Something is rotten with the GHE
By Erich Schaffer
The greenhouse effect (GHE) is a well established theory which most people consider a solid fact, even those who are otherwise “critical” over global warming. On the other side there are some voices who “deny” the GHE with flatearther-like arguments, which seemingly only adds to the credibility of the theory. This is a very odd situation, since the are huge issues with the GHE hidden in plain sight.
“Without GHGs, the Earth would be a frozen planet with a temperature of only -18°C, or 255°K”. This definition is all too familiar to us all and the experts naming it are legion. The 255°K isthe result of a (relatively) simple formula.
(342 x ((1-0.3) / 1) / 5.67e-8) ^0.25 = 255
342W/m2 is the amount of solar radiation (the exact number may vary), 5.67e-8 is the Stefan-Boltzmann constant and ^0.25 (or the 4th root) represents the Stefan-Boltzmann law according to which radiation is a function of temperature to the power of 4.
Black body assumption trouble
The interesting part however is (1-0.3 / 1). 0.3 is the albedo of Earth and 1-0.3 os thus the absorbtivity, which is the share of solar radiation the Earth absorbs (~70%). The 1 below the comma, which is usually omitted, represents emissivity, which is the share of LWIR emitted by the Earth relative to a perfect black body of the same temperature. In other words, it is being assumed Earth would be emitting just like a perfect black body if it were not for GHGs. And that is where the trouble starts.
The basic problem
Quite obviously there are two factors that “violate” the assumption named above.
- The surface of the Earth, mainly consisting of water, is not a perfect emitter, pretty much like any real surface. Although it is not the scope of this article, it can be shown there is a significant deviation from 1 (in the 0.91 to 0.94 range). One needs to look up Fresnel equations, the refractive index of water and so on to sort out this subject.
- Clouds interfere massively with LWIR emissions. Actually this is common wisdom, as “clear nights are cold nights” and most people have made the according experience. Even the IPCC states clouds would block a 50W/m2 of SW radiation, retain a 30W/m2 of LWIR and thus have a net CRE (Cloud Radiative Effect) of -20W/m2 . Of course those -50W/m2 of SW CRE are already included in the formula above (part of the 30% albedo), while the 30W/m2 of LW CRE are not.
For this reason we need to make some minor corrections to the GHE as presented above. Basically Earth receives some 240 W/m2 of solar radiation (= 0.7 x 342) and is meant to emit some 390 W/m2 at 288°K at the surface. Next to a temperature of 33°K, the GHE would thus amount to about 150W/m2 respectively (=390-240).
Since in reality the surface is not a perfect emitter, the 390W/m2 are totally inaccurate. In fact it is easy to call it “fake science” whenever someone claims this number, or an even higher one. Rather we need to reduce this figure by at least 20 W/m2 to allow for a realistic surface emissivity. Next we need to allow for the 30 W/m2 that clouds provide and thus our GHE shrinks from 150 W/m2 to a maximum of only 100 W/m2 (150 – 20 – 30).
For the sake of clarity we should rename the GHE to GHGE (greenhouse gas effect) as this is the pivotal question. How much do GHGs warm up the planet? It is important so see that GHGs were attributed with their specific role in a kind of “diagnosis of exclusion”. If it were not for GHGs, what would be the temperature of Earth? Any delta to the observed temperature can then be attributed to GHGs.
Such a “diagnosis of exclusion” is always prone to failure, be it in the medical field or anywhere else. Essentially a large number of variables need to be taken into account and the slightest mistake in the process, will necessarily cause a faulty outcome. For that reason it should be considered an approach of last resort, maybe helpful to treat a patient or solve a criminal case. As a starting point in physics it is a no-go, and as we can see, it delivers wrong results. But maybe that is the reason why it was chosen in the first place. Faulty approaches give a certain freedom of creativity.
GHGE being notoriously exaggerated
Still, we have not broken any barriers so far. Yes, the GHGE is notoriously being exaggerated and anyone who claims Earth would be 255°K cold if it were not for GHGs, is either incompetent, or simply lying. You cannot excuse such a claim as “simplification”, since exaggerating the GHGE by some 50% at least is certainly beyond negligible.
