Saturday, January 23, 2016

Redistribution of Enthalpy

So it snows, and Eli digs into the drafts pile to dig this one out.

Eli understands that the old guy is trying to trash him. Eli would be quite happy to talk it over with him, if the old guy wanted to have a discussion, but he appears to want to talk to others without others talking to him and the Rabett is not interested in that. OTOH, we have been hopping about the net and came across the figures from Mark Jacobson's books on Atmosphereic Modeling (link since disappeared) and came across this interesting figure. Looking at it we see that the difference in absorbed incoming solar energy is about a factor of two higher at the equator than at the poles (100% difference) but the emitted outgoing IR radiation above the atmosphere is only about 25% higher at the equator than at the poles.

As near as the team at Rabett Run can make out the large units on the ordinate are 100 W/m^2, which agrees with these measurements of average solar insolation.

What first caught Eli's eye was the implication that radiation and convection move enthalpy from the tropics to the poles. No surprise there. It's called weather. This, of course, is averaged over the year, different cloud conditions and more.

So the bottom line is that even though the temperature difference between the poles and the equator is ~ 50 K, the solar insolation at the surface is a factor of two higher at the equator and the same is true of the emission from the surface as predicted by the Stefan-Boltzmann relation, the amount of energy emitted by the earth in the polar region is only about 25% less than at the equator. Another hmm . . ..


Fernando Leanme said...

Yep, that means the warmer Arctic is a better radiator as long as it stays fairly dry. On the other hand if humidity increases we get more snow, which really helps keep things cool. I'm agnostic, but sometimes I wonder if we may not be inside some sort of super intelligent alien planet design exercise?

Gingerbaker said...


E. Swanson said...

That graph of solar absorbed vs IR emitted has been around for a while. I think it's based on the ERBE instruments which measured TOA insolation and also reflected SW and emitted LW energy. Those measurements do indicate a surplus of energy flow at tropical latitudes and a deficit at polar latitudes. The surplus feeds the energy transfer via sensible and latent energy from the tropics to the poles to close the yearly energy balance, the result, as you note, is weather.

In winter, the warm, wet air flows toward the pole and the colder, dryer air must return to complete the circulation loop. Those of us in the East are enjoying a blizzard when the air masses collide as a result. But, some of that energy transfer occurs by way of ocean currents, especially so in the North Atlantic seen as the Gulf Stream's northward branch and the THC. If the THC slows or stops, the atmosphere would need to "take up the slack" and move even more energy toward the North, with more interesting results, IMHO.

Tom Curtis said...

Bunnies may prefer this version (with units) via Chris Colose:

Anonymous said...

Thanks, Tom. Most bunnies do prefer Chris' version of that graph, because for one it does proffer units on the Y axis.

If I understand it correctly, the latitudes at +/- 35 deg from the equator are net emitters of radiative energy, and the rest of us further north/south are sinks. Is that a correct understanding of our current situation?

Tom Curtis said...

metzomagic, going of the Colose graph, the range -38 to 39 is a net absorber of radiant energy, while outside that range are net sinks. That probably varies a little from year to year based on the North Atlantic Oscillation, and probably other oceanic fluctuations, but I don't know by how much.

Kevin O'Neill said...

Do GCMs provide a gridded time series of this data? It would be interesting to see if there are changes in time by latitude.

What would equilibrium look like for a pre-industrial 280ppm world and what would be expected to change as CO2 increased?

Bob Loblaw said...

Surely I'm not the only bunny that remembers the one-dimensional zonally-averaged climate models developed in the 1960s (independently) by Sellers and Budyko? North has also used models of this sort in more recent times.

A key element of those models was the parameterization of the poleward energy transport by three mechanisms:

1) atmospheric sensible heat (move warm air to the poles, and cold air to the equator)

2) atmospheric latent heat (evaporate water at one latitude, condense it at another) This one is fun because near the equator this generally moves energy towards the equator, not away. Some bunny that remembers how trade winds work and where the big zone of precipitation is will be able to explain why.

3) sensible heat in ocean currents.

Kevin O'Neill said...

Eli - the poleward energy transport I understand.

What I'm wondering about is where the energy is loss/retention differential is best seen/understood.

I.e., Schröder and Connolley (2007) and Tietsche et al (2011) both showed large negative feedbacks from instantaneous loss of sea ice. While recent losses are not of that magnitude, we can assume the same mechanisms exist for the losses we have seen.

So we transport more energy to the arctic, but more energy is also then lost in the arctic. The net sum is still gloobal warming, but are the differentials between lower and higher latitudes changing or do they remain the same? Do GCMs tease out these details?

Oale said...

Ah, some old general schematic of this. I would think we nowadays have some numbers for this.

Matter said...

It's also interesting to look into the hemisphere differences.

The north is on average warmer than the south, but the albedo is about the same and longwave emission is very similar. The north has generally higher clouds though, which support a warmer surface despite approximately equal outgoing longwave seen from space.

A neat demonstration of how cloud height is linked to surface temperature. Some will know that changes in cloud height are expected to play a role in cloud feedbacks.


"sometimes I wonder if we may not be inside some sort of super intelligent alien planet design exercise?"

Spoken like a true patriot- could I interest you in a life subscription to The Daily Sims?

Brandon R. Gates said...

Kevin O'Neill,

Do GCMs provide a gridded time series of this data? It would be interesting to see if there are changes in time by latitude.

Yes. The variables are:


rlds longwave down surface
rlus longwave up surface
rlut longwave up TOA
rsds shortwave down surface
rsus shortwave up surface
rsdt shortwave down TOA
rsut shortwave up TOA

Latent and sensible heat flux, wind stress and others are there too but I don't recall the variable names off the top of my head.

You can get zonal means for individual members/models or the entire CMIP5 ensemble at KNMI Climate explorer under the Monthly CMIP5 scenario runs menu, but you'd have to query once for each latitude band which is tedious.

You could also go after the same variables in NCEP/NCAR reanalysis in NETCDF format at monthly resolution back to 1948 at the surface (and several pressure levels if you're feeling particularly masochistic) here:

From those I was able to fairly easily obtain net surface flux means for each zone at the grid resolution (2.5 degrees). For net latent, sensible, LW and SW at the surface for the entire 1948-present I got an imbalance 0.69 W/m^2 down, which leads me to believe I didn't completely goof the data processing.