Fig. 7. (a), and (b) are vertical temperature profiles above the tropics
(30°S-30°N). The coldest pressures are shown after labels. (c), and (d)
are zoom temperature profiles in tropics (30°S-30°S). The coldest
temperatures are shown after labels. Blue solid lines are 1955-1964 mean
values and purple dash lines are 1996-2005 mean values.
Considering the zonal asymmetries shown above in the zonally resolved
temperature trends, we provide fields of the trends in eddy geopotential
height and stream function in Figs. 8a and 8b to further explain how the
tropopause changes in a zonally resolved view. Climatologically, the
zonally asymmetric patterns in tropopause temperature are dominated by
equatorial planetary waves in a bottom-up regime (Grise and Thompson
2013). The dynamical regime also applies here. The trend fields for eddy
geopotential height and stream function show a Gill pattern associated
with Rossby-Kelvin waves, which lead to large ascent and adiabatic
cooling in the tropopause (Figs. 8c, 8d). The zonally asymmetric
patterns in circulation and temperature fields are more evident in CAM5
than those in WACCM4, which naturally leads to a lower decrease rate in
the zonal mean tropopause temperature in CAM5 (Figs. 7, 8) (Fu 2013). We
further proved the long-term changes in the coldest point regions (CPR)
in response to the IPWP warming, which is defined as the coldest 10%
between 30°S and 30°N (Garfinkel et al. 2013a; Zhou et al. 2021) and is
a key factor for the water vapour entry in a zonally resolved aspect.
The CPR temperature decreases at the rate of 0.318 K per decade and
0.347 K per decade in WACCM4 and CAM5 respectively. Because CPR
decreases much more in CAM5 than in WACCM4, it causes CAM5 SWV entry to
decrease more with 0.025 ppmv per decade in WACCM4 and 0.037 ppmv per
decade in CAM5 (Fig. 8e and 8f). Thus, the quantification of the IPWP
warming impacts, using both models and observations, shows the
fundamental contribution of the IPWP warming to the drying trend in the
lower SWV entry.