The 2007-2008 Global Cooling Event: Evidence for Clouds as the Cause
September 26th, 2009 by Roy W. Spencer, Ph. D.
As I work on finishing our forcing/feedback paper for re-submission to Journal of Geophysical Research – a process that has been going on for months now – I keep finding new pieces of evidence in the data that keep changing the paper’s focus in small ways.
For instance, yesterday I realized that NASA Langley has recently updated their CERES global radiative budget measurement dataset through 2008 (it had previously ran from March 2000 through August 2007).
I’ve been anxiously awaiting this update because of the major global cooling event we saw during late 2007 and early 2008. A plot of daily running 91-day global averages in UAH lower tropospheric (LT) temperature anomalies is shown below, which reveals the dramatic 2007-08 cool event.
I was especially interested to see if this was caused by a natural increase in low clouds reducing the amount of sunlight absorbed by the climate system. As readers of my blog know, I believe that most climate change – including “global warming” – in the last 100 years or more has been caused by natural changes in low cloud cover, which in turn have been caused by natural, chaotic fluctuations in global circulation patterns in the atmosphere-ocean system. The leading candidate for this, in my opinion, is the Pacific Decadal Oscillation…possibly augmented by more frequent El Nino activity in the last 30 years.
Now that we have 9 years of CERES data from the Terra satellite, we can more closely examine a possible low cloud connection to climate change. The next figure shows the changes in the Earth’s net radiative balance as measured by the Terra CERES system. By “net” I mean the sum of reflected shortwave energy (sunlight), or “SW”, and emitted longwave energy (infrared) or “LW”.
The changes in the radiative balance of the Earth seen above can be thought of conceptually in terms of forcing and feedback, which are combined together in some unknown proportion that varies over time. Making the interpretation even more uncertain is that some proportion of the feedback is due not only to radiative forcing, but also to non-radiative forcing of temperature change.
So the variations we see in the above chart is the combined result of three processes: (1) radiative forcing (both internal and external), which can be expected to cause a temperature change; (2) radiative feedback upon any radiatively forced temperature changes; and (3) radiative feedback upon any NON-radiatively forced temperature changes (e.g., from tropical intraseasonal oscillations in rainfall). It turns out that feedback can only be uniquely measured in response to NON-radiatively forced temperature changes, but that’s a different discussion.
The SW component of the total flux measured by CERES looks like this…note the large spike upward in reflected sunlight coinciding with the late 2007 cooling:
And here’s the LW (infrared) component…note the very low emission late in 2007, a portion of which must be from the colder atmosphere emitting less infrared radiation.
As I discuss at length in the paper I am preparing, the physical interpretation of which of these 3 processes is dominant is helped by drawing a phase space diagram of the Net (LW+SW) radiative flux anomalies versus temperature anomalies (now shown as monthly running 3-month averages), which shows that the 2007-08 cooling event has a classic radiative forcing signature:
The spiral (or loop) pattern is the result of the fact that the temperature response of the ocean lags the forcing. This is in contrast to feedback, a process for which there is no time lag. The dashed line represents the feedback I believe to be operating in the climate system on these interannual (year-to-year) time scales, around 6 W m-2 K-1 as we published in 2007…and as Lindzen and Choi (2009) recently published from the older Earth Radiation Budget Satellite data.
The ability to separate forcing from feedback is crucial in the global warming debate. While this signature of internal radiative forcing of the 2007-08 event is clear, it is not possible to determine the feedback in response to that temperature change – it’s signature is overwhelmed by the radiative forcing.
Since the fluctuations in Net (LW+SW) radiative flux are a combination of forcing and feedback, we can use the tropospheric temperature variations to remove an estimate of the feedback component in order to isolate the forcing. [While experts will questions this step, it is entirely consistent with the procedures of Forster and Gregory (2006 J. Climate) and Forster and Taylor (2006 J. of Climate), who subtracted known radiative forcings from the total flux to isolate the feedback].
