A new study just published in the January 2009 issue of Journal of Climate uses a model to study the effect of warming oceans on the extensive low-level stratocumulus cloud layers that cover substantial parts of the global oceans. This study, entitled “Response of a Subtropical Stratocumulus-Capped Mixed Layer to Climate and Aerosol Changes”, by Peter Caldwell and Christopher Bretherton, is important because it represents a test of climate models, all of which now cause low level clouds to decrease with warming.

And since less low cloud cover means more sunlight reaching the surface, the small amount of direct warming from extra CO2 in climate models gets amplified – greatly amplified in some models. And the greater the strength of this ‘positive cloud feedback’, the worse manmade global warming and associated climate change will be.

But everyone agrees that clouds are complicated beasts…and it is not at all clear to me that positive cloud feedback really exists in nature. (See here and here for such evidence).

The new Journal of Climate study addressed the marine stratocumulus clouds which form just beneath the temperature inversion (warm air layer) capping the relatively cool boundary layer to the west of the continents. The marine boundary layer is where turbulent mixing of water vapor evaporated from the ocean surface gets trapped and some of that vapor condenses into cloud just below the inversion.

That warm temperature inversion, in turn, is caused by rising air in thunderstorms – usually far away — forcing the air above the inversion to sink, and sinking air always warms. The inversion forms at a relatively low altitude where the air is ‘prevented’ from sinking any farther. This relationship is shown in their Figure 1, which I have reproduced below.

Conceptual model of how marine stratocumulus clouds are formed, from Fig. 1 in Caldwell and Bretherton (2009).

The authors used a fairly detailed model to study the behavior of these clouds in response to warming of the ocean and found that the cloud liquid water content increased with warming, under all simulated conditions. This, by itself, would be a negative feedback (natural cooling effect) in response to the warming since denser clouds will reflect more sunlight. At face value, then, these results would not be supportive of positive cloud feedback in the climate models.

But what is interesting is that the authors do not explicitly make this connection. Even though they mention in the Introduction the importance of their study to testing the behavior of climate models, in their Conclusions they don’t mention whether the results support – or don’t support — the climate models.

And I would imagine they will not be happy with me making that connection for them, either. They would probably say that their study is just one part of a giant puzzle that doesn’t necessarily prove anything about the climate models that predict so much global warming.

Fair enough. But a double standard has clearly been established when it comes to publishing studies related to global warming. Published studies that support climate model predictions of substantial manmade global warming are clearly preferred over those that do not support the models, and explicitly stating that support in the studies is permitted.

But results that appear to contradict the models either can not get published…or (like in this study) the contradiction can not be explicitly stated without upsetting one or more of the peer reviewers.

For instance, a paper I recently submitted to Geophysical Research Letters was very rapidly rejected based upon only one reviewer who was asked to review that paper. (I have never heard of a paper’s fate being left up to a single reviewer, unless no other reviewers could be found, which clearly was not the case in my situation). That reviewer was quite hostile to our satellite-based results, which implied the climate models were wrong in their cloud feedbacks.

One wonders whether support of climate models would have been mentioned in the Caldwell and Bretherton paper if their results were just the opposite, and supported the models. Of course, we will never know.

For those who are interested in monitoring how the current month’s global average tropospheric temperature is shaping up, we have a website where you can plot daily global average satellite-based temperatures since August 1998. The data come from the AMSU instrument flying on the NOAA-15 satellite, and the updates are made automatically once a day in the late afternoon; they run about 1-2 days behind real-time.

The following screenshot shows an example I took from the website today.

NASA Discover project website screenshot of tool to monitor daily global-average temperatures measured by the NOAA-15 satellite.

The webpage tool uses Java, which can be kind of slow loading (clearing your browser’s cache first seems to help…e.g., in Firefox use Tools => Clear private data).

Use the drop-down menu to pick “ch5” (AMSU channel 5) which is the channel John Christy and I use to monitor mid-tropospheric temperatures. The fairly large fluctuations seen within individual months are usually due to increases (warming) or decreases (cooling) in tropical rainfall activity, called “intraseasonal oscillations”.

Channel 9 is the channel we use for lower stratospheric temperatures. Channel 13 is in the upper stratosphere, and is kind of interesting since it is expected to cool with increasing atmospheric carbon dioxide concentrations.

The check boxes allow you to choose which years to display. You can also display the daily record highs and lows measured during the first 20 years of our MSU-based record, 1979-1998, but only for AMSU channels 5 (mid-troposphere) and 9 (lower stratosphere).

What we usually watch is how the current month is shaping up compared to the same calendar month in the previous year. For instance, in the screenshot above we see that early January 2009 is running roughly the same as January 2008, which ended up being close to the 1979-98 average.

This web page should be used as only a rough guide, because there are some data adjustments made before we officially post the UAH monthly updated data. (I post a plot of those data here.) The biggest adjustment is the fact that we don’t even use NOAA-15 right now…we are using the AMSU data from NASA’s Aqua satellite in the final UAH product.

