and in the Atlantic Multidecadal Oscillation data (AMO in cell D). The reason: ENSO is the dominant mode of natural variability on annual and interannual timescales. Muller et al (2013) in fact note that the AMO signal lags the ENSO signal:

For reference, the maximum correlation between AMO and ENSO in these data is 0.50 ± 0.04; with AMO lagging ENSO by 0.70 ± 0.25 years. This is a somewhat larger lag than previously reported in a more detailed analysis of ENSO by Trenberth et al. [2002].

We can see in cell B that the annual variations in the AMO and global land surface temperatures are remarkably similar. And we can see that the yearly changes in the land surface air temperatures (cell C) and the AMO (cell D) do not correlate as well with the ENSO signal.

Why?

Let’s examine the period from the mid-1980s to the early 2000s in cells C and D. The reasons for the differences show up quite well during that period. We can see that the Tave and AMO signals do not cool proportionally to the ENSO signal during the 1988/89 and 1998-01 La Niña events. Also, the Tave and AMO both respond to the eruption of Mount Pinatubo in 1991 with a multiyear dip and rebound, while the ENSO signal does not—or shows little response to the eruption.

In other words, the reason the land surface air temperatures and AMO agree so well is that, in addition to responding to ENSO, they’re also responding similarly to volcanic aerosols and to ENSO residuals, the latter of which prevent the land surface air temperatures and the AMO from cooling proportionally during the 1988/89 and 1998-01 La Niñas.

Nothing magical there. And it introduces an interesting question for other papers and blog posts?

Many papers and blog posts attempt to remove the impacts of natural variables from the global surface temperature record. When they add AMO data to the ENSO, volcanic aerosol and solar data in their multiple regression analyses, do they recognize and account for the fact that the AMO data and global land surface air temperature data have similar responses to ENSO and volcanic aerosols?

MULLER ET AL COMPARED APPLES AND ORANGES

A good portion of Muller et al (2013) deals with comparisons of global surface temperatures with the Atlantic Multidecadal Oscillation index (AMO) and with the Pacific Decadal Oscillation index (PDO), to emphasize their findings.

Muller et al (2013) failed to recognize that the AMO index data is detrended sea surface temperature anomalies of the North Atlantic, while the PDO index is NOT detrended sea surface temperature anomalies of the North Pacific. The PDO is an abstract form (a Picasso version, if you will) of the sea surface temperature anomalies of the North Pacific. So Muller et al (2013) were comparing apples to oranges.

Because I’ve discussed what the Pacific Decadal Oscillation (PDO) index is—and more importantly what it is not—in numerous posts, I’m not going to go into a detailed discussion here. If this topic is new to you, refer to the following posts (listed in order of most recent to earliest):

Yet Even More Discussions About The Pacific Decadal Oscillation (PDO)

An Inverse Relationship Between The PDO And North Pacific SST Anomaly Residuals

An Introduction To ENSO, AMO, and PDO — Part 3

But in scanning the preprint version of Muller et al (2013) I was struck with an idea of how to present differences between the PDO index and the North Pacific sea surface temperature anomalies. In their Figure 6, my Figure 2, Muller et al (2013) presented correlation maps of NCDC global surface temperature data to a couple of sea surface temperature-based indices. The left-hand map presents the correlation of the AMO data with global surface temperatures and the right hand map, ENSO (NINO3.4 sea surface temperature anomalies) with global surface temperatures. High positive correlations are in dark reds and high negative correlations are in dark blues. Muller et al (2013) noted in the caption the reason they included the correlation maps.

AMO is observed to have positive or neutral correlation almost everywhere, while ENSO shows both strong positive and negative correlations.

Figure 2 - Figure 6 from Muller et al 2013

Figure 2

And based on our earlier discussion, the reason the AMO has a positive or neutral correlation with global surface temperatures almost everywhere is, the AMO and global surface temperatures respond similarly to ENSO, ENSO residuals and volcanic aerosols. Simple.

Back to the idea I was talking about: The top two maps in my Figure 3 are correlation maps of global surface temperatures (NCDC data, same as Muller et al) with detrended North Atlantic sea surface temperature anomalies (the AMO) and NINO3.4 sea surface temperature anomalies (ENSO). I prepared the maps using the KNMI Climate Explorer. These correlation maps are similar to the maps presented by Muller et al. I’ve added the correlation map of global surface temperatures with detrended North Pacific (north of 20N) sea surface temperature anomalies as the lower left-hand map, and presented the PDO correlations in the lower right-hand map.

Figure 3

Figure 3

Full-sized version of Figure 3 is here.

Note that on the maps I’ve marked the locations of the sea surface areas of the respective datasets with very fine black boxes. Also note that I used ERSST.v3b sea surface temperature data for all but the PDO data. The ERSST.v3b dataset is the sea surface temperature component of the NCDC surface temperature dataset. On the other hand, Muller et al used the ESRL AMO data—it is based on Kaplan sea surface temperature data, which also includes Reynolds OI.v2 data over the last decade or so. I believe Muller et al used ERSST.v3b sea surface temperature data for their NINO3.4 data (ENSO index), but the preprint version of paper provides a link to the weekly Reynolds OI.v2-based ENSO data, which starts in 1991—and they could not have used it for the comparisons from 1950 to 2010.

