Rapid cloud clearing phenomenon – another piece of the climate puzzle?


Rapid cloud clearing phenomenon could provide another piece of climate puzzle

Researchers from North Carolina State University have described rapid and dramatic clearing of low cloud cover off the southwest coast of Africa. This newly observed phenomenon could help climatologists understand how clouds affect Earth’s heating and cooling.

While reviewing many years of satellite data from NASA and the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT), NC State atmospheric scientist Sandra Yuter and her team noticed abrupt clearings of cloud cover off the coasts of Namibia and Angola.

“Large areas of low cloud are a normal feature over the ocean in the subtropics west of continents,” Yuter says. “What is unusual in this case is that the cloud erosion occurs rapidly along an organized line hundreds of miles long, like a sun shade being pulled away. The lines can move west for a day or more and clear out a cloud area more than twice that of California.”

The rapid cloud removal events happen throughout the year and peak in May – occurring on over half the days of the month. Yuter and her team noted that the wind direction at cloud level was often perpendicular to that of the clearing. This shows the cloud is not pushed away by the wind.

Video: Movie of 15-minute visible imagery for 26 May 2014 from 05:00 to 18:15 UTC showing (left) regional view off the southwest coast of Africa and (right) close-up view corresponding to the yellow box on the left. CREDIT Sandra Yuter, NC State University. Meteosat satellite data from EUMETSAT.

The clearings start off the coast around midnight and continue through the night and the following day, suggesting that solar heating does not cause these events. “This kind of cloud erosion has never been documented before,” Yuter says. “How it happens is still a mystery, although we theorize that atmospheric gravity waves are the most likely mechanism.

“These waves cause up-and-down motions in the atmosphere, akin to a boat bobbing on the ocean. We hypothesize that offshore winds from Africa interact with stable air over the ocean to make these waves, which then move through the cloud field and promote mixing and cloud evaporation.”

Yuter believes that this discovery will improve scientists’ understanding of climate processes. “We were very surprised by what we found,” she says. “This is a way of changing large cloud areas that has not been documented before. Scientists are interested in anything that changes the area of low clouds over the ocean because these clouds reflect sunlight and cool the Earth.

“Understanding how this happens will help us predict what might happen to clouds in a warming climate,” Yuter adds. “It’s about the balance of warming and cooling effects – and this is one of the pieces that make up that balance.”


The paper: http://science.sciencemag.org/content/early/2018/07/18/science.aar5836


newest oldest most voted
Notify of

Almost appears as a front shearing its way north.


Didn’t Willis expound on this? Though his musings were for the South Pacific


More settled science. Awesome!




More sodding scientific navel gazing.

Gordon Jeffrey Giles

That was funny. Thanks.

Kind of like the fog along the west coast.


I despair! What is wrong with these climate scientists. They think they have discovered some new phenomenon but in reality it is well known, well studied and nothing to see here folks. Move along. Next time ask a Met.


‘Cimate science’ is to actual Science what lame out-of-key drunk Karaoke singing by a Philippino president is to a live-performance by AC/DC.

Steven Mosher

“Next time ask a Met.”

Ok, describe the physics that produces this phenomena
Describe the physics using known physical laws.
next, see if your mathematical decription can be used to retrodict past observations
and then predict future observations.

Bottom line: METS use weather models and observations

please explain the role of c02 in weather models?

99% of METS cannot


I don’t know about this cloud activity but there have been more than a few announcements over the past few years about some “new” discovery that has actually been long known about. Commenters have often provided references to previously published papers about the exact thing, sometimes published decades previously, sometime only fairly recently, but still published well before the “new” announcement. The comment you are decrying could easily be a statement that the thing is already well known even though references were not given.


Every western north to south aligned subtropical coastline both noth and south of the equator has this same cloud evaporation phenomenon.

This is primarily caused by subtropical easterly trade winds blowing off a dry continental land mass meeting shallow maritime subtropical high pressure inversion cloud.

