From the “one more thing climate models don’t do well” department…
A new study published in Advances in Atmospheric Sciences has shed light on a previously overlooked but common atmospheric phenomenon: the transition zone (TZ) where clouds and aerosols blend together, making it difficult to tell where one ends and the other begins. These conditions in the sky are more prevalent than previously thought and could be a key to reducing uncertainty in future climate projections.
Clouds and aerosols (tiny suspended particles in the air) are critical players in regulating Earth’s temperature, but they remain one of the largest sources of uncertainty in climate models. Traditionally, scientists have classified atmospheric layers as either cloud or aerosol. However, this new research confirms that the real atmosphere is not so binary, often featuring a gradual transition filled with features like wispy cloud fragments or hydrated aerosols that defy simple classification.
The research, led by Jaume Ruiz de Morales from the Universitat de Girona, was honored with the Best Poster Prize at the 2024 International Radiation Symposium (IRS2024) and has been solicited for publication in the symposium’s special issue. The work was conducted by an international team from the Universitat de Girona (Spain), the Karlsruhe Institute of Technology (Germany), and the Universitat de Barcelona (Spain).
“Our findings show that the atmosphere is far less black-and-white than climate models assume,” said lead author Jaume Ruiz de Morales. “Nearly one in ten measurements reveals this ambiguous transition zone. Ignoring it means we might be missing a critical piece of the puzzle in understanding how the atmosphere manages the Earth’s energy budget. Our work calls for questioning how we represent these suspended particles in climate models to achieve more accurate predictions, especially regarding their radiative effects.”
To conduct their analysis, the team used a year of high-resolution data from the CALIOP lidar aboard the CALIPSO satellite. They identified these ambiguous TZ layers as those falling within the no-confidence range of a standard cloud-aerosol discrimination algorithm, plus the fuzzy edges of cirrus clouds.
The global assessment revealed that these transition zones are remarkably common, appearing in nearly 10% of all atmospheric profiles measured by the satellite. The study further identified three distinct types of TZ layers:
- Type 1: Layers classified as “cirrus fringes” by the algorithm.
- Type 2: Found at higher altitudes, with properties intermediate between thin ice clouds and aerosols, resembling high wispy clouds.
- Type 3: Found at lower altitudes, with properties between water clouds and aerosols, such as large, hydrated aerosol particles.
Geographically, TZ layers are found worldwide. Type 1 and 2 layers are most frequent in the Intertropical Convergence Zone and mid-latitudes. In contrast, type 3 layers are predominantly found over oceans off the coasts of West Africa and East Asia, regions known for elevated smoke and dusty marine aerosols.
This high frequency of transition zones poses a significant challenge for current climate models, which typically treat clouds and aerosols as separate, distinct entities, that do not fully represent the gradual change detected in observations of the real atmosphere. The research underscores the need to develop more sophisticated representations of atmospheric particles, moving beyond a simple two-category system to improve future climate projections. The team proposed two strategies to improve such parametrizations. To either include an intermediate phase between clouds and aerosols or to treat all suspended particles in the atmosphere as a continuum of states.
Journal: Advances in Atmospheric Sciences DOI 10.1007/s00376-025-5052-y
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Isn’t this crew a bit late to the party?
Wasn’t “THE” science about everything climate settled back in the Obama days?
(good to see that REAL science is still being carried out though 🙂 )
Truth be told, atmospheric science is barely in its infancy, and all the baby can do is spew shit and cry for milk.
The South Atlantic Anomaly is the weakening of the earth’s magnetic field – currently dropping 5% per 20 years – it means more radiation reaches lower into the earth’s atmosphere. Its a big deal and affects most satellites electronics…..and the atmosphere?
What’s the adage? Unknown unknowns! Poof, there goes settled science!
A certain Mr A Gore pronounced in 1989 that “The science is settled”. As the science has been settled for 36 years, this article must be quashed as misinformation (Ref. COP30).
“Our work calls for questioning how we represent these suspended particles in climate models to achieve more accurate predictions, especially regarding their radiative effects.”
Those “more accurate predictions” for long-term climate studies are pointless, as explained here in a comment I submitted concerning the DOE Critical Review report.
https://wattsupwiththat.com/2025/08/24/10328448/#comment-4110125
In short, ANY time-step-iterated, pre-stabilized, parameter-tuned-to-hindcast, large-grid, discrete-layer model will incur about +/- 4C uncertainty in surface air temperature after a year’s worth of 30-minute steps. This is based solely on the published uncertainty of observed values of TSI. (Yes, TSI also varies, but that is a different issue.)
