OK, so commenter “Norman” asked me at Roy Spencer’s blog to clarify my position on whether a “more IR active atmosphere” would necessarily produce a higher average annual surface temperature at the bottom of that atmosphere. His inquiry in full:
Then you would also agree that increasing GHG in the atmopshere (the quantity makes a difference since it decreases the heat out) will lead to the end result of a warmer surface?
Good. That is what the basic point is all about.
Does the amount of GHG in the atmosphere change the equilibrium temperature of the Earth’s surface?
In your other writings you have states some GHG is necessary but the quantity does not matter. what is your current understanding?
More GHG warmer surface?
Less GHG cooler surface?
Or No change once a certain amount is present?
If in both cases the solar flux to surface remains the same.
So what do we mean by a “more IR active atmosphere”? Well, a proponent of the AGW idea (that of the anthropogenically “enhanced GHE”), like Norman here, would simply say: more “GHGs”. But what does this actually entail? It would lead to an atmospheric column that is more opaque (that is, less transparent) to outgoing surface IR. The idea is that the so-called “GHGs”, the IR active gases (and clouds, mind you), would absorb it more strongly, sort of “capture it” on its way out, and reradiate it in ALL directions, not just the upward one, thus diminishing the net flux of IR moving away from the surface and in the direction of space. And what is this net flux of outgoing IR from the surface? It’s the surface radiant HEAT loss, its Qout(LW).
So Norman’s central claim is this one: “(…) increasing GHG in the atmopshere (the quantity makes a difference since it decreases the heat out) will lead to the end result of a warmer surface (…)”
Well, will it? What does empirically based data from the real Earth system have to say about it?
We return to Africa. My original post discussing this particular issue is here.
Here are my two regions up for direct comparison. Note how the map in Figure 1 shows annual average temps at the surface. It is pretty clear from this map alone that the overall surface temperature within the Sahara-Sahel rectangle (upper one) is considerably higher on average than the one witin the Congo rectangle (lower one); I would say, by around 3K:
The Sahara-Sahel region lies outside of the equatorial convergence zone, but still well within the tropics/subtropics, with a lofty thermal tropopause and a high mean solar input. It sits a bit past midway towards the typically arid zone of the Sahara proper at the bottom of the descending limb of the northern Hadley Cell. The Sahara-Sahel itself is climatically a semi-arid region.
The Congo, on the other hand, straddles the equator, right in the convergence zone, at the bottom of the Hadley Cell’s ascending limb, and is thus very much deep within a highly humid environment:
Now let’s check out the radiative fluxes for each region.
I will use two datasets, one for the surface (indirectly derived values; estimated from various relevant direct observational sources, such as tropospheric temperature, humidity and cloud profiles), and one for the top of the atmosphere (ToA) (directly observationally derived values).
- Surface fluxes: CERES_SYN1deg-Month_Terra-Aqua-MODIS_Ed3A (absolute values)
- ToA Fluxes: CERES_EBAF-TOA_Ed4.0 (absolute values)
First we look at the radiant heat fluxes at the ToA. That’s the net incoming SW (from the Sun to Earth), the ASR (Qin(SW)), and the net outgoing LW (from the Earth to space), the OLR (Qout(LW)):
There’s a lot of interesting information to be found in these plots, but we’ll skip most of it for now, for the sake of leading rather quickly up to the main thrust of our argument.
If you take the incoming minus the outgoing SW (Fig.3 and Fig.4), you get the net value, which is the “absorbed solar radiation” (ASR), the actual solar HEAT input to the Earth:
You will notice how the mean solar heat input at the ToA above the Congo is substantially larger than the similar one above Sahara-Sahel, 286.64 vs. 268.5 W/m2, to be exact, an excess of 18.14 W/m2 in favour of the Congo. And don’t forget now, this is after having accounted for total albedo. ASR (the solar heat) always includes the albedo.
It would be apposite to compare now the Qin(SW) at the ToA (Fig.5 and Fig.6) with the Qout(LW) at the same level. Earth’s heat loss to space is – for all intents and purposes – equal to the OLR (“outgoing LW radiation”) at the ToA. Globally, these two terms should balance in a steady state. Regionally, however, they don’t need to. And normally, they won’t. Let’s have a look:
So what do we see?
