Arguments
Tung and Zhou circularly blame ~40% of global warming on regional warming
Posted on 21 March 2013 by Dumb Scientist
If a student asks why Alaska warms from January to July, saying "that's the annual cycle" isn't a real answer. It's just a new name for the mystery. A real answer would involve physics: the northern hemisphere receives more energy in July because it points toward the Sun more directly in July than in January.
Similarly, many people blame the recent 50 year warming trend on various "climate cycles" but don't explain the physics causing Earth to gain energy. Tung and Zhou 2013 is the latest attempt to blame a large part of Earth's warming on the Atlantic Multidecadal Oscillation (AMO), which is a long-term fluctuation in N. Atlantic sea surface temperatures. Previous attempts include Bob Tisdale's claim and Zhou and Tung 2012 which WUWT advertised.
Tung and Zhou subtract any apparent "oscillations" with periods between 50 and 90 years without explaining the physics causing Earth to gain and lose energy, as depicted in their Figure 4. They seem to imply that these cycles are more compelling than the mainstream explanation which involves nonrepeating events. But the universe is full of cycles and nonrepeating events, all of which obey the laws of physics. Because mainstream science addresses the energy balance, I find it more compelling than math without a basis in physics.
Tung and Zhou repeat the analysis of Foster and Rahmstorf 2011 (PDF) in Figure 5 while also removing the AMO, and obtain an anthropogenic warming trend over the last 33 years of 0.07°C/decade, less than half of Foster and Rahmstorf's 0.17°C/decade.
I've discussed a short video based on Foster and Rahmstorf 2011, which subtracted some natural phenomena that cause Earth to gain and lose energy:
- Solar variations, which can add energy to the Earth's surface. Importantly, anthropogenic global warming (AGW) doesn't make the Sun brighter. This means it can be subtracted without ignoring AGW.
- Volcanic eruptions, which can block energy from reaching Earth's surface. Global warming doesn't change how often volcanos erupt, so it can also be subtracted without ignoring AGW.
- ENSO, which can be defined based on pressure differences and mostly trades energy between the deep ocean and the surface. Even though global warming might indirectly affect ENSO, it's important to note that it hasn't yet: the ENSO index doesn't have a significant 50-year trend. This means it can be subtracted without ignoring AGW.
Tung and Zhou also subtracted the AMO index, which is defined as detrended N. Atlantic sea surface temperatures. It mostly adds energy to the surface by cooling the deep ocean (see below). Note that the AMO index is defined solely on temperatures, so anthropogenic global warming can increase the AMO index. This means subtracting it potentially ignores AGW.
Their analysis concludes that:
"The underlying net anthropogenic warming rate in the industrial era is found to have been steady since 1910 at 0.07-0.08°C/decade..." [Tung and Zhou 2013]
First, that's unphysical. Anthropogenic warming is caused primarily by the radiative forcings of greenhouse gases, which didn't skyrocket until about 1950 when our population and energy use per person both skyrocketed. More comprehensive analyses also show total anthropogenic radiative forcings increasing dramatically after 1950. Basic physics show that the anthropogenic warming rate should be higher after 1950.
Second, their unphysical claim isn't really a conclusion; it's actually the assumption that (if true) would have justified removing the AMO to determine the anthropogenic warming trend. Their paper is a circular argument. Here's why:
"The removal of the AMO in the determination of the anthropogenic warming trend is justified if one accepts our previous argument that this multidecadal variability is mostly natural." [Tung and Zhou 2013]
No. Removing the AMO to determine anthropogenic warming would only be justified if detrending the AMO from 1856-2011 actually removed the trend due to anthropogenic warming. But that's unphysical: basic physics show that the anthropogenic warming rate should be higher after 1950. As a result, their approach overestimates anthropogenic warming before 1950, and underestimates it after 1950.
Warming the globe also warms the N. Atlantic. Tung and Zhou have subtracted N. Atlantic temperatures that contain an anthropogenic trend over the last 33 years from global temperatures, and seem surprised to find a lower anthropogenic trend over the last 33 years. I'm not.