If Muller et al (2013) had compared the AMO data (upper left-hand map) with a comparable dataset from the North Pacific (north of 20N) they would have used detrended North Pacific sea surface temperature anomalies (lower left-hand map), because the AMO index is detrended sea surface temperature data from the North Atlantic. Instead they used the PDO, which does not represent the sea surface temperatures of the North Pacific. Notice that there are no similarities between the two lower maps but they’re both derived from the same area of the North Pacific. The PDO index basically represents the El Niño- and La Niña-related spatial patterns in the sea surface temperature anomalies in the North Pacific—for example, warm in the east and cool in the central and western North Pacific (north of 20N) during an El Niño.

That’s why the ENSO (upper right-hand map) and PDO (lower right-hand map) correlation maps are so similar in the North Pacific. The PDO is a statistically created dataset that captures the spatial-pattern effects of El Niño and La Niña events on the North Pacific sea surface temperatures. That PDO spatial pattern is important for fishermen because it impacts where fish are located. The PDO spatial pattern is also important because it impacts rainfall patterns in the United States—we discussed that in the recent post here. But the PDO does not represent the sea surface temperatures of the North Pacific.

A couple of other notes:

The time-series data for the PDO index and the NINO3.4 sea surface temperatures are different for a very basic reason: the spatial pattern of the sea surface temperature anomalies in the North Pacific (warm in the east and cool in the central and western portions of the North Pacific during an El Niño and vice versa during a La Niña) are also impacted by the wind patterns (and interdependent sea level pressures) of the North Pacific, and those wind patterns and sea level pressures vary over time.

In the PDO map (lower right-hand map), note how the area of the central North Pacific east of Japan has the highest (though negative) correlation. That area is called the Kuroshio-Oyashio Extension or KOE. The variations in the sea surface temperatures in the KOE dominate the North Pacific data, and they are inversely related to the PDO index.

Note also in the lower left-hand map that the sea surface temperatures of the North Atlantic are correlated with the variations in the sea surface temperatures of the North Pacific. I discussed this in detail in the post The ENSO-Related Variations In Kuroshio-Oyashio Extension (KOE) SST Anomalies And Their Impact On Northern Hemisphere Temperatures.

AMO HAS LITTLE IMPACT ON U.S. TEMPERATURES?

Notice above in Figure 6 from Muller et al (2013), my Figure 2, that the surface air temperatures in the United States correlate poorly with the AMO data. They mention this in the paper:

Remarkably, neither AMO nor ENSO shows a strong correlation with the temperature in the United States, although ENSO reaches strongly up the west coast of the US.

Curiously, the correlation map for the AMO that I created at the KNMI Climate Explorer (the upper left-hand map in Figure 3), shows a moderate correlation between the AMO and U.S. surface temperatures (using the NCDC global surface temperature data). In Figure 4, I used the Berkeley Earth Surface Temperature (BEST) data in the correlation maps. The U.S. surface air temperatures based on the BEST data also correlate with the AMO data.

Figure 4

Figure 4

Full-sized version of Figure 4 is here.

WOULD USING DETRENDED NORTH PACIFIC SEA SURFACE TEMPERATURE DATA INSTEAD OF THE PDO DATA HAVE CHANGED THE RESULTS OF MULLER ET AL?

Nope. The AMO still has the strongest correlation with land surface air temperatures, because they both respond similarly to ENSO, ENSO residuals and volcanic aerosols.

But Muller et al could have saved themselves some time, since the PDO data does not in any way represent the sea surface temperatures of the North Pacific and there was, therefore, no reason to compare it to the AMO data.

ENSO RESIDUALS?

I mentioned ENSO residuals a couple of times in this discussion. Those residuals are basically the aftereffects of strong El Niño events, and those aftereffects are caused by the warm water that’s left over from those strong El Niños. My illustrated essay “The Manmade Global Warming Challenge” [42MB] provides an overview of the causes and impacts of those leftover warm waters. It includes links to animations of data, which confirm the existence and source of the ENSO residuals.

And, of course, if you’re very interested in learning more about the processes of El Niño and La Niña events, there’s my book Who Turned on the Heat? The free preview is available here. Who Turned on the Heat? is available in pdf form here for US$8.00.

FURTHER INFORMATION ABOUT THE AMO

A link to the NOAA FAQ webpage about the AMO is here. I provided a detailed introduction to the Atlantic Multidecadal Oscillation in my post here.

The short description: the Atlantic Multidecadal Oscillation is a mode of natural variability through which the sea surface temperatures of the North Atlantic can contribute additionally to or suppress the global warming of land surface air temperatures that are occurring in response to the warming of the rest of the global oceans. And as discussed in the “The Manmade Global Warming Challenge” [42MB], the ocean heat content data and satellite-era sea surface temperature data both indicate the oceans warmed naturally.

Refer also to the RealClimate glossary webpage about the Atlantic Multidecadal Oscillation here. There, they write:

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).

ADDITIONAL INFORMATION

An El Niño releases heat into the atmosphere, and surface temperatures around the globe in many places warm in response to the El Niño, and in other parts, they cool—with more locations warming than cooling, so the average global surface temperature warms in response to the El Niño. But the heat released into the atmosphere during the El Niño is not directly warming the surface in those remote locations. The surface temperatures outside of the tropical Pacific warm in response to changes in atmospheric circulation caused by the El Niño. The processes that cause those changes are discussed in minute detail in Trenberth et al (2002) Evolution of El Nino–Southern Oscillation and global atmospheric surface temperatures. Wang (2005) ENSO, Atlantic Climate Variability, And The Walker And Hadley Circulation discusses why the tropical North Atlantic warms in response to an El Niño.

