EN
The paper constitutes a continuation of the studies of spatial differentiation of the short-term air temperature changes in Poland. The study was undertaken with the aim of identification of the cyclical changes of temperature depending upon the season of the year. The starting point for the inquiry was the set of monthly average air temperatures for the years 1951-1990, coming from 58 weather stations of the Institute of Meteorology and Water Management. The basis for statistical inference was constituted by the chronological sequences of consecutive seasonal three-month averages (Spring: March through May; Summer: June through August; Autumn: September through November, and Winter: December through February). The actual cycles of the air temperature were determined with the method of J.Boryczka, the so called "true cycles" method of identification of the dense spectra of oscillations. The results obtained from the study indicate that the short-term changes of air temperature are one of the features of thermal relations in Poland. The determined spectra of oscillations contain, depending upon the season of the year, between three and eight rhythms statistically significant at the confidence level of 0.95. The most frequently repeated cycles are the approximately 2-, 3-, 4- and 7-year ones. Among the cycles determined the strongest one, i.e. the one with the biggest amplitude and the multiple correlation coefficient, is the 7.7-year cycle for Spring and Winter and the 3.8-year cycle for Summer. Autumn is the season of the year with the greatest differentiation of individual cycles, though for the majority of weather stations the strongest cycle is the 7.0-year one. The scope of changes of the seasonal averages of air temperature values in the cycles identified is the biggest during Winter. The amplitude of temperature in the 7.7-year cycle amounts to 2.8-4.3 degrees, while in the strongest cycles of Spring, Summer and Autumn it does not exceed 2 degrees. The cycles identified do not always appear in the chronological series of temperature over the whole area of Poland. There is quite important spatial differentiation of appearance of individual cycles. The most "stable" with respect to periodical changes of air temperatures is Winter, while the least "stable" - Autumn and Summer. The distribution of parameters of individual cycles (parameters of the regression sinusoids), i.e. of length, amplitude and phase, indicates also their spatial differentiation. In spite of the fact that the paper presents the analysis of the spatial distribution of just the strongest cycles on the area of Poland, significant differences appeared.