Tung and Zhou implicitly assumed that the anthropogenic warming rate is constant before and after 1950, and (surprise!) that's what they found. This led them to circularly blame about half of global warming on regional warming. Where'd that heat come from, if not from CO2?
Since the words "energy" and "heat" don't appear in Tung and Zhou 2013, let's approximate the energy they're ignoring. Warming a crude model of the atmosphere and upper layer of the ocean by 1°C requires about 10,000,000 megaton atomic bombs of energy. Tung and Zhou 2013 magic away 0.1°C/decade over the last 33 years, so they've ignored about 3,000,000 megaton atomic bombs of energy.
Internal variability moves heat around the Earth's climate without altering its total heat content. I've already noted that the deep ocean can't be the source of surface warming because the Argo probes show that the deep ocean is also warming. However, that claim was based on global averages. If AMO internal variability were causing surface warming through changes in circulation, the deep N. Atlantic should be losing enough heat to account for the surface warming. But the N. Atlantic as a whole has warmed over the past 50 years (PDF). So where'd that heat come from, if not from CO2?
I'm sorry, I don't understand why you object to Tung and Zhou 2013 assessement. Taking the temperature of the earth in 1910 and 2013 and joining the dots, we get a total temperature rise of 0.8o C, as confirmed by the Hadcrut 4 Global Mean temperature. Isn't that an average of 0.08o C per decade? Isn’t that what the IPCC claims is the extent of Global Warming?
LINK
Some of us have been observing the AMO effect for years, noting the almost perfect coincidence between a plot of the peaks of the 'apparent' AMO cycle, as reflected in the Southern Hemisphere temperature peaks circa 1880,1940, and 2000 and the steady rise in concentration of CO2 in the atmosphere.
LINK
If the CO2 curve is extended with ice core data, the rise in the peaks follows the carbon dioxide rise for 160 years. Meanwhile, the extended period of cooling by almost 0.3oC from 1880 to 1910 has never been explained by conventional climate science. CO2 didn't rise much in that period, but it definitely didn't fall, and yet the temperature declined for 30 years! You might like to offer that explanation today.
After 1940 there is another AMO explicable cooling phase that is always explained away by a magical process involving the increase in light reflecting aerosols. But that explanation is deeply flawed for two reasons. One, the major global cooling in the Southern Hemisphere during the period 1940-1960 takes place in the early part of the period before the aerosol build up, which according to the theory, mostly affected the industrialized Northern Hemisphere. Secondly it magically reverses the cooling trend moment clean air legislation to reduce SO2 emissions is passed, despite the fact that the SO2 emissions have fallen only slowly.
That sudden temperature rise, especially in the Northern Hemisphere, corresponds not to aerosol effects but the return of the warming phase of the AMO, a cycle that began in the Southern hemisphere in the 50s. Since 2000 the Southern hemisphere has shown some slowing in cooling followed by the Northern Hemisphere. This is exactly the same hemispheric pattern that occurred during the 40s and 50s and again in the period 1880, 1905. All the evidence is there to explain why there has been a slowing in global warming over the past decade or so, pointing to a return to rapid waming by 2020 at the latest.
What I can't understand is, why, if as this analysis shows, the AMO makes no difference to the fact that real global warming is being driven by CO2, and it is clearly accelerating alarmingly, why not tell the full story?
This might be a stupid question but does the AMO not just act a heat sink during cold phases?
It is rather telling that Tung & Zhou 2013 is so shy at presenting the actual CET data. The global data over the latter half of the analysis period is shown but no CET. Indeed, the peaks (or rises) they show at 1780 & 1840 do not exist outside their heavily smoothed frequency analysis. Once their analysis is shorn of this fantasy, is this any more than a sophisticated version of the drivel Akasofu produced a few years back?
I don't think AMO has ever definitively been shown to be anything other than SST response to surface warming. AMO lags surface temp by, at least and at most, a few weeks, IIRC.
