A Japanese team from the National Astronomical Observatory, Tokio, performed a mirror seeing experiment ([Iye]) with a 62-cm mirror setup which had been originally conceived to verify the feasibility of a new active optics system.
The test setup is shown in fig. 
: the wavefront from a
65-cm horizontal mirror was measured by means of a Shack-Hartmann
analyzer and expanded by Zernike polynomials. The image quality under
the presence of seeing degradation was expressed by evaluating the
Strehl ratio by the expression
where 
is the wavefront rms error.
The mirror was not actively heated or cooled but its temperature as
well as that of ambient air followed slightly different
diurnal variation cycles so that most of the time there was a
mirror-air temperature difference.
In order to measure the effect of forced ventilation, the upper surface
of the mirror was flushed by the air flow created by an electric fan
with a flow-guide nozzle as illustrated in fig. .
The measurements were carried automatically over 90 full days.
Fig. shows the average Strehl ratio measurements
reported by [Iye]
as a function of the mirror-air 
for different flushing
speeds. Some remarks are worth noting:
) becomes more
and more inaccurate for
or 
.
s up to -3K covered by the experiment.
Figure: Measured Strehl ratios versus mirror-air temperature difference for
the 62-cm mirror (as reported by [Iye]).
For the purpose of comparison we have converted the Strehl ratio
values to FWHM seeing values by means of the SuperIMAQ computer
program for the analysis of mirror image quality [ESO].
The seeing values in term of seeing FWHM are then plotted in
fig. .
Figure: Relationship between seeing FWHM and Strehl ratio for a 62-cm
mirror (computed with the SuperIMAQ program).
Figure: Seeing FWHM versus mirror-air temperature difference for
the 62-cm mirror.