Don’t get me wrong – I don’t oppose research to improve the understanding of suspended particles and how to simulate their radiative effects. But for policy-relevant projection scenarios, or for diagnosis of historical trends, there is no value provided at all because the buildup of uncertainty is so rapid.
Thank you for listening.
A potential uncertainty of 8 degrees ?
But they can somehow tell down to the decimalhow the global average temperature was 120 years ago with little to no data .
Actually no they can’t. They keep “adjusting” past data.
That’s the big joke – they are telling us what’s going to happen 100 years in the future, while they can’t even decide what the temperatures were in the past, even with the instrument measurements in hand.
If you parameterize the TSI to a constant you’ll *never* get it right. Any uncertainty in that “constant” will compound with each iteration! With a multi-component functional relationship like the biosphere, the variability of each component makes using an “average” value for each unworkable, it will never give the right answer.
“Any uncertainty in that “constant” will compound with each iteration!” That is precisely the point of the exercise I posed, which is more fully described at the comment from August that I linked. Granted, the models can vary the TSI seasonally, corresponding to the elliptical orbit around the sun, but the uncertainty in the central value, of course, remains.
“Our findings show that the atmosphere is far less black-and-white than climate models assume”
Indeed the binary world of the computer climate model: 0 = black, 1 = white…or vice versa… leaves much to be desired
The models are crap, and will continue to be crap until they remove the unsupported assumption that atmospheric CO2 levels drive the Earth’s temperature. Among a hundred other issues.
What about the Russian model that appears to lie close to the observational datasets. I understand that it doesn’t support the CO2 control and applies a large water factor in its atmospheric simulation. This suggests that AGW is rubbish and the hydrological ocean cycle is dominant in moderating natural climate effects.
I hike regularly in the mountains in the South Island of New Zealand.
Sitting at about 2000 to 3000m and watching the cloud form, in wisps and larger volumes, then suddenly it all changes in a few seconds with nothing remaining or constant, starts to give you a good understanding of the complexity of it all. Air temperature, relative saturation, particle density and type, direction of the wind, ice or no ice etc etc etc. A great place to eat lunch with friends.
No computer will ever replicate what is happening, because it does not know how circumstances will change, take the weather reports as a good example.
See you in.or above the clouds.
Yeah, we live at 700m altitude at the edge of a mountain range. I watch the clouds. Multi-layered, often moving in opposite directions; sometimes fog rising. Nobody can simulate this interplay.
A 1degree grid scale means they are using a ~100km resolution for an event (cloud formation) that is on the sub millimeter scale. cGCM results are less like dart throws or spitballs and more like monkeys flinging poo at the walls.
Sounds like a good definition.
“Climate scientist” (n): Monkeys flinging poo at the walls.
Your observation is very true and what is not realized is that you can often get three or four different types of cloud at various levels as well as far different wind speed and directions with changing height. It is not uncommon to have wind blowing from the opposite direction 5000′ higher. When it is considered that the pressure is changing all the time everywhere, that has an effect on temperature and cloud formation.
It is astounding to me that the irreproducible results from climate models, along with their total inability to predict reality make it possible to keep playing with atmospheric modeling.
Yes, we should defund “model-mania”; especially anything using implausible scenarios
like SSP8.5 or 7.0.
Wake me when the remaining models start making reasonable predictions a year in advance of say, the West African or Indian Monsoons, or the N. Atlantic hurricane season. And the Holy Grail would be predicting the month & year along with the strength of the El Nino/La Nina (ENSO).
Until then they are not fit for the purpose of making policy decisions.
(The only thing the models are fit for seems to be to needlessly scaring the population, and keeping the climate modelers & activist journalists employed.)
On cool mornings in the tall PNW forests small clouds rise like smoke from the canyons. They’re called “water dogs” and sometimes are mistaken for smoke by fire lookouts. False alarm. Only clouds forming like magic in the lofty canopies.
We see this at a small scale in the trees across the bay. There it seems to be stimulated by several houses that burn wet wood in the area,
‘The team proposed two strategies to improve such parametrizations. To either include an intermediate phase between clouds and aerosols or to treat all suspended particles in the atmosphere as a continuum of states.’
