I have heard the argument that the phenomena that is global warming is infact caused by increased sun output (completely naturalistic) as opposed to pollution (man made).
Does this report add any credibility to that argument?
Response: No. The difference is that this is talking about the sunlight reaching the ground which is much more affected by what’s in the atmosphere (clouds, aerosols etc.) than it is affected by the relatively small changes in the sun’s output. The numbers that have been reported for the ‘dimming’ effect (and now the ‘brightening’ effect) are much larger than can possibly be attributed to changes in the solar input. -gavin]
Comment by Colin JN Breame — 15 May 2005 @ 9:22 PM
We live in St. Louis, Mo when we travel to a nearby suburb
we noticed that its somewhat warmer during the summer months.
Is this due to the land altitude or because of the 3 rivers
Thank you for your reply.
Response: The reasons for any particular ‘micro-climate’ can be hugely varied – going from topography effects (altitude, rain shadows from nearby hills/mountains), surface conditions (the surface albedo, amount of open water, soil mositure) as well as degrees of urbanization (amount of tarmac vs. trees, and relevant to this post, amount of local aerosols). In your case, the local TV station weather forecasters or the local university might have a more detailed answer for you. -gavin]
Wild et al. suggest, may have removed this masking effect and lead to the signal being more obvious in the 1990s.
I would have thought that a 0.66 W/m2 per year increase in solar radiation fully explains all the 1990s warming. It certainly doesn’t leave much room for ghg warming.
[Response:This is a common mistake. There is no such increase in solar radiation ,which has remained fairly constant, as the satellite obs show. What is happening is more likely an aerosol effect, which redistributes the radiation. The changes are so large that they cannot represent forcing directly, or we would be seeing obvious temperature responses, and would have seen obvious coolings before then. This was all said before, and why we cautioned against over-interpreting these results… – William]
What is happening is more likely an aerosol effect …
Yes – I know. I realise it’s got nothing to do with changes in the Sun’s output. But if less radiation is being reflected than previously then the effect is the same.
Also when you say ‘redistributes radiation’ – redistributes Where?
Incidentally I don’t necessarily accept these results.
[Response:The point is, the effect is too large for it to be a direct loss/gain of radiation to the climate system – that would be obvious. So the incoming solar is not just being reflected out to space (at least, not totally or even largely – it may be partly) it must be being redistributed – either scattered and being absorbed somewhere above the surface, or some other effect – William]
Does this global brightening (& previous masking of GW) mean that perhaps the projections for 2100 and the climate sensitivity may have to be recalibrated upwards?
I know you suggest “caution,” and are interested in scientific certainty, but I’m willing to look at possibilities even with very less certainty (say 10% certainty or .90 significance), since climate change is a serious problem for people who live on earth, and might become an extremely serious problem in the future. It is, of course, the ultimate catastrophe even now for the 160,000 people who die from it each year, according to WHO estimates.
In discussing positions on climate change, I figure it this way: there are the scientific positions that sort of fall into a bell-shaped curve with outliers on the negative side being climate change skeptics (reputable scientists whose research is published in peer-reviewed journals); and this bell-shaped curve has over the past 15 years been moving in the direction of acknowledging climate change as scientifically substantiated; the outliers on the positive side reaching 95% certainty in 1995. Beyond this science bell-shaped curve are contrarians on one side, who for whatever reasons do not want to acknowledge the strong likelihood of anthropogenic climate change no matter what (or perhaps need 99% or 101% certainty). And on the other extreme are those concerned about life on earth, and do not like to see even 160,000 people die each year from this problem (& think the figure might even be higher in reality), much less millions in the future. I am on this latter side, and I understand that most climate scientists are stressed as scientists on the one hand, having to maintain their credibility & not cry wolf, and on the other hand being concerned people. I know you are working overtime on this, and that you understand the slowness and limitations of science, especially in predicting the future, especially regarding abrupt or highly complex changes. Thanks for your hard work & effort to communicate to us laypersons & for your patience with contrarians and concerned persons on either extreme of “science.”
So, what do your gut instincts tell you? Might the figures have to be recalibrated upwards – is this a possibility in light of this brightening evidence, or pretty much out of the question?