CLOSING

Compared to a number of other sea surface temperature-based indices (and sea level pressure-based indices, which we didn’t discuss in this post), Muller et al (2013) found that global land surface temperatures correlate best with the Atlantic Multidecadal Oscillation. We illustrated and discussed the reason for this—the AMO data and land surface air temperatures respond similarly to ENSO, ENSO residuals, and volcanic aerosols.

We also discussed and illustrated why Muller et al (2013) should have used detrended North Pacific sea surface temperatures instead the PDO data for a proper comparison to the AMO.

And we used correlation maps to show the differences between the PDO and the sea surface temperature anomalies of the North Pacific. We also used the correlation maps of the PDO and ENSO with global temperature anomalies to help explain what the PDO represents.

Advertisements

  Subscribe  
newest oldest most voted
Notify of
Edim

There’s a global multidecadal oscillation and it can be seen in all global and non-global temperature indices. AMO is just a regional (north Atlantic SST) manifestation.
http://www.woodfortrees.org/plot/esrl-amo/plot/esrl-amo/trend/plot/hadcrut4gl/detrend:0.76/plot/hadcrut4gl/trend/detrend:0.76
One can use any global, hemispheric, land or sea temperature index and find the same oscillation.

sailboarder

Nice critique. I am amazed though that Mullers co-authors have not been reading your posts here on WUWT and thus did not see to use the detrended North Pacific sea surface tempertatures.(instead of ENSO)

johnmarshall

Thank you Bob for another good piece of analysis.
Reality can be a damn problem.

Paul Vaughan

“We’ll discuss why Muller should have included detrended North Pacific sea surface temperatures instead of the PDO in their comparisons with the AMO data, and we’ll use correlation maps to help show what the PDO represents—and what it doesn’t represent.”
It’s like pushing water uphill Bob. You inform repeatedly. People ignore repeatedly.
This has gone on for years — that’s too long. My patience with those who can’t or won’t get it has expired. I’m permanently writing them off as dumber-than-a-post. Under the bus they go…

Stephen Wilde

“Before it is safe to attribute a global warming or a global cooling effect to any other factor (CO2 in particular) it is necessary to disentangle the simultaneous overlapping positive and negative effects of solar variation, PDO/ENSO and the other oceanic cycles. Sometimes they work in unison, sometimes they work against each other and until a formula has been developed to work in a majority of situations all our guesses about climate change must come to nought.
So, to be able to monitor and predict changes in global temperature we need more than information about the past, current and expected future level of solar activity.
We also need to identify all the separate oceanic cycles around the globe and ascertain both the current state of their respective warming or cooling modes and, moreover, the intensity of each, both at the time of measurement and in the future.”
from here:
http://climaterealists.com/index.php?id=1302&linkbox=true&position=10
“The Real Link Between Solar Energy, Ocean Cycles and Global Temperature”
Wednesday, May 21st 2008, 8:20 AM EDT

Bill Illis

The AMO is indeed a very important natural climate cycle driver.
Have a look at the monthly Raw AMO index versus Hadcrut4 back to 1856 (The Raw Index is not detrended and is not smoothed – I sometimes use this metric just to show how similar monthly temperatures are to the AMO).
http://s18.postimg.org/9uar3ow0p/Hadcrut4_vs_Raw_AMO.png
Now on a scattergram, not 100% close but the R^2 is 0.48 which on its own explains more of the temperature variability on a monthly basis than any other index).
http://s9.postimg.org/yi12i63z3/Hadcrut4_vs_Raw_AMO_Scatter.png
AMO data here.
http://www.esrl.noaa.gov/psd/data/timeseries/AMO/

Resourceguy

This paper does serve a purpose later when the AMO is in obvious decadal decline. It helps blunt the cooling temperature deniers the random excuse such as declining or shifting Gulf Stream and other nonsense.

michael hart

That’s the problem of an arbitrary index. It is chosen, and sometimes defined, by the person who wants to use it.
If a securities broker thinks the Dow Jones index isn’t saying what the broker wants to tell the clients, then he might choose the all-share index instead.

I’m really uncomfortable with the statement in the Muller paper: “<iThe long-term variability was suppressed by removing the least-squares fit 5th order polynomial from each curve.”.
That appears to me to be very arbitrary and very risky. Since a 5th order polynomial has no real-world meaning, there is no way of knowing whether it really does represent “long-term variability”. The act of removing it may be removing meaningful data, and could even actually be adding in meaningless data. What’s left after its removal does not necessarily have any real-world meaning since its value is the difference between “some real world stuff” and “some non real world stuff”.
In particular, the period studied was 1950-2010, which is only 60 years. It seems inevitable that some of the ‘5th order polynomial’ would necessarily represent much of whatever effect the AMO has – the very effect that is supposed to be visible after the 5th order polynomial has been removed. I don’t see how the Muller paper can be taken seriously.

Espen

Very interesting post! I was surprised to see that land temperatures in most of Europe correlate better with detrended North Pacific SSTs than with the AMO.

Henry Galt

Comparing what Muller (et al) does with what science actually is is comparing apples with oranges.
Comparing what Muller (et al) does with funding … not so much.
Yes, they read WUWT. Anyone in their right mind would. Oh, yeah, sorry.