The method is simple: The forcing equals the Net flux minus the feedback parameter (6 W m-2 K-1) times the LT temperature variations shown in the first figure above. The result looks like this:
What we see are 3 major peaks in radiant energy loss forcing the system: in 2000, 2004, and late 2007. If you look at the features in the separate SW and LW plots above, it is obvious the main signature is in the SW…probably due to natural increases in cloud cover, mostly low clouds, causing internal radiative forcing of the system
If we instead assume a much smaller feedback parameter, say in the mid-range of what the IPCC models exhibit, 1.5 W m-2 K-1, then the estimate of the radiative forcing looks like this:
Note the trend lines in either case show a net increase of at least 1 W m-2 in the radiant energy entering the climate system. The anthropogenic greenhouse gas component of this would be (I believe) about 0.4 W m-2, or a little less that half. I’ll update this if someone gives me a better estimate.
So, what might all of this mean in the climate debate? First, nature can cause some pretty substantial forcings…what if these occur on the time scales associated with global warming (decades to centuries)?
But what is really curious is that the 9-year change in radiative forcing (warming influence) of the system seen in the last two figures is at least TWICE that expected from the carbon dioxide component alone, and yet essentially no warming has occurred over that period (see first illustration above). How could this be, if the climate system is as sensitive as the IPCC claims it to be?
bill. My argument is that oceans control the air temperature through water vapour which has some 100 times the atmospheric ghg capacity of c02. Effectively, this is water vapour from oceans which are 70% of the surface, and which store huge amounts of solar heat. On this basis, the oceans regulate the temperature, meaning that atmospheric temperatures cannot regulate the oceans
clouds from the oceans are a different thing of course… they regulate the air temperature in quite a different way, which is the point of Dr Spencer’s work. On top of all this is precipitation, PDO’s etc.. There is little room for the atmosphere to govern these events.
“most climate change – including “global warming” – in the last 100 years or more has been caused by natural changes in low cloud cover, which in turn have been caused by natural, chaotic fluctuations in global circulation patterns in the atmosphere-ocean system.”
Good opening sentence for any text on the subject of climate.
Leif Svalgaard (23:59:31) :
Nasif Nahle (21:31:17) :
the thermal energy absorbed by the gravity field
Is still utter nonsense, in spite of your references that do not address this. Perhaps, help us by extracting the precise statement from the reference you think supports your statement and posting it here.
The references address the energy density of the gravity field on my second post. About the basis for my hypothesis:
Guth, Alan H., Lightman, Alan P. The Inflationary Universe: The Quest for a New Theory of Cosmic Origins. Perseus Books Group, 1997, New York, New York. Pp. 335-339
Killing the Anthropogenic CO2 link:
http://climaterealists.com/index.php?id=4059&linkbox=true
http://jennifermarohasy.com/blog/2009/09/why-i-am-an-anthropogenic-global-warming-sceptic-part-3/ (see also part 1 and part 2)
Calling 212 scientists to account! (It’s about time)
http://climaterealists.com/index.php?tid=116
After three decades of continual increase, the mean Earth temperature has been decreasing steadily since 2002, as precise satellite measurements show. As a result, the steady rise in sea level has stopped since 2005.
World climate is a regular quasi-periodic phenomenon (see Fig. 2.1 of http://www.fao.org/DOCREP/005/Y2787E/y2787e03.htm ) that is driven by solar activity with a period of 70 – 80 years (Gleissberg cycle). Because of this regularity, it can be stated with absolute certainty that the mean Earth temperature will continue to decrease until 2040.
1. There exists an extremely close correlation between the changes in the mean global temperature and the small changes in the rotational velocity of the Earth – two physically unrelated geophysical quantities – (see Fig. 2.2 of http://www.fao.org/DOCREP/005/Y2787E/y2787e03.htm), which has been ignored by the mainstream climatologists, and leaves little room for a human influence on climate. Note that temperature lags rotation by 6 years. This close correlation results from the action of galactic vacuum density waves on the Sun and on the Earth (see http://www.icecap.us/images/uploads/Lobert_on_CO2.pdf ).
2. The orbital periods of all Solar System planets are very close to integer multiples and integer fractions of the periods of the well-known solar cycles. This provides further evidence for the existence of super-Einsteinian gravitational waves and of their action on all celestial bodies of the Solar System. In the post of September 26, 2009 in http://www.pakteahouse.wordpress.com/2008/09/15/a-new-book-elucidates-the-life-and-work-of-Dr.-Abdus-Salam it is shown that all orbital periods of the planets of the Solar System are very close to integer fractions and integer multiples of the periods of the Hale (22.14 years) and the Gleissberg (78 – 84 years) solar cycles.
