Maximum Entropy Production Principle in Atmospheric Circulation

Povzetek

The maximum entropy production (MEP) conjecture holds that nonequilibrium systems with sufficiently many
internal degrees of freedom, such as the atmosphere-ocean system, organize their steady-state heat transport to
maximize the rate of thermodynamic entropy production, but direct tests against observed climate transport remain
limited in scope. We evaluate the MEP conjecture using a latitudinally resolved (5-degree bands, 36 zones) energy
balance model of Earth's climate, comparing MEP-predicted meridional heat transport and zonal temperature
profiles against 30 years (1991-2020) of reanalysis-derived top-of-atmosphere radiation and surface temperature
data. The MEP-optimized model, which selects an effective meridional diffusivity profile that maximizes total
planetary entropy production subject to the observed absorbed solar radiation distribution, reproduced the observed
equator-to-pole surface temperature gradient with a root-mean-square error of 2.4 K, compared to 5.1 K for a
fixed-diffusivity baseline model calibrated only to match global mean temperature. Predicted total poleward heat
transport at 35 degrees latitude, the approximate latitude of maximum transport, was 4.8 PW under the
MEP-optimized model versus an observed value of 5.1 PW, a 6% underestimate, compared to an 18%
underestimate (4.2 PW) for the fixed-diffusivity baseline. Entropy production rate integrated over the whole planet
under the MEP-optimized configuration was 54.2 mW/(m^2 K), within 4% of the value implied by observed radiative
fluxes (56.3 mW/(m^2 K)), whereas the baseline model implied only 41.7 mW/(m^2 K), 26% below the
observationally implied value. Seasonal decomposition showed the MEP model's advantage over the baseline was
largest in boreal winter (RMSE improvement of 3.6 K versus 1.9 K in boreal summer), coinciding with the season of
strongest meridional temperature gradient and largest eddy heat transport. These results provide quantitative
support for the maximum entropy production conjecture as a useful, though imperfect, organizing principle for
large-scale atmospheric heat transport, and suggest that MEP-based closure schemes could improve simplified
climate model representations of poleward energy flux without resolving explicit atmospheric dynamics.