has been scaled to the number
of z-points, 
, in order to filter out the resolution effects and show
the distribution effects instead (scaled to show the actual grid
for 
).
In Fig. 1 I compare the z-scales of sol32, sol63 and sol125,
together with sol100 which differs from the others in the chemical composition.
The grid spacing has been scaled to the number
of z-points, , in order to filter out the resolution effects and show
the distribution effects instead (scaled to show the actual grid
for ).
Notice the ragged for the sol125, which apart from that follows
the sol32 closely. The sol63 on the other hand, has comparatively lower
resolution in the atmosphere but the ``kink'' lies deeper.
Figure 1:
The vertical grid-spacing, as function of depth, z, for the
three comparable resolutions, as well as for the new solar simulation
with an z-scale optimized to capture the temperature structure.
In Fig. 1 I have also over-plotted the
-ratios (on an arbitrary scale) to illustrate the
region which needs the best vertical sampling. Note how the sol100 pressure
ratio lies about 100km deeper than for the three others simulations - most
likely a composition/opacity/mean-molecular-weight effect.
From Fig. 1 we see that comparisons between sol32, sol63 and sol125 are contaminated a little by distribution effects and the ruggedness of the
Figure 2:
This plot shows the maximum of the -ratio
as function of the horizontal (dashed lines) and vertical (solid lines)
resolutions, respectively. +'s indicate the raw data and
's have been corrected for the varying deviations from
the solar 
and 
. 
and 
shows the
position of the sol100 simulation in the diagram. The remaining
differences between this latter and the three other simulations, is the
chemical composition and the optimized z-scale.
sol125 z-scale. The sol100 z-scale has a much better sampling in the high gradient region and is furthermore differentiable, and gives a pressure ratio slightly larger than for sol125. The old and the optimized z-scales are compared in Sect. 3.
When studying Fig. 2, bear in mind that sol100 and sol125 have the same vertical resolution but differ in horizontal resolution which lower the turbulent pressure (cf. Sect. 4) and composition which slightly increases the turbulent pressure (cf. Sect. 5) of sol100 relative to sol125. The rest comes from the optimization of the z-grid, which is further analyzed in Sect. 3.
Figure 3:
, for the three comparable cases. Notice that the
gradient in the photosphere steepens with resolution.
In Fig. 3 I have plotted for sol32, sol63 and
sol125 to judge the effect on the oscillations. The lower 
high
in the atmosphere of sol125 is probably due to bad statistics in the short
9:30 minutes averaging.
Last updated [an error occurred while processing this directive] by: trampedach@pa.msu.edu.