The Close Correlation between Earth’s Surface Temperature and its Rotational Velocity as well as the Close Correlation between the Planetary Orbital Periods and the Periods of the Solar Cycles Prove that Climate Changes are Driven by Galactic Gravitational Waves
Progress in climatological science can only be achieved if the above physical facts are looked into in full depth.
How does one tell the difference between a forcing cloud and a feedback cloud? These two types have different signatures on a phase space diagram, but how would one distinguish them au natural?
The feedback cloud would be the low level convective cumulus clouds that form on a sunny day with some humidity in the air. But how about the forcing clouds? These would have to persist for longer time periods to produce the lags that are observed, so presumably they live at higher altitudes and are non-convective in nature.
Dr R. Lindzen deals with the ratio between these two types of clouds in his paper on the “Adaptive Iris effect”.
http://www-eaps.mit.edu/faculty/lindzen/adinfriris.pdf
From the abstract:
“…it has been found that the area of cirrus cloud coverage normalized by a measure of the area of cumulus coverage decreases about 22% per degree Celsius increase in the surface temperature of the cloudy region.”
Are these cooling events examples of the Iris Effect in action?
Bill,
High cloud is generally accepted as warming
Low cloud reflects SW during dayligh. BUT low cloud also retains LW and this predominates at night (no SW to reflect) i.e. cloudy nights = warm nights
So what is the forcing of low clouds? Do low clouds disappear at night?
It’s always daylight where the sun is shining, which it always is. You are picturing what happens at the spot where you are standing, a better perspective is to look at the earth from space. From that perspective, it’s like the earth is cooking while rotating on a spit. SW radiation is far more energetic than the diffuse LW that the earth radiates. Low (water) clouds lead to cooling.
Don’t think so
Cold sea will never heat warm air.
Whatever water evapaorates will remove the heat vaporization from the body of water, and transfer it to the atmosphere during condensation. This process doesn’t depend on a temp. difference. Naturally the water vapor will not condense immediately above the water, but at some altitude, the vapor will condense and transfer the heat.
I know that when a cold front comes through and the ambient temperature drops below 70F, my swimming pool gets cold very quickly. Do you know how much it costs to heat a swimming pool?
Nasif Nahle (07:35:50) :
The references address the energy density of the gravity field on my second post.
Where do they say that “the thermal energy is absorbed by the gravity field”. Please, just copy paste the statement in the reference that mentions that.
DGallagher (09:09:06)
This process doesn’t depend on a temp. difference. Naturally the water vapor will not condense immediately above the water, but at some altitude, the vapor will condense and transfer the heat
So the water vapour ascends a few hundres metres to find a lower temperature into which it condenses. This means it has passed though layers of warmer air also transfering heat to the molecule. So is it transfering ocen or air heat upwards?
Low clouds at night retain heat (LW radiation) on the surface low clouds during the day reflect SW radiation. The clouds do not follow the sun. So over a day a single cloud both reflects and retains radiation. Which wins?
gravity waves:
http://www.ligo.caltech.edu/~cit40m/Docs/40msqueezer_P070049-00.pdf
Laserinterferometric gravitational wave (GW) detectors such as the Laser Interferometer Gravitational-Wave Observatory (LIGO) [1, 2] are designed to measure distance changes on the order of 10−18 m caused by GWs from astronomical sources, such as coalescence of neutron stars and black holes, supernova explosions, and the Big Bang, providing further verification of Einstein’s theory of general relativity and opening an entirely new window onto the universe [3]. The sensitivity of the currently operational gravitational wave detectors such as LIGO [1], VIRGO [4], GEO600 [5], and TAMA300 [6] is partly limited by quantum noise that arises from the Heisenberg uncertainty due to quantum mechanical fluctuations in the number of photons at the interferometer output. Future gravitational wave detectors such as Advanced LIGO [7], which are planned to be operational in the next few years, are expected to be limited by quantum noise of light at almost all frequencies in the GW detection band (10 Hz – 10 kHz).
There must be a H-U-G-E amplification factor to modify climate from GW that have not yet been detected!
Gravity waves – are they looking for the wron colour of gravity wave – is this why they cannot find one?