On the other side, this does not deny the global warming narrative at all. One might consider downgrading climate sensitivity a bit, which would only result in climate models better matching reality. Even then, this will only put things on a healthier and more appropriate basis, eventually supporting the theory of CO2 induced global warming.
So far I have not introduced anything substantially new, but only pointed out to what is known and yet constantly forgotten. Especially the CRE in its quoted magnitude is pretty much an undisputed fact of science. Although I do not know exactly what the origins of these estimates were, experts like Veerabhadran Ramanathan already zeroed in on it in the 1970s. Satellite driven projects like ERBE or CERES later confirmed and specified those estimates.
The net CRE of -20W/m2 thus can be found in the IPCC reports, NASA gives detailed satellite data on it, and even “sceptics” like Richard Lindzen name and endorse it. Such a solid agreement is not just good for my argumentation above, it is also great for the GHGE itself. In fact the negative CRE is pretty much a conditio sine qua non. If clouds were not cooling the planet, the scope for GHGs might become marginal.
There are indeed some issues with the CRE I need to talk about and things are not nearly as settled as I just suggested.
- Whatever experts name a net CRE of about -20 W/m2, they refer to the same sources, which are ERBE and CERES satellite data.
- This are not satellite data at all, but models which are getting fed with some satellite data, among others.
- These models were largely developed by the same people who predicted the negative CRE in the first place. They might not even have a (significant) GHGE if the result would not turn out how it did.
- A closer look on these model results show totally inconsistent outcomes over time. Regions with massively negative local CREs turned into having positive CREs, and vice verse.
- The only thing which really held constant over time was the overall negative CRE of the named magnitude. Of course, that is a precondition to the GHGE and cannot be put into question, if “climate science” wants to have an agenda.
There is yet another side to it. Obviously the net CRE is the sum SW and LW CREs, which can easily be formulated as CREsw + CRElw = CREnet. Since the CRElw is what is being forgotten so notoriously (as it diminishes the GHGE), we could assume there might be a motivated tendency to minimize the CRElw. Given the logical restrictions, this can be achieved by making the CREnet as negative, and the CREsw as small as possible. In other words, there is a trinity of issues with the CRE.
- The -50 W/m2 of SW CRE. This figure is pretty low as compared conventional wisdom, according to which clouds make up for about 2/3s of the albedo, or almost -70W/m2.
- The net CRE of some -20 W/m2. We are going to have a look into this hereafter.
- The LW CRE of +30 W/m2 which is reducing the GHGE as shown above, but for some strange reason tends to be “forgotten”.
Putting things to the test
Since the net negative CRE is “confirmed” by nothing but models of dubious nature, since logic might suggest the opposite (to cut a long story short) and the whole GHGE theory totally depends on it, this question made a perfectly legit target for fact checking. It is the one pivotal question it all boils down to. Is the CRE negative indeed and how could we possibly put it to the test?
As on my previous works in forensic science it seemed mandatory to pass by any conventional approach subject to predictable restrictions. Rather you will have to go beyond the understanding of those who might conspire so that their possible defences turn futile. And of course this would require brute force of intellect, creativity and a bit of luck to find an appropriate leverage.
At least the latter turns out to be a friendly gift by the NOAA. Under the title “QCLCD ASCII Files” the NOAA provided all METAR data from US weather stations. Regrettably they pulled these valuable data from their site soon after I downloaded it, and the alternative “Global-Hourly Files” is not quite working.
The METAR data, as far I understand, are taken at airports and contain, next to usual meteorologic data, cloud conditions originally meant to assist aircraft operating around these airports. The data are anything but perfect for our scope and are subject to a couple of restrictions. As a rule, cloud condition is only reported up to 12,000 ft, yet individual exceptions may occur. Then this cloud condition is reported in 5 different “flavours”, which are CLR, FEW, SCT, BKN and OVC, or combinations of which. For our purpose any combination will be reduced to the maximum cloud condition.