The equatorward wind flow around the semi-permanent oceanic subtropical highs converge with the offshore dry continental easterly wind flow offshore making distinct cloud clear air boundaries.

Throw into this mix, gravity waves formed by descending air originating from elevated continental air parcels trapped under the subtropical high pressure inversion. This causes wave like cloud/clear areas.

Also throw in strong coastal sea breeze circulations due to the inevitable cold sea current along western continental boundaries meeting hot land. These circulations are deflected by coriolis forces along the coast towards the equator in both hemispheres enhancing the warm dry versus moist maritime air boundaries. This is basic bread and butter knowledge for Mets and explains this phenomenon easily, clearly and has been known about for probably 50-70 years.

Not something most of climate scientists would know about as shown with this comment,

“We hypothesize that offshore winds from Africa interact with stable air over the ocean to make these waves, which then move through the cloud field and promote mixing and cloud evaporation.”

Hypothesized something that has been known for decades implies they have not bothered to read or ask someone who knows.


“Describe the physics using known physical laws.
next, see if your mathematical decription can be used to retrodict past observations
and then predict future observations.

Bottom line: METS use weather models and observations”

Bottom line Mosh is that the maths and coding already exist and reasonably accurately predict these cloud events in high resolution weather models. If these Cli-Scis had bothered to check this then this whole article or research would not be needed.

Kristi Silber

So, what’s the phenomenon?


The GCM climate modelling community will get right on this.

Anyone interested in buying a bridge?


Curiosity, you never know where it might lead.


As a technical comment, this appears to be large regional size roll clouds.

“Roll clouds are a rare treat usually awarded to those living in coastal areas. When a large pool of cooler, sinking air near the ground sweeps under moister, warmer air, the contrast can give the air a spin along a long horizontal axis, much like a rolling pin. The cool air may come from a sea breeze or a decaying thunderstorm. A roll cloud is often detached from all other nearby clouds, giving it extra prominence.”

Andy Pattullo

One of perhaps hundreds or thousands of natural phenomena which were not apparently necessary to be understood in the making of climate models. Such models are most impressive for their refined predictive skill in the absence of any knowledge of most of what makes climate tick. Sort of like how I can tell the future from the tea leaves in my cup.

Steven Mosher

Its a common misconception that you need to understand every bit of a process to model it effectively


A good practical example is a simulation of a bomb falling or missile flying

To drive the cockpit displays required to fire missiles or drop bombs accurately
you actually run flyout models of the weapon ( or models of the bomb dropping)

These models are simplifed physics models. They have to be simplified to run in real time.

nevertheless these models are good enough ( despite the missing physics) to get the job
done of prediction where the weapon will end up.

When you look at climate model predictions, they do very well for the relavant metrics


No, they don’t. They’re for sh!t. Worse than worthless. GIGO kludge.

Comparison with dumb ordnance delivery is beyond idiotic.

Gordon Jeffrey Giles

Spare us, please.

David Chappell

Typically utter Mosher nonsense; the information required for airborne weapon aiming is derived from experimental ballistic observations.

Which is a good thing! If weapon aiming was done to the same accuracy as climate modeling, I would be a very worried man. The Goldwater Practice Bombing Range is only a couple of hundred miles away…

Gary Ashe

Windage and elevation,,,,,,,,

Set tele sights up in the barn to 50 yds, 70yd rabbit slight breeze … 1/2inch to the right of and 1/2 inch above the eye,,,,, pop,…….in one eyeball and out the other.

Windage and elevation you see.


It is a question of what precision is required and what phenomena are being studied. For the trajectory of a bomb dropping, the first order contribution will be the flight speed and trajectory at the moment of release. As a first estimate we could assume crosswind effects to be second order.

I the case of laminar/turbulent transition induced by thermal stratification, we can find ourselves in a neutral state which can easly bifurcate down the route of either laminar flow or turbulent flow. After bifurcation the “end” states are hugely different. When you add the sensitivity of the system to pressure changes which pilot evaporation and condensation it becomes clear why it is so difficult to predict precipitation and cloud behaviour.