Better yet, ‘the team’ could just throw in the towel on their beloved radiant transfer models and acknowledge that ‘thermalization of excited GHGs by collisions with non-IR active gas species within meters of the Earth’s surface kills the upwelling radiation field from the surface, hence the work of heat transport from the atmosphere to space is performed by convection until radiation from thermally excited emission (reverse thermalization) can escape to space at higher altitudes.’
https://andymaypetrophysicist.com/wp-content/uploads/2025/01/Shula_Ott_Collaboration_Rev_5_Multipart_For_Wuwt_16jul2024.pdf
Considering that highly parametrized GCMs include only the atmosphere and the ocean iteratively, i.e. 40% of the climate system, leaving out the terrestrial, cryo- and bio- spheres, it is not a surprise that GCMs are of limited value, no matter how high the ‘confidence level’ of the IPCC.
The referenced paper uses the same acronym ‘HC’ for HC clouds and HC aerosols. My guess is HC may mean highly convective clouds and hydrophilic carbonaceous aerosols, i.e. ‘condensation nuclei’. But, I might be wrong.The paper gave no definition that I found. It is annoying to find ambiguous 2 letter acronyms and no glossary
This can be ignored, as the science was settled years ago.
/sarc
It is highly complex, but can be simplified (with all the errors simplification incurs) to this.
It is a thermal engine.
The sun is a variable power source.
The ocean is the energy storage function.
Circulating air, weather, etc. are the engine functioning.
The clouds are the governor.
I would change a little.
The ocean is the large, long-term energy storage function.
The surface is the small, short-term energy storage function.
That is an improvement. Thanks.
You are welcome.
The atmospheric science community hasn’t noticed one of our papers, but before I retired, we found some pretty strange behavior in liquid water cloud droplets that suggested that they had a surface coating with ice-like structural properties. This could lead to droplets forming in smaller sizes than is currently believed possible. Maybe, it could make cloud particles look like aerosols. Who knew?
Hall, C.M., Castro, M.C., Scholl, M.A., Amalberti, J. and Gingerich, S.B., 2021. Anomalous noble gas solubility in liquid cloud water: Possible implications for noble gas temperatures and cloud physics. Water Resources Research, 57(12), p.e2020WR029306.
I think any saltwater sailor could have told them this. In my youth I did my share of lookout watch standing on USN ships on passages of the Pacific. Later I did years of optical characterization measurements on low slant paths in various locations of the North American continent and the Hawaii Islands.
On the East Pacific Range, the USN test range that extends out from the coast of California, the navy conducts high energy weapon (HEL) field tests. That’s where I really became aware of what I called “mists”. These were aerosols, mostly tiny water droplets, that were not organized into clouds – like clouds in the sky or fog banks on the surface. When immersed in this environment, it looked superficially normal – blue sky and a horizon – but low slant path transmission could be a small fraction of what you’d expect. You could barely see a ship or an island a few miles away. I measured all this; it was very important as the laser performance on target was directly affected. What I found weird was the loss, a combination of Nee scattering and occulting absorption, really what you’d expect, just wasn’t noticeable to my eyes. Nee scatter is spectrally flat (white light) scatter from particles much larger than the light wavelength, as from a cloud. So, there was these blankets of aerosols, not clouds with a discernible boundary, but not clear air. I didn’t have much time or charter to wonder about this, but it wasn’t anything new to sea dogs.
Transition zones are not modelled well, or at all, by current GCMs.
I remember my late father doing a study on his boat. He noticed that, even on still days, there were always ripples. And between two ripples lay a tube of air that is surrounded by water on three sides. So he measured the humidity of the air in the trough and the air over the peaks of the ripple.
As I recall, the air in the trough had about three times more water content than the air over the peak – although separated by mere centimetres and thus at the same temperature.
No model factors in changes to ripples. How can you model the number of seabirds landing in the bay or the number of rowers in the river or raindrops in puddles.
But it has a big effect on humidity changes and thus atmospheric heat content.
I’ve looked at clouds from both sides now
From up and down and still somehow
It’s cloud illusions I recall
I really don’t know clouds at all
— Joni Mitchell
From the second full paragraph of the above article:
“Traditionally, scientists have classified atmospheric layers as either cloud or aerosol.”
Hey, whatever happened to “clear”?
That classification might just be important.
Some of the cleanest (i.e., by water droplet and particle count) air in the world exists in the air that descends over and around both the North and the South poles of Earth.
This, from https://www.ecowatch.com/cleanest-air-earth-scientists-2646150814.html :
“. . . the cleanest air in the world is in the air above the frigid Southern Ocean surrounding Antarctica . . . The Southern Ocean that surrounds Antarctica has the air that shows the smallest concentration of particles caused by human pollution . . . The researchers noted that the air that forms the lower clouds in the area was free from any particles produced by man-made activity or dust from other continents.”