Article: "The... AMO... is a long-term fluctuation in N. Atlantic sea surface temperatures" Sea SURFACE temperatures. Why would such temperatures NOT reflect the global surface temperature trend? Indeed, as the article states: "...anthropogenic global warming can increase the AMO index. This means subtracting it potentially ignores AGW." In like manner, if I believe that the urban heat island effect is responsible for AGW, why can't I just subtract the urban temperature profile from the global surface temperature record and 'whalla', problem solved. ENSO at least says something PHYSICAL about how heat is being entrained in the deep ocean: a La Nina ought to anchor global surface temperatures to the deep ocean and cool it. What a surprise: it does! But if you google 'noaa ocean heat and salt content' and compare the first two graphs ("0-700m global ocean heat content" versus "0-2000m global ocean heat content") you will see that the sea SURFACE temperature is much more reflective of what is going on in the atmosphere than the oceans depths. Hence, subtracting one from the other is a bit like subtracting one from itself: you end up with zero (rate of change) which, quelle surpise, is what Tung and Zhou ended up with.
Tamino has examined the AMO data several times, and in this post he states:
[Emphasis added]
Cause precedes effect - and global temperature changes precede the AMO. Tung & Zhou in their analysis are removing a significant portion of the global warming from the global warming, artificially decreasing the trends. It's an easy mistake to make in MLR, but it is still a mistake.
The IPCC and virtually all scientists agree that the Earth's surface has warmed by ~0.8°C since 1900. However, Tung and Zhou call this anthropogenic warming and claim that it's been steady before and after 1950. That's just unphysical, because our radiative forcings spiked after 1950. As I explained, this circular conclusion is actually a result of their assumption that they could subtract the AMO without subtracting anthropogenic warming signal after 1950.
Whack-a-mole time. When Judith Curry made a similar argument regarding 1910-1940, Gavin replied:
"Your statement is equivalent to stating that we can't find someone guilty of a crime today because some other cases in the past are still unsolved. It is a non sequitur. It is precisely because we have more information about changes today than we have about changes longer ago that attribution gets easier. We know for the recent period that ocean heat content went up, not down. We don't know that for 1910-1940. We know for the recent period that the stratosphere has been cooling. We don't know that for 1910-1940. We know for the recent period that solar has been flat. We don't know that for 1910-1940. Etc. ...
You appear to be demanding a level of certainty about 1910-1940 that doesn't exist before you will accept the level of certainty about recent trends that does. I fail to see how that makes sense. Each attribution problem is unique, and while the same techniques can be used , and the same logic applies, the answer, or even whether an answer can be sensibly given depends entirely on the specifics of the case." [Gavin Schmidt]
Your numerous uncited claims seem to be aimed at someone who disputes the existence of natural multidecadal variability. But as I wrote to a colleague, I just object to Tung and Zhou's assumption that detrending the AMO removes the nonlinear anthropogenic trend, which led them to unknowingly subtract signal after 1950.
I also mentioned multidecadal variability in a comment on the webpage associated with the video in this article. Kevin C's response doesn't involve anthropogenic aerosols and suggests that contributions from multidecadal oscillations are small, and so do other analyses. In the original version of this article I also referenced Isaac Held's analyses of internal variability, which are highly recommended.
In regards to the last paragraph:
You can not assume that ocean circulations only move heat around the earth system and have no effect on the net energy budget. This is because ocean circulations can modulate the distribution of radiativly active clouds, water vapor, and sea ice.
Here is a very simple example to illustrate the point:
(increase AMOC) => (increase north atlantic temperatures) => (decrease artic sea ice) => (decrease albedo) => (increase net energy absorbed at the surface).