Even if this is not an ideal data pool, it meets a lot of necessary requirements. First it is a totally independent data source, which has never been meant to be used for climate research. Second these data have been collected by many people, who may have made individual mistakes in the process, but were certainly not systemically biased. Third these data are thus “democratic” in nature, not controlled by the bottle neck of a few experts. Eventually, and that is the most important point, we need no models here, but we can look straight onto the empiric evidence.
First I need to tell how such basic research gives you amazing insights otherwise not available anywhere.
You will not get to see what you like to, or expect to see, but what there is. Just like Christopher Columbus searching for India and finding America, you will have to take things for face value. Analyzing the data back and forth, using different perspectives, this was not a simple look up to confirm a certain expectation, but rather a process of continuous learning. Accordingly there are lots of results giving excellent insights into the nature of clouds, or their impacts on climate respectively.
This graph was taken from Harvard’s educational site  on the subject. Here, like in later iterations of the ERBE / CERES modelling, the northern Pacific is meant to be one of the areas with a massively negative CRE, which are of special interest to me.
Since the Aleutian islands are US territory, my NOAA data set included 10 stations located right there.
For the years 2016 and 2017, these stations report about 325,000 valid datasets, with almost 60% of which being overcast. So it is indeed a very cloudy region.
Once we resolve the cloudiness/temperature correlation by season, we find a very typical outcome. OVC skies are correlated with lower temperatures in spring and early summer, and with higher temperatures throughout the rest of the year. This pattern has been seen in all subsets featuring distinct seasons and is due to surface temperatures lagging behind solar intensity.
It is an analogy to the day/night cycle, where clouds hold down day time temperatures, while keeping nights relatively warm. It is about the relation of incoming SW to outgoing LW radiation. As clouds interfere with both radiative fluxes, their primary effect will be relative to which of these fluxes is stronger. In spring surface temperatures lag behind solar intensity, LW emissions will be relatively weak and thus clouds are cooling. In autumn this relation naturally reverses and then clouds are warming.
Note: These “tidal effects” are a direct representation of the LW CRE. Although it goes way beyond the scope of this article, such data are very helpful in assessing the actual magnitude of the LW CRE.
A huge surprise
Finally, if we add up the above results and look at the annual average (thus seasonally adjusted), we are in for a huge surprise (or possibly no more at this point). The correlation between clouds and temperature is strictly positive. The more clouds, the warmer it is, and that is in a region where models suggest a massively negative CRE.
Obviously something is totally wrong here.
I am confident the METAR data are correct, and I am certain my analysis is correct, since I have gone over it many times and the outcome is consistent with all the different perspectives. Instead, the ERBE/CERES models are wrong when compared to empiric evidence a.k.a. “reality”. That is not much of a surprise given a track record of inconsistent results.
And as much as the Bering Sea looks like “a perfect match” to fact check these models, the problems go far beyond the region. No matter where ever I looked, a negative CRE could not be found.
Just the tip of the ice berg
Yet this is just the tip of the iceberg. Of course you need to check for biases to see how much a correlation also means causation, and there are a few. Humidity, as much as it may serve as an indicator for the assumed GHG vapour, is indeed correlated with cloudiness (78% rel. humidity with CLR, 85% with OVC), but this delta is a) influenced by rain and b) too small to explain what we see.
More importantly, this analysis is all about low clouds up to 12.000 ft and it is undisputed the net CRE turns more positive the higher up clouds are. Then there is the subject of rain chill, which makes clouds look statistically colder than they are. Finally temperatures are sluggish relative to ever changing cloud conditions and we would certainly see a larger delta in temperatures if respective cloud conditions were permanent.
Systemic GHE failure
Unlike what I named before, this is not just a scratch on the GHE theory, but systemic failure. If clouds warm the planet indeed, and all the evidence points this way, the very foundation of the theory is getting annihilated. Not that GHGs might not play a certain role in Earth’s climate, but the size of the GHGE will be only a fraction of 33°K, and one that is yet to be precisely determined.
 5th AR of the IPCC, page 580
 Documentation does not fit the data format, for some reason I am unable to locate temperature readings, the format itself is hard to read, and finally for some reason these data, station by station, do not correspond to those of the “QCLCD ASCII Files”