[Response:I still think we should be cautious in interpreting these results. I’m not an expert in this area, so I suppose this really is my gut instinct, plus a certain degree of (the right sort of) skepticism towards over-enthusiasm. Having said that, there may be a case for edging up the climate sensitivity values (partly because other work suggests somewhat the same thing), BUT these results have only just been published – they need to be digested and considered by the experts. Happily, they are about in time to feed into the next IPCC report AR4 – William]
Comment by Lynn Vincentnathan — 16 May 2005 @ 2:06 PM
It seems to me that AGD(imming) would precede AGW. Is that not a reasonable assumption (and why)? The earliest AGD information I can find is from the late 1950s. What about before then? Is AGD a possible contributor the LIA? Doesn’t the idea AGD conflict with the measure temperature rise during the first part of the 20th century? After all, if there were AGD at that time, and the temperature rise of the first part of the 20th century is attributed (as it normally is) to natural attenuation, then that would indicate a higher level of natural variation than previously thought.
[Response:Radiometer values are probably too imprecise and too few to make useful estimates pre-1950 – it is (apparently) rather harder than one might suspect to measure downwards solar to the degree of precision required with no long-term drift. Anthro influence on climate is generally considered to start about 1850-ish, so connecting it to the LIA is unlikely – William]
Comment by Michael Jankowski — 16 May 2005 @ 3:05 PM
Couldn’t variations in dimming simply be due to volcanic activity and the amount of matter in the atmosphere caused by eruptions?
So the incoming solar is not just being reflected out to space (at least, not totally or even largely – it may be partly) it must be being redistributed – either scattered and being absorbed somewhere above the surface, or some other effect
Fair point – some may be absorbed elsewhere. But it’s reasonable to assume that some is also reflected back into space. The *some* no longer being reflected could easily account for the 1990s warming.
In fact, what about this for a hypothesis?
Early 20th century – reduced volcanic activity plus increased solar activity leads to increased global temperatures (this is pretty much accepted anyway)
Mid 20th century – increased aerosol effect reverses warming trend to produce slight cooling (again more or less in line with current thinking)
Late 20th century – ‘cleaner’ technology results in reduced aerosol effect in atmosphere and a return to a warming trend.
I think I’ve cracked it.
[Response: I think you’re over-interpreting it – William]
***Anthro influence on climate is generally considered to start about 1850-ish, so connecting it to the LIA is unlikely***
I’m well aware of this, but aren’t people usually thinking in the context of GHGs, not particulates?
As for the pre-1950 dimming values, air pollution was extremely bad at times. Waves of deadly smog and “black fog” have been recorded. Couldn’t these have been significant enough to affect global temps, and if so, then wouldn’t they have masked some of the “natural warming” that we witnessed in the first part of the 20th century?
[Response:The true answer to both your questions is, I don’t know for sure. But… to some extent, particulates/aerosol scales with GHG emissions. Which is why pre-1850 the values are likely to have been small globally, however large they may have been locally. But also… yes, its possible that they may well have had effects pre-1950s, its just that the obs to verify this are hard to come by – William]
Comment by Michael Jankowski — 17 May 2005 @ 8:46 AM
I’ve had climateprediction running on my computer for a few months, but I’m not sure I’m getting my money’s worth. Why hasn’t global dimming/brightening shown up in the models? Is this sort of thing beyond the predictive capability of climate science?
[Response: It depends on the kind of experiment you are running. If it is either a control or a CO2 doubling run, you probably won’t see it (though that will depend on the cloud feedbacks in your configuration). If you are doing a greenhouse gas + sulphates run (though I’m not sure that is on offer), then it’s likely to be clearer. You should probably ask on their message boards how you could diagnose this (you need to find a diagnostic that sounds like ‘incident solar radiation at ground’). – gavin]
[Response:HadAM3 certainly includes “sfc downwards SW” as a diag (1,235 I think), but whether its available from within the CP.net framework is doubtful – William]
There is a simple reason why dimming or brightening have not so much to do with forcing.
Changes in daytime solar radiation reaching ground due to changes in atmospheric albedo by anthropogenic (short-lived tropospheric) aerosols are large, but they are compensated quite a lot by more thermal insulation (working day and night, for longwave infrared radiation}: clear sky insulates much worse. Amount of compensation is difficult to compute.
(This is the main idea which remained in my mind after reading all the papers and discussions about the problem the last time and forgetting them again. I hope I don’t oversimplify it. Of course, soot is warmed directly by sun, so there is some redistribution too, dimming is not the same as increased albedo.)
As if climate modeling weren’t complicated enough!
Comment by Michael Jankowski — 18 May 2005 @ 8:35 AM
I’ve had climateprediction running on my computer for a few months, but I’m not sure I’m getting my money’s worth. Why hasn’t global dimming/brightening shown up in the models? Is this sort of thing beyond the predictive capability of climate science?
Sulphate aerosol features are not available as standard in current versions of climateprediction. These will, however, be available in climateprediction version 22.214.171.124 which is due for release on April 1st 2006.
