How much warming does the greenhouse effect provide?

[Android]

I used to answer this simply as "33K", but I don't really believe this anymore.

"Some of the heat given off by the Earth is radiated back into space. If this were the entire story the average temperature of the Earth would be -18°C (-0.4°F), which is about 33 C degrees (59 F degrees) colder than it actually is."

http://www.seed.slb.com/en/scictr/watch/cl.../greenhouse.htm

"It has been calculated that if we did not have our natural greenhouse gases in the atmosphere, the average global temperature would be 33C lower at minus 18C"

http://www.bbc.co.uk/weather/features/az/alphabet19.shtml

I suspect now that this 33K figure is actually not the greenhouse effect at all, but the difference between Earth's average surface temperature and Earth's effective temperature as viewed from space. The Earth absorbs and emits about 240wm-2, so it's effective temperature is 255K (-18C). The Earth's surface is about 15C. The difference is 33K.

That's not the same as the difference between the surface temperature with and without a greenhouse effect. I tried to see how it could be but couldn't get it to fit.

What does Earth "without a greenhouse effect" mean?

I think an Earth without a greenhouse effect is an Earth without gases absorbing longwave radiation in the atmosphere. This could be achieved in two ways:

1) Realistically it's achieved by removing greenhouse gases in the atmosphere. But this would mean far more than a lack of greenhouse gases in the atmosphere which absorb longwave radiation. It would also mean no clouds, which would only be accomplished by a lack of significant evaporation which in turn requires no oceans on the planet and probably not much significant life. Mars without co2 would be a close example. I estimate below that Earth's surface temperature in such a case would be about 20K cooler than present.

2) Unrealistically an Earth without greenhouse effect is achieved by leaving greenhouse gases in the atmosphere but disabling their longwave absorption property. This means water vapor can still exist and so can clouds but water vapor is no longer able to absorb longwave radiation. I estimate below that Earth's surface could as much as 55K cooler than present in this case. Ie the warming provided by greenhouse gases in the current atmosphere could be as high as 55K with cooling they provide about 22K.

For estimating these two cases (and they are estimates based on assumptions and perhaps flawed reasoning) I am basing nearly all of this off the K&T energy budget diagram:

and the paper it's based on:

http://www.atmo.arizona.edu/students/cours...ationBudget.pdf

An Earth without greenhouse gases

First remove the 324wm-2 back radiation from the diagram, that won't exist anymore without greenhouse gases. This immediately leads to a surface emission of 66wm-2 and an average surface temperature of -88C rather than -18C.

Next is to remove clouds, because water vapor is gone. Without clouds 77wm-2 of incoming sunlight would not be reflected. Additionally the paper explains that clouds are reponsible for 7wm-2 of the 67wm-2 sunlight absorbed by the atmosphere. Of the other 60wm-2 sunlight absorbed by the atmosphere the paper explains that H2O and ozone absorb 58wm-2 of it! That leaves just 2wm-2 sunlight absorbed by oxygen O2. The effect of removing greenhouse gases (and as a side effect clouds) is that 340wm-2 sunlight reaches the surface instead of just 198wm-2.

How much of this is 340wm-2 is reflected? I will assume the albedo of this imaginary Earth without clouds, oceans or plantlife is similar to that of Mars, about 0.15 (http://nssdc.gsfc.nasa.gov/planetary/factsheet/marsfact.html).

Last of all is Thermals and Evapo-transpiration on the diagram. From the paper I gather that thermals is heat lost from the surface by convection and conduction, while Evapo-transpiration is heat lost by evaporation and transpiration. We can strike out Evapo-transpiration because there will be no water vapor in the atmosphere. In a cooler world would there be less less convection? conduction? I am not certain on this, but as the atmosphere without greenhouse gases cannot emit energy into space (http://www.netweather.tv/forum/index.php?s...t&p=1328974), energy emitted into space must come from the surface alone. So doesn't this suggest the surface is gaining energy from the atmosphere? In which case heat transfer from atmosphere to the surface is 2wm-2 if the atmosphere is in energy balance.

Putting it all together:

340wm-2 sunlight reaches the surface.

51wm-2 of this is reflected. 289wm-2 is absorbed.

Surface also absorbs 2wm-2 from atmosphere.

Surface is about 268K (-5C)

So based on all this, without greenhouse gases the Earth's surface would be about 20K cooler (very rough estimate given the assumptions)

An Earth where greenhouse gases cannot absorb longwave radiation

Again remove the 324wm-2 back radiation from the diagram. Again this leads to a surface emission of 66wm-2 and an average surface temperature of -88C.

The case against Thermals and Evapo-transpiration still stands. The atmosphere can only lose energy to the surface, so the net energy transfer between the surface and atmosphere must be from the atmosphere to the surface (because the atmosphere still absorbs sunlight). The atmosphere still gains about 67wm-2 from absorbed sunlight, so this must be transferred to the surface.

That leaves the amount of sunlight reaching Earth. This is mostly affected by the clouds which reflect quite a bit. My wild (and perhaps not correct) assumption is that a colder Earth = less cloud cover = more sunlight reaching the surface? I will also assume (more likely correct) that surface albedo increases as ice covers more of the Earth = less sunlight absorbed by the surface. So I assume both effects oppose, but I don't know the net effect of this. But if the net is zero (big assumption) the surface is absorbing about 168wm-2 sunlight as currently but also with about 67wm-2 also absorbed by the atmosphere. Which means a surface emitting 235wm-2 which gives a surface temperature of 253K (-19C).