RichardLH

Black Diamond or Black Swan?
Take a moment of your time. If you can see that the logical extension of the UAH data series in this presentation should next contain a Black Diamond (a falling node) and can place it somewhere on the graph with good arguments as to why and when, then welcome to the world of short term climate prediction!
http://s1291.photobucket.com/user/RichardLH/story/70051
This is effectively just a presentation of the low frequency (1 to 15 years) natural cycles visible in the present in the satellite data (i.e. from 1979). That is, measured cycles, not proposed ones!
Nyquist limits the presentation to cycles less than 15 years in the output for now.
A black swan moment?

Richard M

Here’s my view (again). The PDO is actually an indicator of the structure of global atmospheric pressure zones (GAPZ) – (think Bermuda high, or Aleutian Low). The average position of these GAPZs affect the jet streams. For example, a strong Bermuda High in 2012 led to a more northerly jet stream in the US which led to less precipitation and warmer temperatures.
At this geological blink of time the GAPZs have been cycling through a ~60 cycle. The PDO appears to be a good proxy for the GAPZ’s positions. When the PDO is positive as it was from 1976-2005 then several things occur:
– El Niño are more prevalent (which releases more heat into the atmosphere)
– The AMO index starts to increase. (which leads to more Arctic sea ice loss)
– More zonal jet streams (Which reduces global cloudiness/albedo slightly)
The opposite happens when the PDO becomes negative as is the case right now.
This can have interesting effects. For example, with a positive PDO the jet stream position over the N. Atlantic is driven higher (more northerly) allowing more sunshine over the water and the AMO to increase. However, this also causes the jet stream to dive down over Europe leading to cooler temperatures even while the North Atlantic is warmer.
Keep in mind that all of this refers to “average” positions.
This explains almost everything we’ve seen temperature-wise over the last 100 years. Of course, it doesn’t explain what drives the changes. There are several possibilities both terrestrial and non-terrestrial.

Stephen Wilde

Richard M says:
June 18, 2013 at 7:48 am
Much as I’ve been suggesting since 2008.
Next step, consider what causes longer term variations beyond the 60 year cycle.
I suggest top down solar effects altering global albedo to skew ENSO towards El Nino or La Nina via cloudiness changes.
The latitudinal positions of the climate zones and jet streams serve as a proxy for changes in the energy budget such that zonal / poleward is a sign of warming and equatorward / meridional is a sign of cooling and in each case the circulation change is a negative system response to whatever the net forcing effect of all relevant mechanisms is at any given moment.
So, if the system tries to warm then the poleward zonal pattern lets energy flow through the system faster (less clouds) and if the system tries to cool then the equatorward meridional pattern causes energy to flow through the system more slowly (more clouds).

RichardLH

Richard M says:
June 18, 2013 at 7:48 am
“Keep in mind that all of this refers to “average” positions.
This explains almost everything we’ve seen temperature-wise over the last 100 years. Of course, it doesn’t explain what drives the changes. There are several possibilities both terrestrial and non-terrestrial.”
I would propose that natural cycles or 37 months, 4 years, 7 years (3+4) and 12 years (3*4) are of more than passing interest.

Greg Goodman

“We identify one strong narrow spectral peak in the AMO at period 9.1 ± 0.4 years and p-value 1.7% ”
Hey, that sounds a lot like Scafetta’s 9.01 +/-0.1 , which he identified as be lunar in origin. 😉

Greg Goodman

My article on lunar-solar influence:
http://climategrog.wordpress.com/2013/03/01/61/
identified this as a significant frequency in many basins. It also showed how Hadley manage to remove it from the SST record.
http://climategrog.files.wordpress.com/2013/03/icoad_v_hadsst3_ddt_n_pac_chirp.png
Maybe that is why it took a study of land temps to find this.

Theo Goodwin

What a polite but total take down of the essay by Muller and friends. There are two major highlights:
“Many papers and blog posts attempt to remove the impacts of natural variables from the global surface temperature record. When they add AMO data to the ENSO, volcanic aerosol and solar data in their multiple regression analyses, do they recognize and account for the fact that the AMO data and global land surface air temperature data have similar responses to ENSO and volcanic aerosols?”
Let me restate this. Many people use purely statistical techniques to accomplish what they call “removing the impacts of natural variables” only to reveal that they are utterly clueless about the natural processes from which their data was taken. These people have the annoying habit of excluding from their work any and all hypotheses about the natural processes. (Even my version comes across as polite – must be the influence of Tisdale.)
The second point is that not only are Muller and friends uninterested in underlying physical processes but are inexplicably careless with the data obtained from them. Mr. Tisdale writes:
“A good portion of Muller et al (2013) deals with comparisons of global surface temperatures with the Atlantic Multidecadal Oscillation index (AMO) and with the Pacific Decadal Oscillation index (PDO), to emphasize their findings.
Muller et al (2013) failed to recognize that the AMO index data is detrended sea surface temperature anomalies of the North Atlantic, while the PDO index is NOT detrended sea surface temperature anomalies of the North Pacific. The PDO is an abstract form (a Picasso version, if you will) of the sea surface temperature anomalies of the North Pacific. So Muller et al (2013) were comparing apples to oranges.”
At this point, politeness is really difficult though Mr. Tisdale managed to be polite. After recovering from a powerful “face palm,” I am pretty much speechless. How could Muller and friends have overlooked this? How could journal reviewers have overlooked this? (Pardon my lack of politeness, but this matter raises a serious question of trust.)

