A summary of the variables in this file is given in
Section 2.4.1.
The parameters in qtcmpar.in are generated by a Fortran 90 program, par.f90.
To recalculate qtcmpar.f90 (and qtcmpar.gs) uncomment the build lines
in the makefile and execute make qtcmpar.f90
It produces the file, qtcmpar.f90. The input reference profile for par.f is
Tqo3.ref which includes SST (K), air temperature (K), water vapor mixing
ratio (kg/kg), and ozone mixing ratio (kg/kg) in 1 hPa interval from 1000 hPa
to 1 hPa.
For calculating a1
, 900 hPa cloud base and 150 hPa cloud top are used.
Below the cloud base, the atmosphere is dry adiabatic; above the cloud base,
the atmosphere is moist adiabatic. Relative humidity profile is a linear
fit to the observed reference relative humidity, which is 85% near surface
and 60% at 150 hpa. Tqo3.ref is used to calculate reference values of the
gross moist stability and the gross moisture stratification.
However, for some sensitive quantities, the values from the single profile
in qtcmpar.f90 are used as a guideline balanced with other considerations.
In particular, reference values of the gross dry stability and gross
moisture stratification, GMsr and GMqr, are assigned in qtcmpar.in based
after sensitivity testing. Relevant factors include: (i) the choice of dry
stability affects numerical stability. Sometimes a larger value
of this parameter is useful to avoid sporadic blow-up after decades
of run. (ii) the moisture tends to adjust nonlinearly in the model so the
climatological value in the warm pool region may be somewhat off the
values in the reference profile. (iii) Estimates of the gross moist
stability M in @CROSSREF[11] differed in absolute value among different
data sets, but the relative value of M/Ms was more consistent and is more
dynamically important. There is thus some freedom to choose GMsr and GMqr
within known constraints to attempt to obtain better numerical stability
and reasonable output values of M and GMq.
In version 2.3 GMsr=3.5 and GMqr=3.0.
Model top in the calculation of all basis functions was set at 200mb
in v2.2 and earlier. This was partly motivated by the values of
that result above 200 mb from the contribution of
in the
vertical integration of the hydrostatic equation and the
impact of these in
and
coefficients. In v2.3, the velocity
basis function
is held constant above 280 mb and the
model top can thus be extended to 150mb, more typical of the tropical
troposphere. Since temperature at the uppermost levels is not strongly
held to quasi-equilibrium, it is reasonable to cap
values there
to prevent impact in the higher order terms. This impacts momentum
transport terms, including effects of the vertical advection of
momentum.