It is possible, though, to get additional climate factors by downloading the Fudge Factor suite of plug-ins – absolutely free of charge.
[Response:JF is correct. Maybe many people don’t realise exactly what they are running from CP.net. If you look at http://climateprediction.net/science/strategy.php you’ll see that the existing runs are 15 years each of calibration, pre-industrial CO2 then 2*CO2. Phase 2, not yet available, is simulations of the last 50 years, but this requires a full ocean, which is rather harder to set up, and slower to run (factor of 2 at least, and possibly 4) – William]
Scientists have gut instincts as much as anyone else, but we know, or at least ought to know, better than to rely on them.
Sometimes nature takes us by surprise. All scientists must always be vigilant about not letting our expectations color our results. I think climate scientists should be especially scrupulous to avoid reporting our intuitions and having them enter into public discourse.
Since nobody else appears to have mentioned it, let me throw out a possibilty which would combine the global brightening results with the recent finding that the earth’s energy balance is off. IR is absorbed in the top few mm of ocean, but visible light is only absorbed at greater depths which depend on the exact wavelength. Now if we’ve had a recent (post space age) clearing in the atmosphere, then there will be, on average, more visible light absorbed by the ocean. Since this energy is absorbed at depth, it will much more easily mixed with deeper water, resulting in not being sent back to space. OTOH, additional IR from GHGs in the atmosphere will be largely used to evaporate more water vapor from the water’s surface. The net result is that the ocean surface will be absorbing more energy than it emits until such time as either more cloud cover occurs or when sufficient heat is absorbed to equilibrate the oceans to a higher temperature.
[Response:Solar does penetrate into the ocean, but I suspect that the bulk of it is absorbed in the top, say, 10 meters. Over a fairly large fraction of the ocean, the “mixed layer” (wherein the ocean is close to isothermal) is deeper than this. So that probably means that the solar stays close to the surface and is thermally in contact with the surface values. Its also the case that we don’t need anything else to “explain” the heat balance of the ocean being off – that is already explainable. Indeed, any additional explanation would probably leave the balance off again – William]
Oh my, William! I have to laugh at your supposed point. Reread what you wrote and tell me it doesn’t amount to “We’ve gotten the results we want, let’s quit while we’re ahead!” This is rather an inside joke among skeptics concerning warmers, but to actually hear one say it, even if just as a slip of the tongue, is risible.
However to pass on to the factual point, could someone point to where we can find how mixing in the top layer of the ocean occurs and how fast? It may well be that the top few meters of ocean mix quickly, but even so, that doesn’t mean we’re dealing with something which can be ignored. After all the total amount out of balance is less than 1 watt/M2. For instance, might we assume that a given amount of solar radiation which penetrates to 5 meters is equally likely to mix downward to 10 meters as to rise to the surface?
[Response: Williams point is accurate. The maximum penetration of solar radiation is down to about 80m, but anyone who has ever done any diving will have noticed that most of the red and yellow parts of the spectrum have been absorbed even in crystal clear water by about 15m depth. Mixed layer depths range from a few tens of meters in the tropics to hundreds of meters in the North Atlantic, and so the penetration of solar radiation directly below the mixed layer is very small. Heat gets into the mixed layer (which mixes quite fast, as a function of the wind and buoyancy forcing) through long wave and solar radiation, as well as through sensible heat. Those heat anomalies effectively diffuses into the deeper ocean, and that is what’s being seen in the ocean heat content diagnostics. – gavin]
[Response:Rather than laughing, you might be better off thinking. We have a physically and observationally consistent theory and model for the radiation inbalance. Why would we want to substitute for that a physically unrealistic mechanism? If there was a puzzle needing explanation – as, for example, the mid-tropospheric temperatures – then speculation looks more sensible – William]
Perhaps RealClimate could address some of the most visible interpretations that are popping up in response to these papers? Thanks.
[Response:WCR perceptively notes that “Obviously something is very wrong here” but its probably their own analysis: which makes the std.mistake of assuming that these changes in direct solar translate directly into changes in the radiative forcing of the earth – William (belatedly signed)]
Not having ever studied Oceanography, I went looking to find out a bit about Bouyancy Forcing. Found an interesting paper: http://opd.apl.washington.edu/~dasaro/FLOATTECH/DAsaro01.pdf What I liked is a set of data at the bottom of page 8 which describes the contributions to buoyancy flux for various factors. It shows: “Buoyancy flux is the sum of that due to short and longwave radiation and that due to sensible and latent heat fluxes. Average values of these quantities are 41, -22, 4 and -27 W/m^2 respectively.”
