The astronomers had not waited for the MMT to realize that the seeing of the 4-m class telescopes developed in the 60s and 70s was much worse than the natural value of the site. The drawback of bad seeing is greater with large telescopes since it tends to offset for many types of observation the advantage of the large aperture.
Great efforts were put in trying to improve the situation. Anedoptical evidence and some early experiments ([Hoag], [Murdin]) pointed to convective heat transfer from heat generating equipment and warm surfaces as the main cause of dome seeing. A "theory" of dome seeing was formulated ([Woolf 79]), in which the phenomenon was described by the rising of bubbles of warm air in the dome. Thus the corrective actions consisted in insulating the surfaces of floors and walls, trying to eliminate by active cooling all the heat produced by electric equipment items on the telescope and elsewhere within the dome. In some cases ventilators were installed to generate a "sucking" effect through the slit, supposedly to counteract the upward motion of the bubbles.
In some observatories a floor chilling system was installed in order to cause a stable vertical temperature gradient inside the dome to damp natural convection. This course of action was taken in particular for the 3.6-m Canadian-French-Hawaii-Telescope (CFHT). The telescope, erected on Mauna Kea on the Hawaii island, which is deemed to be the best astronomical site in the world, had during the first years of operation a seeing in excess of 2 arcsec. As a result of many precautions taken to eliminate heat transfer to the telescope air volume, the average seeing has been improved to 0.6 arcsec ([Racine 84], [Racine 92]).
These improvements, as in other observatories, were almost totally empirical, as only occasional measurements of seeing in association with with physical quantities of the dome atmosphere were taken. For a long time little information was gathered on correlations usable in engineering work and the effectiveness of single actions was often controversial. For instance, an air conditioning system was installed at the ESO 3.6-m telescope to homogenize and lower the air temperature inside the dome. The author measured once the air temperature at 16:00 on a clear summer afternoon and found only 0.1 K difference between the floor and the top of the dome. So the system was certainly effective in eliminating temperature gradients inside the dome. However, the seeing improved only slightly.
Also the effect of dome ventilators was controversial.
[Gillingham 82] reported an improvement (not precisely quantified)
of seeing in the 3.9-m
Anglo-Australian-Telescope after the installation of two
ventilators
with a total rated flow of 40 m
/s; this appeared not to depend on
the direction of the flow, either up or down the slit.
In a later report however, [Gillingham 89] reported on the basis of
records totalling a few hours that blowing upwards
resulted in better seeing than sucking downward.
Nevertheless it was obvious that the various technical changes done in the domes were globally effective even if no clear engineering guidelines could be drawn. Eventually the hypothesis was expressed that a large part of dome seeing was in fact mirror seeing, generated in the immediate vicinity of the surface of the primary mirror ([Woolf 88]). This was confirmed by the analysis of a large data set taken at the CFHT ([Racine 92]). It now appears that the general improvement of seeing at the CFHT was due in a large part to the fact that chilled floor and all the heat extraction precautions implemented were also helping to keep the temperature of the primary mirror closer to that of the air volume, a welcome effect that had not been deliberately sought. Recently the CFHT observatory started a project to control more directly the temperature of the mirror.
In the meantime the MMT had already announced its good news. It then appeared that seeing could be dramatically reduced also by letting the wind blow away any natural convection flows arising from the mirror and the surfaces of the telescope room.