I assume the coldest case is where Earth is entirely covered in ice. The albedo of Europa, covered in ice is about 0.67 wheras the surface albedo of Earth is about 0.15. A greater than fourfold increase in surface albedo of Earth would mean sunlight absorbed by surface = 90wm-2 * (1 - cloud albedo). This suggests that even if all clouds are removed, the maximum amount of sunlight absorbed by Earth's surface is under 100wm-2 if it is covered in ice. In which case assuming it still gains 67wm-2 from the atmosphere it's gaining only 167wm-2 for a surface temperature of about -40C. With cloud it would be less (but could there be cloud in such a cold atmosphere?). In this case the greenhouse effect on Earth actually provides about 55K warmth.

How can the Earth's surface can be 255K (-18C) without greenhouse effect as claimed?

It seems too much of a coincidence that 255K is Earth's effective temperature.For the Earth's surface to be -18C without a greenhouse effect it would have to be absorbing 240wm-2. The Earth is absorbing about 240wm-2 solar energy, but is the Earth's surface isn't. Is the origin of the 33K figure so simple as a confusion between Earth's surface and the entire Earth?

Conclusions

- I am very skeptical of the 33K figure commonly cited. I am unsure what it is supposed to mean.

- The presence of Greenhouse gases could provide as much as 55K warmth.

- All this is why computer climate models are useful and can be used to answer complicated questions like this. The few mere calculations based on massive assumptions I have done based on a 0D energy budget diagram are insufficient.

But if anyone can make me unskeptical of the 33K again that would be nice.

[Chris Knight]

Apart from intrinsic geothermal heat as the earth still cools from its initial hot origins, there are other heat sources.

One is tidal friction, which has geothermal (through plate tectonics and bodily tides) and atmospheric effects, estimated at 5TWplus. Without the oceans, the bodily tides would become more extreme, resulting in possibly increased volcanic and seismic activity. Without the distribution of mass of the oceans, what was seabed would tend to rise, and the rate of continental drift may increase in order to balance the mass of the continents around the earth.

Another is due to radioactive decay, natural nuclear reactors within the earth.

Another is radiation downwards from the thermosphere and stratosphere, which would become more important if the surface was cold.

Another is heating due to interaction between the earths magnetic field and that of the sun and the solar wind. Without an electrically conductive ocean, any electromagnetic currents would be restricted to the mantle and ionosphere, with resistive heating taking place there.

Another is the frictional heating of solar system debris as it falls through the atmosphere, i.e meteors and meteorites, ca. 3 kilotons/day.

In the absence of water, many endothermic and exothermic chemical reactions would not take place, but oxidation (i.e. electron loss) is commonly endothermic, releasing heat and/or light energy.

Without water, life would be very restricted, if not impossible, so the energy that drives the biosphere would be available as heat.

Its an interesting thought experiment, Android. I really have no clue how much difference all of these processes and any others that may go on under such circumstances would make to the temperature, globally.

Using the moon as a comparator, since it shares it's mean solar orbit with that of the earth, but of course has no atmosphere to speak of, and rotates only once every 27 odd days with respect to the sun, the temperatures are discussed here.

(edited to remove quote)

Edited September 22, 2008 by Chris Knight

[Devonian]

I used to answer this simply as "33K", but I don't really believe this anymore.

....

But if anyone can make me unskeptical of the 33K again that would be nice.

Phew! Or is that brrr....

I've myself have wrestled with and tried to understand this fully for a long, long time. As ever I go with the view I'm not an expert (thus I need to wrestle with it more) but that the experts are experts.

This chap (one of the few who seems to really know this stuff and post it on the net) might be worth asking? Or perhaps one of the RC people?

Edit: the Moon is in the same place, effectively, as the Earth, isn't it's mean temperature -18C?

Edit: this is heavy but helpful?

Edited September 22, 2008 by Devonian

[Chris Knight]

Phew! Or is that brrr....

I've myself have wrestled with and tried to understand this fully for a long, long time. As ever I go with the view I'm not an expert (thus I need to wrestle with it more) but that the experts are experts.

This chap (one of the few who seems to really know this stuff and post it on the net) might be worth asking? Or perhaps one of the RC people?

Edit: the Moon is in the same place, effectively, as the Earth, isn't it's mean temperature -18C?

Edit: this is heavy but helpful?

According to several sources:

The mean lunar daytime temperature is +107deg C. The mean lunar nighttime temperature is -153deg C. The overall mean is therefore -23 deg C.

The moons surface is of course rotating much slower than that of the earth with respect to the sun, so any point on the moon heats for just under 14 days, and cools for a similar time. There is also about half a million miles difference between apogee and perigee with respect to the sun.

There are a few points near the poles which are in perpetual shadow due to crater walls etc.

And of course, no atmosphere to conduct and transfer heat by convection.

Edited September 24, 2008 by Chris Knight

[Android]

That's a good suggestion to ask eli or RC people. It can wait though until they put up a somewhat related post, im in no rush to find this out. I am ok with the 33K figure being ballpark, just wondered if I was missing something obvious.