This shows (I think) that more short wavelength radiation will produce more mixing while more long wavelength (IR) radiation (for instance from more GHGs) will result in less mixing. It would seem to me that this difference in mixing should result in global brightening sending more heat to depth quite irrespective of how much short wavelength radiation is absorbed below the mixed layer.
[Response: Err…no. Any increase in heating (from whatever source) causes less mixing (because it makes the surface more buoyant than before). There is a slight difference in this effect because the solar radiation penetrates a little further in, but since wind mixing is such a strong factor, in practice this is not a significant effect. However, the changes in ocean heat content are not being driven by big changes in mixing depths, they are mainly driven by the diffusion of heat anomalies at the base of the mixed layer, and that is the same whatever the mechanism is for warming the mixed layer itself. -gavin]
With some delay (Science needs two weeks to arrive in the library here!), I have read the different articles about global dimming and I must say, I am confused.
First the magnitude of the changes. These are enormous, compared to what can be expected for the change in greenhouse gases and aerosols in the period of interest, according to models.
About the suggestion that aerosols may be involved: Although the emissions in Western Europe have sharply declined (over 50% since 1975), the timing doesn’t correspond to the decrease in insolation until 1990 and the recovery thereafter. Neither does that correspond to the increase in solar radiation in Australia and Antarctica since 1990, where human made aerosols have no measurable impact. Moreover, the global emission of sulphate aerosols is near steady since 1975, but with a large shift from Europe (and North America) toward SE Asia. This may be seen in the data of India, but China with an explosively increasing industry shows increasing insolation…
Further, more insolation also means less reflection of incoming light, as good as by clouds as by the surface itself. But that is contradicted by the “earthshine” measurements, which measure how much visible light is reflected by the earth’s atmosphere on the dark side of the moon. These trends show exactly the opposite of what is measured at the surface and by earth orbiting satellites…
Further, direct and reflected solar light is only one part of the balance, downward IR radiation (from water vapour and other GHG’s) and IR radiation from the surface, water evaporation, precipitation,â?¦ all play a role in distributing energy from warmer to cooler places and what is received from the sun and emitted back to space. Interesting is that in the 20N-20S (even 30N-30S) area, the amount of extra (solar) energy on the surface (2 W/m2) in the last decades is more than compensated by more IR radiance back to space (~5 W/m2, measured at the top of the atmosphere). How the balance for the whole earth is, is not known to me and I haven’t found the data yet.
Thus what causes global dimming and back? I suppose that most is by changes in water vapour, as can be seen in data from the Alps. The change in water vapour there is three times larger than expected from CO2-induced water vapour feedback by models and the downward IR radiation more than compensates for insolation losses.
It is impossible to distinguish between natural water vapour variability and feedback induced by CO2. But by searching for surrogate water vapour levels, as indicated by discharge river flows, there is no trend in the river Rhine figure #13 (North Alps discharge), but a solar cycle related trend in the river Po (South Alps discharge)â?¦ Thus probably induced by the prevailing SW winds from over the (warmer) Atlantic…
As the Swiss have surface based upward and downward IR measurements on different places between 370 m and 3580 m elevation in the Alps, some of them equipped with several filters on different wavelengths of interest, it should be possible to make a difference in trends between CO2 induced downward IR radiation and that induced by water vapour. This may give a better indication of what happens in the atmosphere today…
[Response: Be careful that you don’t compare apples and oranges. Flux changes at the surface are NOT the same as flux changes at the tropopause (the definition used for the forcings bar chart). The surface numbers can be quite large and yet have little or no impact at the tropopause. In the case of ‘global dimming’, aerosols and clouds have an effect both at the ground and at the tropopause, but the ground numbers are much bigger. Secondly, 1990 is not a magic turnaround date. Instead, it is the date that BSRN network was established to provide better quality data for these questions, while the older data only went up to 1990.
The Earthshine measurements are indeed contradicted by the CERES measurements presented by Wielicki et al. Thus there is some more work to do on this. I think that the Wielicki results have a better sampling and a simpler calculation, however we will have to wait until the dust settles before coming to a conclusion.
However, the amount of ‘global dimming’ that can be attribtuted to water vapour changes is tiny – there is a small amount of NIR absorption by water vapour, but changes to that is completely dwarfed by changes because of clouds or aerosols. -gavin]
Re 13: “Response:HadAM3 certainly includes “sfc downwards SW” as a diag (1,235 I think), but whether its available from within the CP.net framework is doubtful – William”
If you download Martin Sykes’s excellent CPView program and load one of the pc files, the fifth set of data is called “Net Down Surface Shortwave Flux: SW TS only”. I assume that this is what you are referring to. So you can look at a pretty map of it if you want. Ken, does that make you feel any less short changed?