RichardLH

Greg Goodman says:
June 18, 2013 at 8:54 am
“Maybe that is why it took a study of land temps to find this.”
This is a summary of observed short term (< 15 years) cycles in the satellite data.
http://s1291.photobucket.com/user/RichardLH/story/70051

Thanks, Bob.
You have reported on reality, but many deny the real, too complex world and then device models that reflect only their own preconceptions.

Kristian

Paul Vaughan says, June 18, 2013 at 5:00 am:
“You inform repeatedly. People ignore repeatedly. This has gone on for years — that’s too long. My patience with those who can’t or won’t get it has expired.”
I couldn’t agree more. People still act as though Tisdale’s strictly data-based explanation of global warming since the 70s doesn’t exist and has never been put forward, on this blog or his own, or elsewhere. You can see posts back-to-back with Tisdale expositions of how it all went down acting all confused about the recent ‘pause in warming’ or how there really has been warming over the last 35 years so there should be a certain CO2 component in there somewhere (read: climate sensitivity studies). People are still lamenting the level of scientific knowledge about what really rules the climate, as if there are somehow huge gaps in our understanding, some missing and as of yet unknown mechanisms pulling the strings. And yet they’re bound to have read or at least heard of this guy called Bob Tisdale at some point during the last four years showing us all that ENSO is the natural process doing the pulling – the Great Puppet Master – not just on an interannual or decadal scale, but on a multidecadal one. There is hardly a gap in our understanding of what caused the global warming since the 70s, which clearly and obviously is contained in its entirety within three abrupt shifts. Not for those of us who care to have a look at what the real-world data are actually telling us. Bob Tisdale has, once and for all, rid us of the need for any CO2 ‘God of the Gaps’ … And he’s still summarily ignored or dismissed. I wonder what kind of psychological or sociological phenomenon that lies behind.

Matthew R Marler

Muller et al “suppressed” the “long-term variability” by subtracting the values of a 5th order polynomial curve from the data.
Curious wording on their part. Usually that is called “detrending”.

Kristian

Espen says, June 18, 2013 at 6:49 am:
“I was surprised to see that land temperatures in most of Europe correlate better with detrended North Pacific SSTs than with the AMO.”
I think you will find that European temperatures correlate most tightly with the AO/NAO mode.

Matthew R Marler

Muller et al (2013) failed to recognize that the AMO index data is detrended sea surface temperature anomalies of the North Atlantic, while the PDO index is NOT detrended sea surface temperature anomalies of the North Pacific. The PDO is an abstract form (a Picasso version, if you will) of the sea surface temperature anomalies of the North Pacific.
I take it you were not satisfied with the detrending carried out by Muller et al. What can you possibly mean by “a Picasso version”?

Matthew R Marler

And based on our earlier discussion, the reason the AMO has a positive or neutral correlation with global surface temperatures almost everywhere is, the AMO and global surface temperatures respond similarly to ENSO, ENSO residuals and volcanic aerosols. Simple.
I noticed that here and elsewhere you accepted a main result of Muller’s analysis while exploring interpretations that they have slighted or ignored.
It’s a shame you won’t submit your work for publication.

Theo Goodwin

Matthew R Marler says:
June 18, 2013 at 9:31 am
“It’s a shame you won’t submit your work for publication.”
Why?

“Greg Goodman says:
June 18, 2013 at 8:50 am
“We identify one strong narrow spectral peak in the AMO at period 9.1 ± 0.4 years and p-value 1.7% ”
Hey, that sounds a lot like Scafetta’s 9.01 +/-0.1 , which he identified as be lunar in origin. ;)”
#########
that would kinda explain why we cite him and discuss it. doh. the benefits of reading a paper before commenting are sometimes large.
“Scafetta [2010] reported a forest of 11 spectral
peaks based on a multitaper analysis; to each of these peaks,
he calculated 99% confidence intervals. He reported seven
peaks with periods in the range from 5.99 to 14.8 years.
One of these is at our period of 9.1 years; he suggests that this
cycle could be induced by lunar tidal variations. However,
we find that when we use our Monte-Carlo methods to estimate background, none of his claimed peaks are statistically
significant except for the 9.1 year peak; we do not find them
in the AMO, PDO, or ENSO.”

Theo Goodwin

Bhb Tisdale says:
June 18, 2013 at 10:39 am
When creating a portrait, Picasso would on occasion show the subject in multiple perspectives at the same time. To the untrained eye, some of his subjects seemed to have three noses trying to fit in the same spot. Picasso’s work gave rise to an industry of critics who explained how these portraits are to be viewed. Time revealed that there is no rational explanation of how to view these portraits. Mr. Tisdale chose a very good analogy.