Some meteo phenomena are easier to predict that others. The main difficulty is neutral equilibria. If the “relevant metric” is related to clouds or precipitation then climate model predictions do not “do very well” or are not “good enough”.


“Natural Equilibria” or neutral equilibria? Either way physical states or math fictions? Least action, physical, yes, but not some algebraic equivalence.


No they don’t – look at the B61-12 upgrade. It is a common misconception that all physical effects are even known. Military hopefully are still aware ot that in spite of Pentagon Innovation now headed by Goole’s Eric Schmidt, but IPCC definitely not.
And the new Russian hypersonics caught the Pentagon off -guard! So much for the allusion.

Andy Pattullo

Steven Mosher, If you believe climate models “do very well for the relevant metrics” I can only conclude that we are looking at different metrics or your standards are very low. Can’t predict regional patterns of temperature/precipitation/drought etc. at all, can’t get the magnitude of global change even close, don’t predict ocean cycles including ENSO which appear to major drivers, not to mention the inner guts of models which you say perform well have almost no similarity to each other across groups of model structures and the scatter they produce looks like random noise. Didn’t predict pause, didn’t predict current cooling of oceans, didn’t predict the current stabilization and even increases in arctic sea ice or ice accumulation in Greenland and Antarctica.

Robert of Texas

It depends on the complexity of the system you are modeling, the number of relevant variables, and the overall contribution of each variable. To model a bomb falling, you need to understand its aerodynamics (lift, drag), its speed, its acceleration due to gravity, and to a small degree the conditions of the atmosphere. Its a fairly basic system, and most of all its not chaotic (the bomb never just decides to fall up instead).

Weather and climate are hugely complex, many more variables, many variables have a large contribution, and is chaotic given enough time (so mostly climate). Chaotic systems defy long term prediction. We don’t even know what variables contribute to climate, let alone the correct weights to use. And given that it is a chaotic system, we can only give probabilities to a long term prediction of a model – it will never be accurate given enough time to run.

I have yet to see a climate model make any prediction that was even close to correct. There may be a few outliers on the fringe, but certainly the Frankensteinian average of their models is not.

Bill Treuren

These bomb flight paths are history matched with actual flight paths and adjusted to accommodate the errors.
I suspect that the history matching of climate models indicate issues as yet not adjusted.
The issue being that the adjustments would render them less frightening. Easier to adjust the measured flight paths of the actual bombs, yeh right, as they say in NZ.
But yes I agree the simplified models could provide important projects into the future.


Steve Mosher is right – the only model that would simulate a system in every detail, is the system itself. Especially if we are talking about complex-chaotic systems. Simulations cannot simulate climate events completely – but they can give us an idea of what processes might be going on.

old construction worker

‘refined predictive skill’? I may be wrong but doesn’t the IPCC say climate models don’t predict outcome? I believe I read someplace climate models breaks so many “Rules of Forecasting” they are nothing more than opinions of their creators.

Steven Mosher

Climate models evaluate scenarios.
When you cannot control the conditions of an experiment you run scenarios.
standard practice is operations research, business modelling, any observational science.
rules of forecasting? forecasting rules and skill scores work fine for high frequency ( like daily) short term ( like 2 weeks) models. Not appropriate for long range scenario


Scenarios along the lines of “If CO2 reaches X ppm and ECS is Y deg/doubling then we can expect temp to increase by Z degrees”, like the IPCC ones? For the argument that follows the number and choice of metrics does not matter.
In such a game, we can’t predict Z because we know neither X nor Y accurately. This is Mosher’s point. So we run it for several pairs of X and Y and see what sort of Z that leads to.
Can we expect that for a given set of real actual values of X and Y close to an envisaged scenario, we should expect a value of Z close to the one suggested by that scenario? If that is not the case, is it permissible to conclude that the dynamics and data embodied in the scenario’s calculation of Z do not match observed reality?

Ian W

The clearings start off the coast around midnight and continue through the night and the following day, suggesting that solar heating does not cause these events.

