Because 2 + 2 = 5

I'm done with making fun of Skeptical Science, it's shooting fish in a barrel, so I dipped my toe into  Science of Doom again. That blog has scientific pretensions, which makes it  harder to spot the contradictions. But they put hard numbers on things, which makes it easier to nail down the sleight of hand once you've spotted it (which took me a couple of days this time):

Globally and annually averaged, clouds cool the planet by around 18W/m² – that’s large compared with the radiative effect of doubling CO2, a value of 3.7W/m². The net effect is made up of two larger opposite effects:
* cooling from reflecting sunlight (albedo effect) of about 46W/m²
* warming from the radiative effect of about 28W/m² – clouds absorb terrestrial radiation and reemit from near the top of the cloud where it is colder, this is like the “greenhouse” effect.

They are obsessed with 'radiation', It is merely one form of energy, which like most forms of energy, can turn into other forms of energy in an instant. Start with chemical energy in your body. You convert it to electrical energy when you throw a ball into the air. Then watch the kinetic energy of the ball turn seamlessly into potential energy and back again. Let the ball fall to the ground and you get a bit of sound and thermal energy. There is not a fixed amount of 'radiation' in the atmosphere that has to be accounted for; there is a fixed amount of total energy, a lot of which is not thermal energy or radiation.

What we care about is not 'radiation' in itself, we care about thermal energy i.e. temperature i.e. 'global warming'. In plain English, a net reduction of 18 W/m2 incoming radiation means that surface temperatures are about 3 degrees cooler then they would be if there were no clouds. Three degrees seems to be on the low side (because 46 W/m2 is on the low side, a back of theh envelope calculation says about twice as much), but let's accept it for now.

See footnotes for further musings on this fascinating topic.

By magic, they can disaggregate the 18 W/m2 (3 degrees) of cooling into minus 46 W/m2 (8 degrees) of cooling (clouds reflecting sunlight back into space and casting shadows) and plus 28 W/m2 (5 degrees) of warming due to the 'top of atmosphere' effect. While the net 18 W/m2 is probably about right, I don't see how it is possible to disaggregate without making dozens of assumptions and guesses, seeing as both things happen simultaneously and have the same cause.

Example - you leave the fridge door open. You can independently measure the normal kitchen temperature; the temperature of the radiator behind the fridge; and the temperature in front of the fridge. If behind is warmer and in front is cooler, you know that the fridge must be on (or was  turned off shortly before), if they are the same, it must be off. You know that the warming and cooling effects should (nearly) cancel out; you know the fridge's wattage and efficiency; so you can get a good estimate of each effect and a sense-check in both directions. With atmospheric water vapour/warming, you can only measure the temperature of the whole atmosphere. That one measurement gives you no clues as to what the warming and cooling effects of clouds are. If you just take one temperature measurement in our kitchen, nowhere near the fridge, you can't even work out whether the fridge is on or off.

But hey.

I explained the 'top of atmosphere' effect at point 2 here. What it boils down to is that the effective emitting altitude (in this case the upper surface of clouds) is (or would be) pushed up by about 1.5 km. That is a heck of a lot, surely airline pilots would have noticed if clouds are higher than they used to be?

The 'top of atmosphere' is an intellectually and mathematically pleasing theory, but complete nonsense nonetheless. If clouds are higher, then of course they are emitting less radiation. Not just because they are cooler (being higher up), but because the radiation energy which they would otherwise emit has been converted to potential energy (the clouds are higher). Potential energy is not thermal energy and there is no warming as a result.

Common sense tells us that water vapour and clouds must dampen temperature swings, they are largely self-cancelling. The Alarmists insist that more CO2 = higher temperature = more water vapour and clouds; water vapour in turn is a 'greenhouse gas' so this pushes up temperatures even more in a vicious circle. This is clearly nonsense, if water vapour caused more water vapour, the oceans would have boiled dry very quickly (or more likely, never formed in the first place).

The Alarmists get round this (read the article) by saying, aha, the 46 W/m2 (8 degrees) of cooling (clouds reflecting sunlight back out to space) is fixed and unaffected by how much water vapour and clouds there are (not plausible), but the 28 W/m2 (5 degrees) of warming will increase with increasing water vapour and cloud altitudes. So above a certain level of moisture; the warming effect exceeds the cooling effect. Which is of course not borne out in real life or plausible.
----------------------------------
The article also get a good kicking in the comments, well worth a visit and a read.
----------------------------------
Footnotes:

1. Just as noticeable/measurable is that day-night swings are much lower where/when it's cloudy/moist than where/when it's clear and dry. Then there is the thorny issue of distinguishing between 'reflection' and 're-radiation':

a. Clouds don't just absorb and re-emit radiation, it also just bounces off. Clouds physically reflect (i.e. like a mirror) sunlight back up by day (i.e. 46 W/m2) and reflect radiation from the earth back down (particularly noticeable by night, and the latent heat released by condensation adds to this - how do you even disaggregate these effects?). For reflection, this must mean a net cooling.

b. Then there is the lesser effect of re-radiation - clouds are warmed from above by the sun (after deducting the amount reflected) and from below by the ground (after deducting the amount reflected). They radiate equally in both directions (we assume). If the radiation absorbed by clouds from above is more than the radiation absorbed by clouds from below (seems likely), then the half of the total radiation absorbed which is re-emitted back down (28 W/m2) will be less than the radiation which would have hit the surface in the absence of clouds = cooling of surface (duh). If the reverse is true (unlikely but possible), then this is indeed a net warming of the surface. I just don't see how you can disaggregate these separate effects.

2. The 'official' figure is an extra 3.7 W/m2 radiation if CO2 doubles - somebody made up this number decades ago and it is now Alarmist Gospel. This - even if were true, which it isn't - equates to about 0.7 degree of warming at the surface (and cooling higher up). Sorry, still not scared.