The radiative heat balance at the ToA above the Congo looks like this: IN, 286.64 W/m2; OUT, 225.2 W/m2. Surplus: +61.44 W/m2.
The radiative heat balance at the ToA above Sahara-Sahel looks like this: IN, 268.5 W/m2; OUT, 279.25 W/m2. Deficit: –10.75 W/m2.
How can this be? And how can it be sustained in a steady state?
The simple answer is of course that each region also has other heat gains and losses beside the radiative ones. The TOTAL heat balance must always be close to 0 in a steady state, even regionally, otherwise the average temp would just spiral away indefinitely and out of hand. Above the Congo, the tropospheric column appears to “hold back” radiative heat to space, but the energy does in fact escape the region, transported out and away rather via atmospheric divergence high up towards the tropopause (Fig.2). Conversely, above the Sahara-Sahel region, the tropospheric column appears to “over-release” radiative heat to space, but what really happens is that the excess energy thus radiated is transported there advectively from the Equator; from the Congo, so to say.
This process is very well described here:
Air convected to the top of the troposphere in the ITCZ [InterTropical Convergence Zone] has a very high potential temperature, due to latent heat release during ascent in hot towers. Air spreading out at higher levels also tends to have low relative humidity, because of moisture losses by precipitation. As this dry upper air drifts polewards, its potential temperature gradually falls due to longwave radiative losses to space (this is a diabatic process, involving exchanges of energy between the air mass and its environment). Decreasing potential temperature leads to an increase in density, upsetting the hydrostatic balance and initiating subsidence. The subsiding air warms (as pressure increases towards lower levels), further lowering the relative humidity and maintaining clear-sky conditions. However, although the subsiding air warms, it does not do so at the dry adiabatic lapse rate. Continuing losses of longwave radiation (radiative cooling) means that the air warms at less than the dry adiabatic lapse rate (i.e. some of the adiabatic warming is offset by diabatic cooling).
In colloquial terms, one could put it like this: The tropospheric column above the Congo heats (very strongly) from solar input, and cools (weakly) from radiation AND (quite strongly) from advection, while the tropospheric column above the Sahara-Sahel region heats (quite strongly) from solar input AND (weakly) from advection, and cools (very strongly) from radiation.
Well, on to the surface.
The heat IN (Qin) to the surface is again simply equal to the solar heat, the actually absorbed solar radiation, the ASRsfc, which is the downwelling SW minus the reflected SW (albedo) at the surface itself. (Note, we always use the All-Sky values, because this is what represents the situation as it really is (clouds included)):
Figure 9. The downwelling SW to the surface in the Congo is relatively small. The intervening atmosphere (between ToA and sfc) has absorbed about 112 (36%) of the 312 W/m2 worth of downwelling SW entering the Earth system as a whole (through the ToA) that was not reflected by/in the atmosphere itself (~105 W/m2). However, the surface albedo in the Congo also happens to be remarkably low, so as you can see, a relatively small portion of the downwelling SW reaching the surface is reflected back out.
Figure 10. In comparison, the downwelling SW reaching the surface in the Sahara-Sahel region is much higher. The intervening atmosphere has only reflected ~37 W/m2 and absorbed ~89 W/m2 (25%) of the originally incoming SW at the ToA, but at the same time, the surface albedo is much higher than in the Congo, and so a much larger portion of the downwelling SW reaching the surface is reflected back out.
What, then, does this all mean to the average solar heat inputs to the Congo vs. the Sahara-Sahel surfaces?
As you can see, the mean values end up being close to identical. The average solar heat input (Qin(SW)) to the surface is 0.88 W/m2 larger in the Sahara-Sahel region than in the Congo, which means the two are within about 0.5% of each other. Practically the same …
Now we’re finally coming to the crux of the matter. What do two tropospheric columns of such vastly differing IR opacities, one semi dry and clear and one highly humid and cloudy, do to the surface radiant heat LOSS (Qout(LW)) in two regions such as the ones under study? We now know that the average surface heat GAIN from the Sun is pretty much equal in both areas. And we know that this heat gain must somehow, in a steady state, be balanced by the average heat LOSS from the surface.