Bill,
So the water vapour ascends a few hundres metres to find a lower temperature into which it condenses. This means it has passed though layers of warmer air also transfering heat to the molecule. So is it transfering ocen or air heat upwards?
The heat of vaporization is massive compared to any slight comvective heating from the lower level air.
Low clouds at night retain heat (LW radiation) on the surface low clouds during the day reflect SW radiation. The clouds do not follow the sun. So over a day a single cloud both reflects and retains radiation. Which wins?
Lets say that we were together somewhere and that we were effectively blindfolded. It is daytime and we are lying in the sunshine. I believe that both of us would be able to tell when a cloud goes over. Now let’s repeat the same experiment, but this time at night. I doubt that either of us would be able to detect a cloud passing from the slight increase in reflected LW.
“Even a blind man knows when the sun is shinin'”
The effect of the reflected SW is far more significant that the reflected LW, they have drasticaly different energy levels.
Leif Svalgaard (10:34:53):
Nasif Nahle (07:35:50):
The references address the energy density of the gravity field on my second post.
Where do they say that “the thermal energy is absorbed by the gravity field”. Please, just copy paste the statement in the reference that mentions that.
“The references address the energy density of the gravity field on my second post.”
“About the basis of my hypothesis:
Guth, Alan H., Lightman, Alan P. The Inflationary Universe: The Quest for a New Theory of Cosmic Origins. Perseus Books Group, 1997, New York, New York. Pp. 335-339″
“…when an object is lifted the gravitational field stores the energy exerted by the lifter as potential energy.”
http://www.wlsyw.com/thesise/13.asp
I wrote in my message to P. Wilson [Nasif Nahle (19:46:45)]:
I have speculated a bit about the thermal energy absorbed by the gravity field, for example; however, we don’t know a bit about it because we have not gone yet beyond our near surrounding cosmos.”
So, as you say, they do not mention my speculation. It would be a fluctuation of the electrochemical field. Although, it could be that someone else has mentioned it somewhere. Nevertheless, my speculation finds its sources on those articles, from which we find that:
1. There is a measurable energy density in the gravity field.
2. Work is transferred to the gravity field.
3. For a gravity field is created, a release of energy from the new gravity field is needed.
4. The gravity field absorbs thermal energy every time an object acquires potential energy, i.e. every time an object converts its thermal energy into potential energy.
Thus, we are not expecting to find my hypothesis in any Cosmology treatise. Not yet.
Nasif Nahle (12:46:14) :
I have speculated a bit about the thermal energy absorbed by the gravity field, for example; however, we don’t know a bit about it because we have not gone yet beyond our near surrounding cosmos. […]
Thus, we are not expecting to find my hypothesis in any Cosmology treatise. Not yet.
So, you expect we’ll know more if we go beyond our near surrounding cosmos… Might take a while…:-)
Everyone is familiar with the argument that smoking causes lung cancer. A personal friend of mine, recently retired from U of Miami; as a Professor of Psychology (behavioral) and Pediatrics; but he was more of an epidemiologist attached to the med school. The way he put the smoking thing was; “There’s a body of evidence that smoking causes lung cancer; and there’s a body of evidence that sex causes children; it’s just that the lung cancer data is much more convincing.
So back to the global climate; and we see there’s a body of evidence that clouds cause cooling.
Now Doctor Roy cautiously puts the little word “low” in there as a cloud selector.
Well I’m not as cautious as the good Doctor is, so I can afford to be rash. I’ve never met a cloud, high or low, that warmed me up when it passed in front of the sun; they all block sunlight from reaching me, either by reflection or absorption or both; in any case the solar irradiance that strikes me, and the ajacent surface when a cloud passes in front of the sun; always goes down; it NEVER goes up. And I even have taken meter readings (crude to be sure) with my Gossen Lunasix light meter to prove that it always goes down.
And I don’t buy the notion that the higher the clouds form the more they warm the surface; because I simply don’t believe that it is the clouds that cause the surface conditions; I’m convinced that it is the other way around.
Those high clouds are the RESULT of the surface conditions, and the warmer the surface; the higher the moist air has to rise to reach the dew point and form clouds, and the lower the humidity (less moisture content) the higher yet those ethereal clouds will form.
So Doctor Spencer I am on your team; it’s the CLOUDS; all of the clouds.