This paper is interesting for several reasons.
The first reason is that it was received by JGR on the 22nd of April 2013; it was revised on the 28th of April 2013; and it was accepted on the 30th of April 2013.
I would love to receive such speedy review for my own papers.
The second reason why this paper is very interesting is because it contains one important result, that is that the AMO and the PDO are characterized by a significant quasi 9.1 year oscillation. The only major problem with this important “discovery” is that this about 9.1 year oscillation in the AMO and PDO indexes was actually discovered and extensively discussed in one of my papers, see for example figure 9 in:
Manzi V., R. Gennari, S, Lugli, M. Roveri, N. Scafetta and C. Schreiber, 2012. High-frequency cyclicity in the Mediterranean Messinian evaporites: evidence for solar-lunar climate forcing. Journal of Sedimentary Research 82, 991-1005. DOI: 10.2110/jsr.2012.81.
http://people.duke.edu/~ns2002/pdf/991.full.pdf
which, of course, was not referenced. My analysis is actually better.
Fortunately, they referenced one of my past papers
Scafetta N., 2010. Empirical evidence for a celestial origin of the climate oscillations and its implications. Journal of Atmospheric and Solar-Terrestrial Physics 72, 951-970.
DOI: 10.1016/j.jastp.2010.04.015.
http://people.duke.edu/~ns2002/pdf/ATP3162.pdf
where the 9.1 year oscillation is also discovered in the global temperature records around he earth and it was demonstrated that it is very likely a soli-lunar long-range tidal cycle which should be particularly strong in the ocean related record such as the PDO and AMO.
A better explanation of this soli-lunar tidal cycle is discussed in page 35-36 of the supporting material of
Scafetta N., 2012. Testing an astronomically based decadal-scale empirical harmonic climate model versus the IPCC (2007) general circulation climate models. Journal of Atmospheric and Solar-Terrestrial Physics 80, 124-137.
DOI: 10.1016/j.jastp.2011.12.005.
http://people.duke.edu/~ns2002/pdf/ATP3533.pdf
which they do not reference either.
Then there are other minor issues such as their proposed ““prewhitening” process made with a 5-order polynomial that is a very rude and very dangerous high-pass filter that would partially deform also the high component of the spectrum (fortunately only a little bit). And other details such as their claim that I found a “forest” of 11 spectral based on a “multitaper analysis”, which is false because I am mostly using the Maximum Entropy Method which is far more accurate, and other small details such as the AMO and PDO present an oscillation closer to 60-year than a 70-80 year one of the model.
However, given the fast review of the paper, I do not expect that the reviewer had much time to give a look around or study statistics.
But the paper is interesting, nevertheless. A better review would have improved it.

jai mitchell

1. The correlation of .65 +/- .04 is a very moderate correlation, to show strong correlation one must show >.85 — Even with strong correlation the following quote from the paper stands, “Correlation does not imply causation.”
2. The author’s suggestion of only using north pacific SST during the PDO phase is intriguing and will show stronger correlation though the correlation will be negative. The reason this is true is because the south hemisphere’s effect (cooling) is stronger. . .this is based on the following quote from the article,
“An El Niño releases heat into the atmosphere, and surface temperatures around the globe in many places warm in response to the El Niño, and in other parts, they cool—with more locations warming than cooling, so the average global surface temperature warms in response to the El Niño. But the heat released into the atmosphere during the El Niño is not directly warming the surface in those remote locations. ”
conversely, a La Nino cools the atmosphere by allowing more atmospheric heat energy to enter the oceans. (please note that the atmospheric energy is also entering the oceans during El Nino, just not as much as neutral and La Nina phases.
3. The Berkeley Earth Temp Project summary of findings. says,
“The good match between then new temperature record and historical carbon dioxide records suggests that the most straightforward explanation for this warming is human greenhouse gas emissions.”
This indicates that the relatively low correlation of El Nino and AMO to Land Surface temperatures does not IN ANY WAY detract from the reality of anthropogenic climate change as currently understood and caused by the increase of greenhouse gasses. (the low correlation of El Nino and AMO only displays a shift in the energy balance between ocean and atmosphere and does not in any way show more or less warming).
4. This is only another “wake up call” that indicates we need to be more concerned with land surface temperatures and not sea surface temperatures and definitely not lower-correlation satellite tropospheric temperatures that do not show northern hemisphere land temperature amplification (where most of humanity lives and grows food).
I am a big fan of Dr. Muller. . .by the way.

Ashby

I’m a little confused here. If they removed this signal from the temperatures:
http://bobtisdale.files.wordpress.com/2013/06/amo-with-5th-order-polynomial.png
…didn’t they remove precisely what they might be looking for?

Bill Illis

New paper on North Atlantic SSTs published today in the Holocene going back to 1000 AD. Looks a little different than other AMO reconstructions we have seen but 10 different reliable-type proxies were used in this study.
http://hol.sagepub.com/content/23/7/921.abstract?rss=1&utm_source=feedly
Reconstruction (with a little help from me on conversion to temp-anomaly type numbers).
http://s8.postimg.org/5a2g4xnc5/North_Atlantic_SST_Proxies.png
Data here.
ftp://ftp.ncdc.noaa.gov/pub/data/paleo/paleocean/by_contributor/cunningham2013/cunningham2013-data-series.txt

jai mitchell

Nicola,
on section 6 page 36 you show the 9.1 year harmonic cycle. It shows a temperature variance of .09C peak-to-trough as caused by the sol-lunar cycle that you hypothesize causes this temperature variability.
questions:
-you have no error bars on your graphic
1. what are the uncertainties of these values?
In your paper you state,
“Finally, there may be an additional natural warming due to
multisecular and millennial cycles as explained in Introduction. In
fact, the solar activity increased during the last four centuries
(Scafetta, 2009), and the observed global surface warming during
the 20th century is very likely also part of a natural and persistent
recovery from the Little Ice Age of AD 1300–1900”

and yet, you offer no hypothesized mechanism for this recovery, however, you penalize the “observed anthropogenic warming” based on this unsupported claim, you go on to say,
“Thus, the above estimated 1:30 1C=century anthropogenic warming
trending is likely an upper limit estimate. As a lower limit we
can reasonably assume the 0:6670:16 1C=century, as estimated in
Loehle and Scafetta (2011), which would be compatible with the
claim that only 0.2 1C warming (instead of 0.7 1C) of the observed
0.5 1C warming since 1970 could be anthropogenically induced”
and then use this in your models.
2. Don’t you think it is intellectually dishonest to assume an effect, without determining causality and in opposition to the preponderance of supportive evidence, and then use that assumption as an input for the production of a modeled result? Isn’t this, in effect, the epitome of “quasi-science”?
Unless you have a provable theory of causality for the “Jupiter effect” that you use as an excuse for a “recovery” from the little ice age, your results are incredibly suspect.

michael hart

jai mitchell says:
June 18, 2013 at 11:12 am
3. The Berkeley Earth Temp Project summary of findings. says,
“The good match between then new temperature record and historical carbon dioxide records suggests that the most straightforward explanation for this warming is human greenhouse gas emissions.”

