“…One of the most important meteorological variables affecting both the formation and degeneration of lake ice is air temperature (Williams, 1971;Palecki and Barry, 1986). Consequently, historical information on both freeze-up and break-up dates can represent a valuable source of proxy data for air temperature (Arakawa, 1954;Simojoki, 1940;Gray, 1974;Tanaka and Yoshino, 1982;Pfister, 1984;Gordon et al, 1985;Tramoni et al, 1985;Palecki and Barry, 1986;Ruosteenoja, 1986;Skinner, 1986Skinner, , 1993Kuusisto, 1987Kuusisto, , 1993Robertson et al, 1992;Assel and Robertson, 1995). Some authors (e.g., Robertson et al, 1992;Assel and Robertson, 1995) are of the opinion that historical lake ice records may be a more reliable indicator of past local and regional climatic changes than even air temperature records themselves, since the latter are frequently subject to inhomogeneities and bias resulting from station alterations and observer changes.…”