Where did the extra heat come from? It is POSSIBLE that it did come from internal variability changing radiativly active parts of the system. This possibility can not be dismissed out of hand as simply being "unphysical"
See the following paper for a rigorous explanation for this:
http://connection.ebscohost.com/c/articles/17627276/why-ocean-heat-transport-warms-global-mean-climate
I'm almost embarrassed to pipe up in these threads because my science knowledge is so sub-par, but I can't resist pointing out: at WUWT right now this AMO is undoubtedly the hot topic.
Does it ever occur to anyone there that this is the God-only-knows-how-many-eth "alternative explanation" they've embraced for warming? First it was the sun, then it was volcanoes, then it was cosmic rays, then it was the Pacific decadal oscillation, now it's the ~Atlantic~ decadal oscillation (I might have the order wrong).
Funny how the one unifying "theory" seems to be Whatever It Is, It ~Ain't~ Greenhouse Gases! That right there to me ought to be a huge red flag to them that motivated reasoning and not genuine curiosity is in the driver's seat.
dvaytw - I always think of it as Anything But Us (ABU), an attempt to disavow all responsibility so that we can keep on with what we're doing without change.
And yes, the holding of multiple, contradictory crack-pot hypotheses is a red flag. To everyone else. But apparently not to those who push such nonsense...
Lee and I already discussed the fact that AMO can change surface temperatures through strictly internal variability and by changing radiative forcings. I'm only focusing on internal variability because Tung and Zhou used that term six times in their paper.
Note that I'm not dismissing AMO radiative forcings as unphysical. Instead, I've repeatedly pointed out that it's unphysical to assume that anthropogenic radiative forcings are the same before and after 1950.
what if the 0.08/decade increase is not attributable to new emissions but the effect of deforrestation, deforrestation has been occurring at a significant rate since 1850s which corresponds to the data set time period
The article linked to a radiative forcings chart that includes deforestation as a small "land use" component. It's negative because clearing rainforests to plant endless fields of identical crops increases the albedo, reflecting more sunlight and producing a slight cooling effect.
I've already noted that deforestation also releases CO2 but in much smaller quantities than fossil fuel emissions. If deforestation before 1950 warmed the planet at the same rate as fossil fuel emissions after 1950, then why did atmospheric CO2 increase faster after 1950?
significant deforrestation has not been replaced with other plants, significant deforrestation are in urban areas, the deforrestation reduces the co2 cycled through natural systems leading to more atmospheric co2
Atmospheric CO2 increased faster after 1950.
slasher@14 it is straightforward to show that argument is incorrect. Let C' be the annual change in atmospheric CO2 (which we can measure directly) and Ea is annual anthropogenic emissions (fossil fuels are taxed, so we have good estimates of that) and En and Un are natural emissions of CO2 into the atmosphere and natural uptake of CO2 from the atmosphere (neither of which we can measure directly). Fortunately, the carbon cycle obeys conservation of mass, so we know that
C' = Ea + En - Un
Technically there should also be a Ua representing anthropogenic uptake (i.e. carbon sequestration), but that is essentially negligible, so it has been neglected. We can rearrange this to give
En - Un = C' - Ea
which means we can estimate the net natural carbon flux from things we can actually measure. If we plug in the measurements, we get this:
The green line is the natural contribution to atmospheric CO2, and we can see that it is not only negative (meaning the natural environment takes up more carbon from the atmosphere than it emits into it), but that it has been taking more and more carbon out of the atmosphere each year as time progresses. So if it were not for fossil fuel emissions, CO2 levels would currently be falling, rather than rising.
According to the IPCC reports, the carbon cycle has actually intensified, with the natural fluxes both into the atmosphere and out of it increasing. See below, the figures in red show the changes in the fluxes since the pre-industrial equilibrium.
This shouldn't be too much of a surprise, if only because moving a substantial amount of carbon from the lithosphere into the atmosphere (via burning fossil fuels) means that there is more carbon sloshing about. See the IPCC report for details, or better still, David Archers excellent primer on the carbon cycle.
I suspect any further discussion of this topic should be on a more appropriate thread, such as this oneDr. Tung responds.
When I try to watch the video I get a message "This video is private". Any chance of fixing that?