Eric,
I gave the paper a quick read and have a few questions of clarification and suggestions.
page 1. The chimney model of Lindzen(1990) and Sun and Lindzen(1993) has already
been pretty heavily criticized, e.g. Betts(1990).
Anyway Hartmann and Larson(2002) argue that the temperature at the detrainment level
should remain constant when the climate changes because it only depends on the
saturation vapor pressure, which is only dependent on air temperature (the Fixed Anvil
Temperature Hypothesis). FAT is holding up very well.
I agree. Citing Lindzen (1990) is sort of beating a dead horse. I’m picking this up from
Held and Soden and from Iwasa, who use the Lindzen hypothesis in contrast to more
complex air parcel trajectories. But, the real point in Iwasa is that w.v. is transported from
tropics to subtropics in a way that preserves RH under warming. The layers somewhat
confirm the mechanisms this model requires. I think it is worth contrasting the layers, as
evidence of transport from lower levels, with detrainment near tropopause. I’ve added
paragraph on Hartmann and Larson to make things more current.
page 4, line 2. It seems that it is not just moist air coming out of convective regions, but
also dry air coming in. So replace ‘out of’ with ‘in and out of’ ?
Changed to:
“This complex structure in the moisture field suggest that the transport of moisture
between the convectively active regions and the dry environment is highly variable and
depends upon complex mesoscale circulation patters surrounding tropical convection.”
page 6, para 3: Were these 10, 380 cloud-free soundings, or was that the total including
the cloudy ones. How many cloud free ones were there?
Clarified:
10,380 is total soundings
Only 8729 soundings passed quality control, so I use this number
Total number of cloudy soundings = 3621
Total number of clear soundings = 5108
Total number of layers
= 2871
Page 7, para 2, Fig.. 4a: Were there really 30 layers per cloud-free sounding? This
sounds too high. Is this worded correctly?
Figures 4a and 4b are reversed – layers are like 0.3 per sounding (or one layer in every
third sounding). Criteria were fairly strict, missing a lot of features that qualify by eye,
but also minimizing features that do not qualify by eye.
30% is UTH.
page 7, para 2, line4-5: In order to see a moist layer, the mean sounding has to be
comparatively dry, right? In order to have a sharp RH gradient the mean must be low, or
you must have dry layers. What is the relationship between the dry layer analysis and the
wet layer analysis?
Dry layer vs moist layer is a matter of viewpoint. It’s really the interleaving of moist and
wet air masses, which is what I try to emphasize in section 2. By forcing the layer to be
less than 200mb thick, I mostly weed out cases of a moist sounding with dry layers.
I concentrate on the moist layers since water vapor feedback is a matter of transporting
moist air out of convection and into the dry environment. Mapes looks at dry layers
because interested in how dry air intruding into the tropics affects convection.
page 8, last line: I think it makes more sense to use the insolation averaged over the day
and the zenith angle, insolation weighted (over the day), than to use a really high zenith
angle. This requires a little bit more fussing with the code, of course. It’s probably not
crucial to the calculation, it just looks funny. The insolation weighted zenith angle is
about 40 degrees at the sub-solar point (Hartmann, 1994, page 32, Fig. 2.8.
See comments on fig 6
Fig. 3 Can you extend this vertically to show some vertical extend outside the layer?
The average layer RH is very high 85%. You must be pretty close to convection.
This is due to funny normalization I did. I took that out, so RH is 30% in environment
and 55% in layer. Composite extends 0.2 normalized pres units above and below layer.
Fig. 4b. Does the dry tongue relate to any known climatological feature?
Fig 4a and 4b reversed…. This is tongue of suppressed layers….
The best explanation I have is that mean meridional winds (ncep) for this period are
northerly along 125E and southerly along 155E. These winds bring dry into the region,
and it is where dry air is supplied that we get layers. This creates asymmetry, and the low
is just the diagonal between.
Fig. 5 So the peak frequency of occurrence at 400 mb is about 13%?
Right. I replotted the histogram with better axes.
Fig. 6 You have not taken into account adiabatic warming due to descent. You can see
this because the RH increases. So I assume that T goes down and p stays the same, since
theta is decreasing in the layer. It seems like the layer has to entrain high theta air from
above or it would become negatively buoyant and convect into the layer below. I think
Dessler and Sherwood did some calculations like this. Complicated to do more than you
have, I guess.
I think this is not worth fixing (ie I don’t have time right now). I think the main point can
be inferred from Mapes et al and the pot temp composite. All I really want to argue is that
layers appear to be stable at the top, which would tent to allow them to persist rather than
get mixed quickly.
Dennis
Betts, A. K., 1990: Greenhouse warming and the tropical water budget. Bull.
Amer. Meteor. Soc., 71, 1464-1465.
Hartmann, D. L. and K. Larson, 2002: An important constraint on tropical cloud climate feedback - art. no. 1951. Geophys. Res Lett., 29, NIL_42-NIL_45.
For GRL need to cut 3300 chars or 2 figs. (~1 page)