Every winter I start thinking about the processes that lead to brutally cold air masses over regions far removed from the warming influence of the oceans…Siberia, interior Canada, etc. Temperatures in interior Alaska have been routinely dipping into the -50’s F over the last week or so. Europe has been hard hit by unusually cold weather, and Vladimir Putin decided this would be a good time to cut off natural gas supplies to a number of European countries.

As of this writing (January 9), it looks like the coldest temperatures in the Lower 48 are yet to come, as the coldest airmass over northwest Canada finds its way down into the central and eastern U.S. starting around next Wednesday (January 14) or so. Gee, where is global warming when you really need it?

The ‘scientific consensus’ is that these frigid air masses are the ones that should warm the most with manmade global warming. The reasoning goes that since they contain very little water vapor (Earth’s main greenhouse gas), the warming effect of the extra carbon dioxide should be proportionately greater there.

But what causes these air masses to get so cold in the first place? Well, little or no sunlight is the most direct reason, which means they radiatively cool to outer space without any solar heating to offset that infrared cooling.

But what limits how cold they can get? Why do these temperatures seldom fall below -60 or -70 deg. F….temperatures reached fairly early in the winter, but which then level off? The answer is mostly related to the water vapor content of the air.

There is an interesting issue of causation involved with these cold and dry air masses. Contrary to what some meteorologists think, the air doesn’t become dry because of the cold. If that was the case, the air would become continuously saturated with clouds and fog as it keeps cooling, rather than clear and relatively dry as is observed.

No, rather than being dry because it is cold, the air instead becomes cold because it is dry. And the reason the air is so dry is because it has been slowly sinking from high in the atmosphere, where there is very little water vapor. And why is THAT air so dry? Because precipitation processes have removed the water vapor as relatively warmer and moister air ascends in low pressure areas — snowstorms — which move around the periphery of the high pressure zones that are created by the strong cooling.

So, ultimately, it is precipitation processes in regions remote from these cold high pressure areas that mostly determine how cold surface temperatures will get. And since we have little understanding of how these precipitation processes in the upper atmosphere might change with ‘global warming’, there is (in my mind) more uncertainty about water vapor feedbacks than the IPCC has led us to believe.

I thought it would be interesting to examine the behavior of the coldest temperatures in the climate models that are tracked by the IPCC. Monthly gridded data are archived at PCMDI for transient CO2 simulations from these models, and we simply took the minimum monthly average temperature anywhere in the Northern Hemisphere for each month in the first 70 years of those simulations. Some of the results are shown in the figure below (note the temperature scaling is relative, not absolute). The red lines are 12-month running averages.

Across those 22 models, the minimum monthly temperatures warmed by an average of 0.43 deg. C per decade (0.77 deg. F per decade), which is somewhat less than double the average global warming rate in those models. Thus, the coldest air masses in the models do warm much faster than the global average temperature does.

Also, the average minimum monthly temperature for the first February across all 22 models was -50 deg. C (-58 deg. F), indicating that the models do indeed create very cold airmasses.

The model with the greatest rate of warming is also shown: the INM CM3.0 model warmed at an average of 0.80 deg. C per decade (1.44 deg. F per decade). The model with the least warming (actually, it had a zero warming trend) was the FGOALS 1.0, which is also the least sensitive of all the IPCC models analyzed by Forster & Taylor (2006 J. Climate).

What does all this mean for the theory of manmade global warming? How fast have these coldest airmasses warmed, compared to the IPCC models? Well, the location in Siberia that is traditionally the coldest, Ojmjakon, hit -60 deg. C (-76 deg. F) twice last month (December, 2008), a temperature that has been reached only one other time in the last 25 years. So, I suspect that global warming isn’t happening nearly fast enough for the folks who live there.

(Jan. 10 update: A few people seem to have missed the point of this satirical post. It is a counterpoint to Al Gore’s use of “millions of tons” when talking about CO2 emissions. I’m pointing out that relative to the total atmosphere, millions of tons of CO2 is miniscule. And even a 50% increase in a very small number [the CO2 content of the atmosphere] is still a very small number.)

Now that there have been 50 full years of atmospheric carbon dioxide concentration monitoring at Mauna Loa, Hawaii, I thought January1, 2009 would be an appropriate time to take a nostalgic look back.

As you well know from Al Gore’s movie (remember? It’s the one you were required to come to English class and watch or the teacher would fail your kid), we are now pumping 70 million tons of carbon dioxide into the atmosphere every day as if it’s an “open sewer”.

Well, 50 years of that kind of pollution is really taking its toll. So, without further ado, here’s what 50 years of increasing levels of CO2 looks like on the Big Island:

As you can see, there has been a rapid…what? You can’t see it?…oh, I’m sorry. It’s that flat line at the bottom of the graph…here let me change the vertical scale so it runs from 0 to 10% of the atmosphere, rather than 0 to 100%….

Now, as I was saying…you can see there has been a rapid increase…what? what NOW? You still can’t see it?? It’s that blue line at the bottom! Are you color deaf?

