Part of the density discussion is an extension of one from UKweatherworld; however, it relates to the heat
exchange by convection between two different density mediums via physical contact (wind / water energy exchange).
It is not «conduction» but
exchange of radiation; if you keep your hands parallel at a distance of some cm the right hand does not (radiatively) «warm» the left hand or vice versa albeit at 33 °C skin temperature they
exchange some hundreds of W / m ² (about 500 W / m ²) The solar radiation reaching the surface (for 71 % of the surface, the oceans) is lost
by evaporation (or evapotranspiration of the vegetation), plus some
convection (20 W / ²) and some radiation reaching the cosmos directly through the window 8µm to 12 µm (about 20 W / m ² «global» average); only the radiative heat flow surface to air (absorbed
by the air) is negligible (plus or minus); the non radiative (latent heat, sensible heat) are transferred for surface to air and compensate for a part of the heat lost to the cosmos
by the upper layer of the water vapour displayed on figure 6 - C.
Without atmosphere the surface of the ocean or land would lose o (T ^ 4 — Ts ^ 4)(1) where Ts is the temperature of the space (about 4K) while in the presence of the atmosphere the heat losses are hc * (T — Tl)(2) and o (T ^ 4 — Tl ^ 4)(3) where (2) represents the heat transfer
by convection (inclusive conduction) through the air layer and (3) corresponds to the net flow due to the heat
exchange by radiation, Tl being the mean temperature of the air layer.