In either of the upper level or the lower level cooling experiment,
precipitation at the equator does not occur randomly
but appear as sequences of grid-scale precipitation events
that propagate eastward or westward.
In order to examine the relationship between the
grid-scale precipitation activity and other variables,
we perform composite analysis, i.e., calculate mean fields
referring to the location of precipitation maximum.
The procedure of the analysis follows
that of Numaguti and hayashi (1991).
The way of selecting locations of maximum
precpitation is described in Appendix B.1.
Fig.3.10 is the composite structure
referring to the eastward propagating grid-scale precpiptation
of the upper level cooling experiment (kuo-c).
The structure of the eastward propagating coherent disturbances
described in the previous subsection is clearly shown.
In the vertical section along the equator
(upper right panel of Fig.3.10),
we can observe a westward tilt of the structure as a whole.
The westward tilt is a representative structure of a propagating unstable
mode of the wave-CISK.
A warm region overlaps with the region of the upward motion
(key precipitation area)
in the middle and the upper troposphere.
The overlap between positive temperature anomaly and upward motion
assures positive conversion from potential to kinetic energy.
In the horizontal sections (lower panels of Fig.3.10),
we can recognize that the pressure anomaly is localized around the equator
as was described in the previous subsection.
The wind anomaly is dominated by the east-west component, but
has some north-south component, which converges equatorward to
the precipitation region in the lower and diverges poleward
in the upper troposphere.
It is indicated that the eastward propagating disturbances do not have
a circulation structure identified as pure equatorial Kelvin wave
like. This may be explained by the fact that,
in addition to the effects caused by friction at the surface and
nonlinearity in the upper layer,
the circulation is forced by condensation heating
associated with a precipitation area which is
smaller than equatorial radius of deformation.
Those characteristics of the composite disturbance
and its slow eastward propagation speed
compared to that of the first mode of free Kelvin wave in the troposphere
as observed in subsection 3.2,
imply that
the grid-scale precipitation structure in the upper level cooling experiment
can be interpreted as the cooperative combination
of precipitation activity and equatorial Kelvin wave responses.
In general, this result confirms
that of Numaguti and hayashi (1991)
(see Appendix C. for the detailed comparison).
|
|
|
|
|
Fig.3.10:
Composite structure referring to
the eastward propagating grid-scale precipitation events
for the upper-level cooling experiment with Kuo scheme (case Kuo-c).
(upper left)
longitude time section of precipitation at the equator.
Unit is [kg m-2 s-1]
Black dots indicate the points of reference chosen for the composite.
(upper right)
Longitude height cross section of
temperature anomaly at the equator from the longitudinal mean
(unit is [K]).
Arrows indicate wind velocity anomaly from the longitudinal mean
(the arrows at the lower right corner indicates
reference magnitudes of [30m/s, 4x10^-6 s-1], respectively).
(lower left)
Surface pressure anomaly from the longitudinal mean
(unit is [Pa]).
Arrows indicate wind velocity anomaly from the longitudinal mean
(the arrows at the lower right corner indicates
reference magnitudes of [12m/s, 12m/s], respectively).
(lower right)
Height anomaly from the longitudinal mean
in the upper troposphere (¦Ò=0.23) (unit is [m]).
Arrows indicate wind velocity anomaly from the longitudinal mean
(the arrows at the lower right corner indicates
reference magnitudes of [12m/s, 12m/s], respectively).
|
Fig.3.11 is the composite structure
referring to the westward propagating grid-scale precpiptation
of the lower level cooling experiment (kuo-a).
The composite structure of the westward propagating precipitation activity
differs significantly from that of the eastward propagating (Fig.3.10).
In the upper left panel of Fig.3.11,
the temperature anomaly is weaker and mostly localized
at the location of the grid-scale precipitation.
The upward motion is also weak and its area
overlaps that of the grid-scale precipitation
over the entire troposphere from the low altitudes to the top.
No phase tilt is observed in either temperature
or wind anomaly.
In the horizontal sections (lower panels in Fig.3.11),
the wind anomalies are shown to be much weaker and more isotropic
around the precipitation event than
those of the eastward propagating disturbance in upper cooling experiment.
It should be noted that the intensities of the grid-scale precipitation
in the upper and lower cooling experiments do not differ greatly.
This means that the reason for the significant difference
between the intensities of circulation anomalies in the two experiments
cannot be attributed to the difference of precipitation amounts.
More probable cause can be the difference of the vertical structure of heating.
In the lower level cooling experiment,
the heating associated with
precipitation is stronger in the lower level
where the static stability of the atmosphere is stronger,
while in the upper level cooling experiment,
it is stronger in the upper level
where the static stability is weaker.
Considering the approximate balance between heating associated
with precipitation and adiabatic cooling, i.e.,
the product of static stability and vertical motion,
in the tropical atmosphere,
the vertical motion in the lower cooling experiment
is expected to be weaker than that in the upper cooling experiment.
Those characteristics of the composite disturbance
and the rough coincidence of westward propagation speed
with that of the lowlevel background zonal mean westward wind
as observed in subsection 3.2,
imply that
the westward moving grid-scale precipitation structure
can be interpreted as
moist convective activity occurring at the smallest horizontal
scale resolvable in the model,
i.e., conditional instability of the first kind (CIFK).
The westward moving convective activity is associated
with the positive humidity anomaly which
is advected by the mean easterly wind
and is favorable for its emergence.
|
|
|
|
|
Fig.3.11:
The same as Fig.3.10 but for
composite structure referring to
the westward propagating grid-scale precipitation events
for the lower-level cooling experiment with Kuo scheme (case Kuo-a).
|