But fortunately this paper neither mentions carbon dioxide, nor makes such a foolish statement.

4. This is only another “wake up call” that indicates we need to be more concerned with land surface temperatures and not sea surface temperatures and definitely not lower-correlation satellite tropospheric temperatures that do not show northern hemisphere land temperature amplification (where most of humanity lives and grows food).
When ocean cycles are removed, land surface and sea surface temperature should be strongly correlated if CO2 is the cause (assuming constant humidity). They are not. Which tells us several things, including that land surface warming isn’t cause by a uniform global phenomena, and hence CO2 can not be the cause of land surface warming.
Otherwise, show me a correlation in the raw data and I’ll take notice.

jai mitchell says: June 18, 2013 at 11:51 am
I think that you are quite confused and should study the papers better.
1) the error on the 9.1 year is about 0.2 year and it is reported in Table 2 of
Scafetta N., 2010. Empirical evidence for a celestial origin of the climate oscillations and its implications. Journal of Atmospheric and Solar-Terrestrial Physics 72, 951-970.
DOI: 10.1016/j.jastp.2010.04.015. http://people.duke.edu/~ns2002/pdf/ATP3162.pdf
2) the other calculations you are talking about are not arbitrary, but based on the existence of a 60-year oscillation that very likely contributed 0.3 C warming between 1970 and 2000, which imply that AGW has been overestimated by 50-60% at least.
3) ” Don’t you think it is intellectually dishonest to assume an effect, without determining causality and in opposition to the preponderance of supportive evidence”
Not really, the models fail to properly interpret the 1970-2000 period for the same reason they fail to interpret the 1850-1880 warming, the 1910-1940 warming and the 1880-1910 and 1940-1970 cooling, and also the standstill since 2000. They also fail to get many other things, which would be to long here to say. So, your “preponderance of supportive evidence” does not exist. It is political propaganda, indeed.
For example if you do not want to look at my papers with any care, look at this figure prepared by Spencer that shows the epic failure of 73 Climate Models vs. Measurements, Running 5-Year Means:
http://www.drroyspencer.com/2013/06/still-epic-fail-73-climate-models-vs-measurements-running-5-year-means/
and try to respond to my question: which model reproduce the temperature?
Tell me one name and model run number
So it is legitimate to assume the existence of “causes” that the modelers do not know yet, and there are papers showing this pattern in the solar records. What I do is to investigate the dynamical implication of this “causes”. This is a valid argument in physics. And my models are tested in their forecasting capability and are good indeed. See here
http://people.duke.edu/~ns2002/#astronomical_model-1
What is really dishonest in science is to assume that everything should be already known and “settled” even when the proposed models do not reproduce the data and accuse those that investigate the missing physics of “dishonesty”.

“Nicola Scafetta says:
June 18, 2013 at 11:06 am
This paper is interesting for several reasons.
The first reason is that it was received by JGR on the 22nd of April 2013; it was revised on the 28th of April 2013; and it was accepted on the 30th of April 2013.”
##############################
The paper was submitted to JGR in 2011, reviewed and accepted at that time. JGR added the stipulation that final publication would be contingent on the publication of Rohde. once rohde 2012 was published the process went quickly as the paper had already been reviewed and accepted.
“However, given the fast review of the paper, I do not expect that the reviewer had much time to give a look around or study statistics.
But the paper is interesting, nevertheless. A better review would have improved it.”
#############################
From 2011 till today the paper has been available on our website. I searched my files for your comments. I find no emails from you.
So basically we made the paper freely available.
the data is freely available
the code is freely available.
The only thing stopping people from sending a mail or asking a question is….? I dunno.
It’s pretty simple. We post up our work in progress. Anybody who wants to ask me a question about it can do so. steve@berkeleyearth.org
If you are in the area and would like to come to my wen meeting to discuss issues with the head scientist, just let me know. Door’s open.

jai mitchell

Nicola,
thank you for your reply.
I was not asking for the periodicity error that is in table 2. I was asking for the AMPLITUDE error which is the temperature error (which I mentioned as being .09C from peak to trough). Please provide (what you believe is) the temperature error sensitivity for the oscillation chart found on page 36 of the supplementary data.
————-
with regard to “quasi-science” you have answered the question phenomenally. . .you believe that it is ok to ascribe a causation based simply on periodicity and your understanding of the model. You even assume that this will work on a multi-centennial timescale with no corroborative evidence.

Tonyb

Bill illis
Your post at 11.51
Your middle graph headed S8 posting bears more than a passing resemblance to my cet reconstruction to 1538 and (very) interim generalised findings to 1210.
Can you tell me any more about it?
Tonyb

jai mitchell

Phil Bradley,
When ocean cycles, solar cycles and volcanic eruptions are removed as “noise” from the data the temperature record absolutely pairs with the CO2 increase. This can be clearly found using multiple sources.

