A positive temperature difference between the mirror and ambient air creates seeing with a rate between 0.3 to 0.4 arcsec per degree K in calm air. A forced air flow decreases this effect quite rapidly, but the pressure fluctuations may cause mirror deformations, as discussed above. The combined error analysis of seeing + wind buffeting and its consequences on the design and operation strategy of the telescope enclosure are treated in section 6.3.1 below. Here we will outline the possible technical means demanded by the problem.
Considering the constraints given by the technology of the mirror and its support system, at the time of writing the only certain technical solution requires the active cooling of the mirror. Two methods, and their combination, are basically available. The first method requires that the entire telescope is kept very cold during the day in the closed enclosure. The second method involves providing a cooling disk under the primary mirror.
All the available data indicate that over-cooling the mirror does not create seeing, at least down to -3K with respect to ambient air. Therefore daytime cooling may be considered sufficient for those sites where the air temperature does not drop exceedingly during the night. A detail analysis on the basis of the site meteo data will be required to assess the suitability of this solution.
Direct cooling of the mirror is potentially more reliable and flexible, although one should still consider its large heat capacity. Because of this thermal inertia, a temperature control of the mirror will likely present limits with respect to the seeing effect.
This leads us to forecast that ultimately the solution to the mirror seeing problem will consist of the unhindered ventilation of the mirror by the wind. This design approach will be made possible by the expected progress in extended active systems which will maintain an acceptable mirror figure with an airflow up to at least 7-8 m/s.