I don’t know but I suspect that in the current experiment, this diagnostic will not do anything more interesting than simply follow the cloud cover. However, there is a sulphur cycle in alpha testing, which might cause it to do something more interesting.
[Response:Net down SW is (up-down) (or down-up, I can never remember), whereas SW down is just the down bit. But indeed, without the aerosols and stuff, it won’t have much of interest to say about GD – William]
The Alps network is an interesting case, as they measured several incoming and outgoing fluxes.
That aerosols haven’t had much influence in the past decades is quite clear from the Swiss data: While SO2 levels steadily declined over Switzerland, incoming SW decreased 1981 to 2002 and 1995 to 2002. If 2003 is included (with an exceptional clear, warm summer), the trend still is negative for 1981-2003 but positive for 1995-2003.
Cloud cover slightly increased until 2002, but had a sharp decline in 2003. The net effect of clouds on total heating radiation is near zero (as less SW seems to be nearly balanced by more downward IR at this latitude).
The measured increase in water vapour caused a calculated extra absorption of 2.4 W/m2 in clear skies.
And the recent increase in surface temperature in Central Europe (three times faster than average NH) is the result of increased downward IR radiation (in the order of 5 W/m2 1995-2003). Most of it is caused by increased water vapour, and it is some three times larger than what is expected by GCM’s for the increase of CO2 and it’s water vapour feedback.
Thus in part cloud cover and mainly increased water vapour are the cause of global dimming and increased surface temperatures, at least in Central Europe. What has caused the high increase of water vapour, that is another question that needs an answer, as we see the huge changes in heat transfer that are involved.
And of course, radiation fluxes at the surface may be completely different of those measured at the top of the atmosphere. But if the TOA measurements show a net 3 W/m2 loss of energy to space in the tropics, then it should come from any other place on earth, or the earth as a whole is/will be cooling?
[Response: The Swiss data are useful, and do seem to indicate that locally the water vapour SW absorbtion is more significant than I suggested above. However, the increases in humidity reported in that paper are far in excess of what can be expected globally from the basic water vapour feedback to GHG increases (indeed, if water vapour feedabcks were that strong at a global scale, we would not still be arguing about it!). One more interesting point is that much of the absorption of SW in the atmosphere by water vapour (and indeed aersols) is re-radiated as LW radition, accounting for at least some of the increased downward LW seen in the Swiss data. -gavin]
Re #23 by Ferdinand Engelbeen, it is indeed unlikely that the global aerosol loading has decreased. Emission reductions in the US, EU and Japan have been more than offset by increases in the developing countries. The US just hit 1 billion tons per year of coal burning, but China just hit 2 billion and is growing fast. India is shooting for 400 million tpy, etc. Most of this is for electric power, which many people are getting for the first time.
The upshot of this discussion seems to be that the dimming and brightening amounts are big numbers and we don’t know at this point quite what to make of them. That is, we cannot explain them, although there are several speculations. Is this a fair assessment?
[Response: I think a fairer assessment would be that there are a number of different effects, none of which is globally applicable. It is more that different places may be seeing similar results but it may be for different reasons. Locally some of the numbers have seemed large, but on a regional to global level they are not as big. This is neither as exciting, nor as aesethetically pleasing as it could be, but that’s the real world for you. -gavin]
I saw short shrift given in response to WCR’s commentary(see comment #20), with no subtantive rebuttal(and anonymous to boot) combined with a subsequent ad hominem being allowed(#22). Can we expect a more detailed line of reasoning, or is it better to not address problematic material?
[Response:The response to #20 points out that this is the std.mistake. I’m sure its elsewhere, but to repeat: the large (a few percent) changes to the measured direct solar radiation can’t possibly be changes to the Earths energy balance, because such large changes would have immeadiately obvious effects. Instead, they represent redistributions of the energy within the system.
#22 *Is* an ad hom, but unless you know where this particular hom (Michaels) is coming from, you won’t understand why he writes this stuff – William]
I see no ad hominum attack in post #22. It was solely a backgrounder on the WCR personnel, who they are affiliated with and receive funding from. There is no harm in that. (In fact, in the journal Nature, there is a “conflicting interests declaration” which all authors are obligated to fill out.)
Saying that Pat Michaels is a part of the WCR, which receives money from ExxonMobil is no different from saying George W. Bush is a member of the Republican Party, which receives money from Halliburton or Texaco. It is not an ad hominum attack, but a statement of fact.