Theo Goodwin

jai mitchell says:
June 18, 2013 at 1:25 pm
So, you too believe that the “forcings and feedbacks calculation” is complete and settled. Would you care to present it? What is the calculation for water vapor? How about cloud albedo?

jai mitchell say: June 18, 2013 at 1:20 pm
as I said, you may need to study my paper better. The amplitude uncertainties are reported for example at page 131 and there is a 20% error in the amplitudes of
Scafetta N., 2012. Testing an astronomically based decadal-scale empirical harmonic climate model versus the IPCC (2007) general circulation climate models. Journal of Atmospheric and Solar-Terrestrial Physics 80, 124-137. DOI: 10.1016/j.jastp.2011.12.005.
http://people.duke.edu/~ns2002/pdf/ATP3162.pdf
“You even assume that this will work on a multi-centennial timescale with no corroborative evidence.”
Not really, you need to study better the extended model here:
Scafetta N., 2012. Multi-scale harmonic model for solar and climate cyclical variation throughout the Holocene based on Jupiter-Saturn tidal frequencies plus the 11-year solar dynamo cycle. Journal of Atmospheric and Solar-Terrestrial Physics 80, 296-311.
DOI: 10.1016/j.jastp.2012.02.016
http://people.duke.edu/~ns2002/pdf/ATP3581.pdf

Steven Mosher says: June 18, 2013 at 12:46 pm
Ok Steven, I see what you want to say. However the paper appears “submitted” and received by JGR on the 22nd of April 2013; it was even revised on the 28th of April 2013; and it was accepted on the 30th of April 2013. As it is written on the web-site. Even if the paper was available in some draft form in 2011 the fact is that if the paper was resubmitted in 2013 it had to be updated with the available literature that extended since then. And written better given the large time they had to improve it and the numerous “email” comments that you say they received to improve it.
For example, when they say that they could find only the 9.1 year oscillation, somebody should have let them notice that by filtering the records with a 5-polynomial fit they were killing for example both the 20 and 60 year oscillation that are the other two major oscillations that I found and that I used the MEM and not the “multitaper analysis” etc.
In any case, please note that the fact that the ocean temperature records (AMO is a subset of the SST) presents a major 9.1 year oscillation has been clearly demonstrated by me in 2010.
Scafetta N., 2010. Empirical evidence for a celestial origin of the climate oscillations and its implications. Journal of Atmospheric and Solar-Terrestrial Physics 72, 951-970.
DOI: 10.1016/j.jastp.2010.04.015.
http://people.duke.edu/~ns2002/pdf/ATP3162.pdf
see figure 6B.
I hope that you acknowledge that 2010 is earlier than 2011. Because they reference my 2010 paper they should have noted that I was talking about the ocean records too and referenced my paper more properly instead of just writing a naïve and imprecise comment about my paper.
On another topic, about Rhode 2012 paper you may be interested in
Scafetta N., 2013. Discussion on common errors in analyzing sea level accelerations, solar trends and global warming. Pattern Recognition in Physics, 1, 37–57. DOI: 10.5194/prp-1-37-2013.
http://people.duke.edu/~ns2002/pdf/prp-1-37-2013.pdf
there is some comment on Rhode 2012 too and their way to use regression to interpret the temperature since 1700. See section 3.

jai mitchell

Nicola
I see the error calculation you have for your sinusoidal coefficients. They are .03 +/- .01 for the first term and .05 +/- .01 for the second.
so is the error range 20% or 33%
and how did you determine these values?
Oh I see! Thanks for the multi-scale harmonic model paper
where you say,
It is not possible to accurately reconstruct the amplitudes of the
solar dynamical patterns. Such exercise would be impossible also
because the multi-decadal, multi-secular and millennial amplitudes
of the total solar luminosity and solar magnetic activity are
extremely uncertain (p297)
So the amplitude of this sinusoidal wave function is basically just a guess, correct?
———-
you have done significant work on these papers, I don’t mean to fault them in toto because they do show a very interesting dynamic. That being said, the methodology of
A. determining a periodicity in the record
B. Knowing full well that other planets have as much or greater influence on the sun (i.e earth exerts 95.3% of the force of Saturn on the sun and Venus exerts more force than Saturn) You still only use the tidal patterns of Jupiter and Saturn.
C. This is all interesting but if you want to perform a best fit model you will have to, I repeat MUST, produce a tidal force model using AT LEAST the first 5 planets combined. You simply cannot produce a reasonably effective solar model without including the orbits of mercury-mars in your calculations.
—————
If you do have a hypothesized effect of planetary tidal forces on the internal circulation models of the sun then you cannot exclude the other planets from the model. It just won’t work and makes the entire exercise seem more like trickery.
I would like to see this done, you have put a lot of work into this!

Steven Mosher says: June 18, 2013 at 12:46 pm
Moreover, Steven, consider that my Figure 6B in my 2010 paper demonstrates a full spectral coherence among alternative regions of the Earth (North and South, Ocean and Land), which demonstrate a full synchronicity of the climate system. I talk extensively about this synchronicity in the paper.
Essentially only figure 6B of my paper by alone says much more than Muller’s paper claiming to demonstrate that there is a link between AMO and temperature.
And my paper says even much more by providing a physical interpretation of the spectral findings that they could not understand.
Muller et al. could only reference my 2010 paper talking about “a forest” about things that they do not understand.