So what happens when the tropospheric IR opacity goes up? What happens to the heat loss? And what happens to the average surface temperature as a result?
The UWLWIR values are independent of All-Sky vs. Clear-Sky conditions, because they are computed directly from the surface temperature only. In most cases, and it seems to also be the case here, the surface emissivity in these calculations (similar to the ones performed in regular pyrgeometer ‘measurements’) is assumed to be 1. If this is true, then the mean sfc UWLWIR value for the Congo (Fig.13) gives (through the Stefan-Boltzmann equation) an average surface temperature of 26.1 °C. Likewise, the mean sfc UWLWIR value for the Sahara-Sahel region (Fig.14) gives an average surface temperature of 28.9 °C. That’s a difference of 2.8 degrees, close to the 3 that I suggested at the beginning of this post from eyeballing Fig.1. You can go back and verify for yourself. It seems to accord very well with that temperature map …
So what do we end up with? Net LW in both cases. The surface radiant heat loss:
Yup. So our suspicion has been confirmed. The average surface radiant heat loss in the Sahara-Sahel region is about TWICE the average surface radiant heat loss in the Congo.
The highly humid and cloudy tropospheric column resting on top of the Congo surface reduces its radiant heat loss by … A LOT! Compared to the radiant heat loss from the Sahara-Sahel surface, which lies rather at the bottom of a fairly dry, fairly clear tropospheric column.
OK. So at this point the situation at the surface looks like this:
- The Congo: Heat IN (Qin(SW)), 177.96 W/m2; radiant heat OUT (Qout(LW)), 51.08 W/m2. Missing: 126.88 W/m2.
- Sahara-Sahel: Heat IN (Qin(SW)), 178.84 W/m2; radiant heat OUT (Qout(LW)), 103.13 W/m2. Missing: 75.71 W/m2.
Still, we know that the heat IN is balanced by the total heat OUT in both regions. The surface heat loss, after all, is not restricted to radiation. We also have conductive and evaporative losses, Qout(cond) and Qout(evap). In the Congo, these other heat loss mechanisms will have to take care of the 126.88 W/m2 that are left after we have accounted for the radiative loss. In the Sahara-Sahel region, however, they will only have to rid the surface of an additional 75.71 W/m2, about 51 W/m2 less.
So what kind of effect on the average surface temperature would such a striking difference have? If the radiant heat loss of the surface is that much reduced from having an atmosphere on top with a much higher level of IR opacity, then wouldn’t we expect this to somehow result in substantial surface warming? Wouldn’t the surface have to warm in this case so that the other available mechanisms are able to work at a higher level of efficiency, simply forced to by the reduction in the radiative loss?
Isn’t this exactly the idea of how the “enhanced GHE” is supposed to force the surface to warm towards a new and higher equilibrium temperature, to make up for a reduction in radiative heat loss?
So how come the average surface temperature in the Congo is almost 3 degrees LOWER than in the Sahara-Sahel region!? With as much heat coming IN, but with much, much less radiant heat going OUT. Why hasn’t this circumstance forced the average surface temperature in the Congo to be much HIGHER than in the Sahara-Sahel? As per standard AGW ‘logic’.
- The Congo, sfc:
– Heat IN, 177.96 W/m2.
– Heat OUT, [51.08 (net LW) + 126.88 (cond+evap) =] 177.96 W/m2.
– Heat balance.
– Mean temp: 26.1 °C.
- Sahara-Sahel, sfc:
– Heat IN, 178.84 W/m2.
– Heat OUT, [103.13 (net LW) + 75.71 (cond+evap) =] 178.84 W/m2.
– Heat balance.
– Mean temp: 28.9 °C.
Increasing the IR opacity of an atmospheric column sure reduces the radiant heat loss from the surface underneath. But it most certainly doesn’t thereby necessarily force the average surface temperature to rise. Because there are also factors OTHER than pure radiation involved.
Theory, meet Reality.