George
Leif Svalgaard (13:27:55) :
“Thus, we are not expecting to find my hypothesis in any Cosmology treatise. Not yet.”
So, you expect we’ll know more if we go beyond our near surrounding cosmos… Might take a while…:-)
Oh, no! I think we’ll know more about it when the existence of gravitons and Higgs’ particles in our near neighborhood is revealed. As near as the LHC begins to operate properly. There are some related discoveries which have given roots to my hypothesis, otherwise I had not thought on it. So the development of my hypothesis is closer than we think. 🙂
Terry (13:05:38) “Re Dr Spencer
While I fully understand your phase-space diagrams, I wonder if they might not be more visually instructive if they were presented as 3D plots with lag time as the 3rd dimension, thus avoiding the need to follow the “strings”.”
I have used the approach suggested by Terry in the past. For the z-axis I used a spinner that allowed control of depth (so I was looking at a slice and spinning the height of the cross-section slicer).
One sees pattern-cycles as one rolls the lag through harmonics. For example, striations become right-slanted-ovals, which become straight-lines, which become left-slanted-ovals, which become striations, etc.
This relates to visualization methods that data analysts call “coplots” (short for “conditioning plot”), which can be used to overcome paradox in areas where intuition is unsuspecting. Parametric approaches are based on assumptions that blind to some varieties of patterns which show up in coplots. Of course, if/when a parametric practitioner becomes aware of such patterns, they can modify their assumptions to easily make parametric methods notice a paradox with crystal clarity …but the problem is that analysts (particularly ones following traditional convention) don’t always [or even generally] run non-parametric exploratory diagnostics.
I have some ideas about the 2000, 2004, & 2007 spikes in Dr. Spencer’s plots, but before commenting, I need to know why there is a gap in the record at 2005. Even a single missing datum in a time series takes out a wide swath in a moving-average — (to avoid this I usually insert an estimate for missing values). The ideas I have might make sense if there is also a spike at 2005 …but if there is no spike there, I have more work to do on the conditioning…
A final note: These striations, ovals, circles, loops, etc. can arise due to relative variations in cyclical amplitude. This becomes evident when running detailed diagnostics on wavelets [which I’ve done to develop awareness of statistical coincidences that drive spurious power cycles in wavelet spectra – a problem that can, fortunately, be largely overcome with filters].
To:
Stephen Wilde (12:09:37) & tallbloke (11:43:29):
Sorry, I posted my original comment to tallbloke w/o thinking much & as a result didn’t make much sense.
I agree that air masses have only a small temp effect on the ocean below. This is seen in everyday weather synoptics — cold or hot air blowing off continents’ coasts is quickly changed to near the ocean temp (at least the shallow air near the ocean surface).
So to tallblokes original question, I was wrong — there’s no way heat can “hide” as long as the ocean-heat content can be measured w/some acceptable accuracy. Whether the current measurements meet those requirements, I’m not sure — it’s a big global ocean, and fractional-degree changes in large volumes of deep-water may be difficult to detect.
But if total ocean “heat” has actually decreased alittle, there’s no other place it could be except to have been radiated out to space (I think chemical/biological energy “storage” would prb’ly be insignificant).
“”” bill (02:00:51) :
High cloud is generally accepted as warming
Low cloud reflects SW during dayligh. BUT low cloud also retains LW and this predominates at night (no SW to reflect) i.e. cloudy nights = warm nights “””
First of all, clouds don’t change their optical behavior based on any clock. Day or night, clouds keep on doing the same things. Of course what the clouds are doing, may result in those clouds changing; and that in turn will have their behavior changing; but those changes are because of the cloud changes; not because a clock says it is night time.
We are taught that during the day, the earth absorbs solar energy and warms up, and during the night the earth radiates long wave IR radiation and cools down.
Actually the earth is cooling (radiating LW) much more efficiently during the hottest hours of the day; not at night.
And where is the evidence that the earth warms up at night because of the presence of high clouds. And the graphs say the warming is greater when the clouds are higher and wispier.
Is that even sensible ? The higher you go in the atmosphere, the lower the atmospheric density becomes, the CO2 and water vapor densities also go down at the same time, so their absorption of LW infrared Radiation also goes down with increased altitude. For most of the important part of the atmosphere where clouds reside, the temperature also goes down with increasing altitude, and since thermal radiation goes as the 4th power of the Temperature (Kelvins), the long wave emission from the atmosphere and or clouds, goes down big time, as the altitude increases.