The timing makes it look like it is an offshore breeze as the land cools down faster than the ocean. So it is caused by solar heating but only as the heat stored in the land is lost while the ocean is still warm. There is a large body of research on such breezes at the coasts. Later in the day the breeze will reverse as the land warms faster than the ocean.


I don’t think that will work. The ocean is quite cold in this area while it is very hot inland, so the sea breeze is virtually constant, even at night. As a matter of fact the dew from this breeze is vital to the Namib desert ecosystems.


How does the speed correspond to planetary movements at the time of obs.?
Gravitational drag lifting the atmosphere?
Nothing new under the sun and the moon has seen a lot too.


This cloud cover is being, in local parlance “burnt off” by the sun as the liquid element is being evaporated; but there is insufficient replacement of the water from surface evaporation. This could possibly be due to very dry winds at the surface surface coming from the dry interior. Elevated wind direction is often often very different from surface winds.

Winds do not necessarily move clouds. Lenticular clouds on top of mountains stay where they are in spite of the prevailing wind. This primarily due to the pressure differences on each side of the mountain.


“This cloud cover is being, in local parlance “burnt off” by the sun ”

At midnight?

By the way the explanation could possibly be the exact opposite of your theory. In this area due to the cold water there is a semi-permanent inversion, i e temperature rises with altitude. Possibly warm air from the interior might spread above the low clouds.

Steven Mosher

he is probably a MET, excuse his errors


Mets are getting Mosh. They know too much.


Yes. My bad. NOT the sun. Probably as you say and as I said later. That fact remains that an energy source is evaporating the liquid element in the clouds, from whatever source and it is not being replaced.
I suppose I was hooked on the local parlance as it were!!

Alan Tomalty

“although we theorize that atmospheric gravity waves are the most likely mechanism.”

These climate scientists are stupider than stupid.

Thomas Homer

” gravity waves are the most likely mechanism”

What was the phase of the moon during this phenomena? Perhaps we should consider potential effects of atmospheric tides before proclaiming ‘gravity waves’ as the ‘most likely mechanism’.

Steve Reddish

“Understanding how this happens will help us predict what might happen to clouds in a warming climate,” Yuter adds. “It’s about the balance of warming and cooling effects – and this is one of the pieces that make up that balance.”

Yuter talks like a new cause of warming has begun that will bring about the warming that models have been predicting. Does she think this is a new warming process that just started up because the climate has warmed to a tipping point and now warming will meet their expectations?


Steve Reddish

“Understanding how this happens will help us predict what might happen to clouds in a warming climate,” Yuter adds. “It’s about the balance of warming and cooling effects – and this is one of the pieces that make up that balance.”

Additionally, if the climate is warming , warming and cooling effects are NOT in balance – unless she means the kind of balance talked about in yoga class.


Smart Rock

Nah, she just says that so the grant-giving organization will fund her next adventure. Doesn’t mean anything, really. You have to put that stuff in or be ostracized by the climate science community.

Richard Patton

This is a **WEATHER** phenomena, not climate. Climate is the ‘average’ of many years worth of data describing the tendency of the weather at certain places. As has been said here many times “Weather is not climate!”

Steven Mosher

Yes. but to get the weather right in a weather model you need to have the physics right.
That physics will tranfer to climate models.. some of the best climate models are weather models run for longer periods and at lower resolution


They’re all for sh!t.

Worse than worthless, GIGO wastes of money, except to show that their assumptions regarding CO2 are obviously all wrong.

Steven Mosher

There are no assumptions about c02. just scientific fact.
facts that have been known for over 100 years
facts that every satellite sensor engineer knows
facts that work.

Luckily science is governed by guys named chimp.
who have opinions.

Ian W

Strange that these models that are so deterministic don’t work with coupled non-linear chaotic systems isn’t it – I wonder why that might be? It is certainly opinionated to claim that the deterministic models work, especially after all their continual and demonstrable failures.
Steven you should spend some time with met forecasters working in areas where getting it right is very important for life and limb, and who work closely with their customers.
Your disdain for them is unmerited.