Obviously, you had too much to drink at the New Years party last night, and your eyes are a little blurry. Here, I’ll change the scale…AGAIN..to go from 0 to 1% of the atmosphere….

Now can you see it? Good. As I was saying, 50 years of carbon dioxide emissions by humanity has really caused the CO2 content of the atmosphere to surge upward. It might not look like much, but trust me, Mr. Gore says….

NOW what?? Carbon dioxide is what? Necessary for life on Earth?

What are you, some kind of global warming denying right-wing extremist wacko? The polar bears are drowning!!

I can see I’m just wasting my time…sheesh.

Back in 1997, Bob Cess (climate researcher, cloud expert) said in an interview with Science magazine’s Richard Kerr,

“…the [climate models] may be agreeing now simply because they’re all tending to do the same thing wrong. It’s not clear to me that we have clouds right by any stretch of the imagination.”

In the last year or so I have become convinced that this is indeed what has happened…..the models are all doing the “same thing wrong”. While I have addressed this before, I am going to continue to harp on this issue until one or more climate modelers finally has a light bulb go on in their head and says, “Ahhh…I see what you’re talking about now…”. The issue is critical, and could completely change our perception of the role of clouds in climate change.

First, though, a little background for the uninitiated. Modern climate change theory is all about radiative forcing (aka, radiative energy imbalance) of the climate system: Something causes either a change in the rate at which solar energy is absorbed by the Earth (for instance, a major volcanic eruption), or the rate at which the Earth emits infrared energy back to outer space (for instance, increasing atmospheric carbon dioxide concentrations). The resulting global average radiative energy imbalance then causes a temperature change. This part of the theory is seldom disputed.

But that temperature change, in turn, causes other elements of the climate system – clouds, water vapor, etc. – to also be altered, which then feeds back on the original temperature change by amplifying it (positive feedback) or reducing it (negative feedback). Those feedbacks are what will determine whether manmade global warming will be either lost in the noise of natural climate variability, or — as NASA’s James Hansen believes — catastrophic. Feedbacks in the climate system are much less certain than the radiative forcing from extra carbon dioxide.

All twenty climate models tracked by the IPCC now have positive cloud feedbacks, which is partly why they project so much global warming in the future. Obviously, we desperately need to know what cloud feedbacks are occurring in the real climate system. And since one needs global observations to do that, it can only be done (if at all) during the modern satellite era.

But some researchers now think the search for the feedback “Holy Grail” is a lost cause. This is partly because researchers get different answers depending on which years of satellite observations are analyzed.

But I think I have discovered why. The issue is not a new one to the climate research community, and it is really quite simple: In order to estimate radiative feedbacks, one must first remove any sources of radiative forcing present in the data.

This ‘radiative forcing removal’ technique has been performed before by Forster & Gregory (2006 J. of Climate) to estimate feedbacks during the global cooling which occurred after the 1991 eruption of Mt. Pinatubo. They removed an estimate of the radiative forcing caused by the volcanic aerosols in the stratosphere in order to estimate radiative feedbacks.

Similarly, Forster & Taylor (also 2006 J. Climate) removed anthropogenic radiative forcings from the output of 20 IPCC climate models in order to diagnose the radiative feedbacks operating in those models.

Well, what researchers haven’t accounted for is that there are natural cloud variations in the real climate system, and these also cause radiative forcing. So, in order to estimate feedbacks in the real climate system from satellite data, one would need to first remove those radiative forcings. Unfortunately, this is not easy because those forcings are somewhat chaotic…but, as I
show, they have distinctly different signatures in the data.

Because this ‘contamination’ of the feedback signature by internally-generated radiative forcing by clouds has never been taken into account before, diagnosed feedbacks have been both quite variable (depending upon the time period analyzed), AND they have been significantly biased in the direction of positive feedback.

The result has been the illusion of a sensitive climate system with positive cloud feedback….when in fact the satellite evidence, after accounting for this effect, reveals cloud feedbacks to be negative.

The article I posted here contains the details on all of this, including what I believe to be the most stringent test of climate model feedbacks ever performed. In it I present proof that this natural radiative forcing by clouds is strong, and ever-present…even in the climate models themselves!

I also provide the first evidence that the short-term feedbacks in the IPCC models are substantially the same as their long-term feedbacks in response to anthropogenic radiative forcing — a key finding if we are to ever apply our short-term satellite observations to the long-term global warming problem.

I challenge modelers to address this important issue, because the current, crude level of model testing has NOT been sufficient to validate feedbacks in climate models.

And yes, some of our early work on this issue has been published in the peer-reviewed literature (Spencer et al., 2007 GRL; Spencer & Braswell, November 1, 2008 J. of Climate). Unfortunately, our work is either being ignored or marginalized.

If anyone has a legitimate objection to my arguments in that article, or think something I’ve presented is not clear, e-mail me (see bottom of this page) and I will post your question and my reply here if I think it would be of general interest. I will not reply to comments or questions which are submitted anonymously.