Also the higher and less dense, and lower temperature the atmosphere is, the narrower is the spectral widths of the major absorption lines of the GHGs water vapor, and or CO2/ozone etc, so the fraction of the surface emitted LW that gets absorbed, will diminish as the altitude increases.
So you have a forever diminishing thermal radiation source that is blocking less and less of the outgoing IR, and returning some to earth through an increasingly hostile denser warmer atmospheric layer, at lower altitudes.
Does this sound like a mechanism for increasing the efficiency of surface warming from cloud emissions ?
Yes warm surfaces, and high clouds tend to occur together. The silly part is trying to present the clouds as the cause, and the surface warming as the result; particularly since it actually cools down after sundown; it does not warm up, clouds or no clouds; UNLESS A BODY OF WARMER AIR MOVES IN FROM SOMEWHERE ELSE. All bets are off in that case; which is differerent from the cloud/surface interraction.
Nasif Nahle (13:59:32) :
Oh, no! I think we’ll know more about it when the existence of gravitons and Higgs’ particles in our near neighborhood is revealed.
So the Higgs is important for our climate…
REPLY: Nasif this has to be one of the most ridiculous things ever said. The Higgs Boson has yet to be revealed. It is only theorized.
Again I grow weary of your constant silliness. Take a 24 hour time out. This continued back of and forth is pointless and a waste of my time and my moderators time. – Anthony
not sure if you have read http://gustofhotair.blogspot.com/ which looks at cloud cover and its effect that it has on Australian temperature?
Here’s a (possibly substantial) clue for Dr. Spencer:
http://www.sfu.ca/~plv/PWP_Detrended_1999-2008.png
For related insight, blink between these two:
http://www.sfu.ca/~plv/PWP&LOD1.png
http://www.sfu.ca/~plv/PWP&LOD2.png
The disturbance shows up in AO, NAM, NH-winter NAO, & Sea Ice – for example:
http://www.sfu.ca/~plv/CumuSumAO70.png
I suspect a connection with contrasts between interannual & annual aa index (or something lurking & confounded with it), but I will need to spend a considerable amount of time refining the diagnostics before commenting further because, at this time, there remains at least one further conditional-dependence that I have not yet worked out.
For now I can say I am convinced of substantial shared interannual variance involving SOI, aa (or something lurking & confounded with it), & regional precipitation patterns based on records dating back to the 19th century; I never would have discovered this had I not spent a considerable amount of time analyzing EOP (Earth orientation parameters).
George E. Smith (14:18:47) :
Oh dear! this is ridiculous!
Obviously clouds actions do not change night and day. In sunlight they affect albedo and also act as a GHG. In the dark There is nothing to albedo so they act as a GHG.
Take away the cloud moisture and the heat stored in the ground radiates away more quickly. No where has it been said that clouds will warm the night. They retain heat!
In UK a cloudy night can make a large difference to the night temperature (5 to 10C)
http://amsu.cira.colostate.edu/kidder/Oskar.pdf
http://www.atmosphere.mpg.de/enid/5dac76137ab0d6f4b731a6ecab3a7398,0/3__Sun_and_clouds/-_Clouds_and_climate_25x.html
There are huge differences in day and night time temperatures in deserts. Very few clouds form over deserts because the air is so dry. This means the surface heats up during the day and escapes rapidly at night. Night-time temperatures can be 35oC lower than daytime ones.
Clouds can begin at 0 metres but more importantly at 300 to 600 metres. The air at these heights is not a lot different to sea level.
http://www.maltaweather.info/clouds3.html
Clouds aren’t really a ‘blanket’ on warm nights
You sometimes hear that “clouds act like a blanket” to keep temperatures from falling as low as they otherwise would if the sky were clear. It’s true that the effect of clouds can be like a blanket. But they work in a different way. …The tiny water droplets that make up clouds radiate even more heat than the gases in clear air. For one thing, the tiny droplets absorb some frequencies of infrared energy that the air’s gases don’t. As the energy is absorbed, it heats the droplets, which causes them to radiate energy. Then too, the cloud has more matter – the cloud droplets – to radiate heat in all directions, including back toward Earth, than clear air.