Pompous Git

“There are no assumptions about c02. just scientific fact.”

The models relied upon by the IPCC are based on future CO2 levels. Different models assume different levels at different times. And you claim these are scientific facts, not assumptions? Makes sense when you believe a”nthropogenic “CO2 behaves differently to “natural” CO2.


Come on Mosh, tell us how to send a missile up the tail pipe of a jet.
I know you know.



My comment was a statement of fact. Which Mosh can’t dispute.

He knows nothing whatsoever of missiles or science. I do.



GCMs make assumptions about CO2 which are not in evidence.

Thus, they are worse than worthless GIGO. Just like you.

spalding craft

Chimp, I ain’t the moderator but you’d be more effective if you calmed down. This is supposed to be a discussion not an ongoing temper tantrum.


Weather models with a 3-5 day accuracy applied to decadal climate? And they are the best? Systems analysis (Bertrand Russell’s dram) gone wild.
It’s not just GIGO, rather Entropy in Entropy Out, creeping in with the maths, No wonder the solutions go ergodic – it’s pre-programmed.
The entire “equilibrium” “balance” mantra springs forth spontaneously in a unknowable way from the maths used, as intended. A logicians fiction.

No, You are wron.g The weather models are valid only for maximum two weeks. They do not even contain the climate change effects like GH gas concentrations.

Pompous Git

“some of the best climate models are weather models run for longer periods and at lower resolution”
Weather models run until the algorithms start generating singularities (infinite values). Doesn’t matter for weather prediction because a fresh model run is started every day that includes yesterday’s data. When my computer algorithms generate infinities, the program crashes.
Kevin Trenberth wrote: “None of the models used by the IPCC are initialized to the observed state and none of the climate states in the models correspond even remotely to the current observed state.” That also appears to contradict your claim since weather models are initialised to an observed state every day.

Johann Wundersamer

These phenomenon is known since

Portuguese henry the navigator –

sea  currents  west  africa:

On the coast of West Africa from equator to South Kap there’s cold winds going northwest.



“The hydrological conditions are peculiar. Not only is the coast dominated by the cold Benguella Current, but in addition to this the upwelling water lowers the temperature still further.” From Ekman (original 1953, Zoogeography of the Sea) writing about what he called the Namaqua fauna. He then talks about the exceptional productivity from the upwelling which sometimes produces mortality from low oxygen. Author does not mention this in abstracts.

Somebody might translate to clouds.


This is research that, for once, is interesting. Have my doubts that it’s anything new, tho. Only newly discovered…..

John F. Hultquist

I think this team has shown something that was not known before, despite the comments here that the comings and goings of such clouds is old stuff. Yes, the clouds were known and also known was that sometimes they were not there.
What is unusual in this case is that the cloud erosion occurs rapidly along an organized line hundreds of miles long, …

If anyone can uncover a similar statement, I’ll buy you a beer, bottle of 2-buck chuck, or a cheap sandwich. So far with 39 commenters and 58 comments nothing much useful has been added.

Sandra Yuter used the phrase “gravity wave” in the (Wiki) sense of “at the interface between two media when the force of gravity or buoyancy tries to restore equilibrium.

It sounds to me like she and her team have discussed what they have seen about every which way they can.
I don’t see this as much of a climate story, but I think it is both new information (for earth science) and interesting.


I explained the mechanism above. It is well known. It is possible the exact words describing the length are not mentioned but who cares. The processes are well understood and known for a long time. Nothing new here.

Bair Polaire

“suggesting that solar heating does not cause these events.”

What about nighttime cooling? The air over land loses more heat during the night. The movie seems to show the Namib desert in Namibia. The cooler airmass slowly moves west over the ocean. Cooler air over the ocean leads to fewer or no clouds. No gravity waves, just cold air.


This phenomenon reminds me of some of the fast transitions seen in thin film Belousov-Zhabotinsky reactions: