cloud/temperature feedbacks

Professor Roger Davies, University of Auckland, NZ gave a talk
yesterday describing satellite [*] global measurements of cloud
coverage, cloud heights, liquid water content,and reflectivities
for a dozen years since ~2000 [**], and their effects on
temperatures at the surface, the latter estimated by
radiative-convective modeling.

Firstly.  Twelve years is too short: the data have large variances
[**] and very small trends, if any.  Correlations between derived
cloud heights and reflectivities appear marginally significant by
visual inspection, but have [I judge] large errors, both random
and systematic, that were not discussed.  The modeling went by me
too quickly.  Getting old.

Tentatively, however:

From simplest energy-balance models, with T and dT being globally
averaged surface temperatures, and GH referring to GreenHouse
effects:

T(radiative balance) = 250 degK, old stuff
T(surface)= 283 degK, observed
dT(GH/total) =  33 degK, by difference, of which
dT(GH/clouds) =  10 degK, computed by models
dT(GH/gases)  =  23 degK, ditto

[Only the last two of these are germaine to this note.]

Modeled small departures of cloud heights and reflectivities ruffly
cancel one another in their effects on surface temperatures.  [Lower
clouds more GH, and more reflective.]  Davies was concerned about
effects of precipitating clouds with local latent heat release, which
his models do well treat.  And, of course, satellites don’t well
measure cloud thicknesses.

What was interesting to me was the partition of greenhouse effects
of clouds and gases [H2O, CO2, ..], which I and others have too often
ignored.  The large fraction of cloud-contributed GH upon surface
temperatures diminishes the relative contributions of GH/gases and
the relative efficiency of their potential control.  Obvious.  The
coupling between future controls [one hopes] and the future partition
between clouds and gases is less obvious.

Serious modelers have known for years and that variations in cloud-
top heights ruffly cancel effects of correlated variations in
cloud reflectivity.  I’m not quite sure what unique contribution
Davies may be making: again, the data record is very short and the
models incomplete.

’nuff
hh