Zróżnicowanie przestrzenne i zmienność wieloletnia występowania burz w Polsce (1949-1998)

Abstract

The main aim of this study was to present the thunderstorms’ spatial and temporal occurrence in Poland in the period 1949—1998. The paper also attempts to assess an impact of atmospheric circulation on the variability of number of days with thunderstorm. At the same time the study fills a gap in Polish climatological research into thunderstorm occurrence and variability based on the long-term series. The research was based on records on thunderstorms from the period 1949—1998, gathered at 56 meteorological stations which belong to the Institute of Meteorology and Water Management. The main indicator analysed in the study was the day with a thunderstorm. In the second part of the study, devoted to the influence of synoptic situations during which the thunderstorms occurred, the types of synoptic situations elaborated by B. Osuchowska-Klein (1978, 1991, and 1998) for the period 1949—1998 were used. Other sources used in the project were the types of synoptic situations by T. Niedźwiedź (1999) for the period 1949—1998, indices of atmospheric circulation by Z. Ustrnul (1999) for the period 1949-1995, index of North-Atlantic Oscillation (NAO) from the period 1949—1997 (North Atlantic Oscillation, 1998) and synoptic maps from the Institute of Meteorology and Water Management published as Biuletyn synoptyczny PIHM (then IMGW) (Biuletyn..., 1959—1979), Mapy synoptyczne (Mapy..., 1952—1981) and Codzienny biuletyn meteorologiczny (Codzienny biuletyn..., 1980—1981) from the years 1952-1981. The analysis of 50-year series of number of days with thunderstorm, carried for stations from all over Poland, revealed some aspects of spatial variability of that element. It also allowed to find areas with different probability of the thunderstorm occurrence and to define the differentiation of noted numbers of days with thunderstorm in shorter observation periods. Spatial analysis of the occurrence of number of days with thunderstorm in Poland showed that even though in Poland there are 24 such days on average, the number varies from 15 in North-West to 33 in South-East of the country. The largest values of the index are noted at the stations in the Carpathian Mts., e.g. 54 days in 1963 at Kasprowy Wierch Mt., while the smallest — at the Baltic Sea seaside, e.g. 4 days in 1976 in Ustka and 5 days in Kołobrzeg and Gdańsk (in 1976 and 1977, respectively). It turned out, however, that in spite of high number of days with thunderstorm in southern Poland, its inter-annual variability is much smaller than for northern areas. In shorter observation series and in particular years, the spatial pattern of number of days with thunderstorm may vary significantly from the mean one from the multi-annual period. Therefore the impression of increase or decrease of the frequency of thunderstorm occurrence at a certain station depends to a large extent on its geographical location and on the weather type prevailing in a certain year. It was also found out that only at the seaside and in the mountains, where the local conditions determine significantly the number of the days with thunderstorm, their values are always the lowest and the highest in the country, respectively. Spatial distribution of number of days with thunderstorm in seasons and months is related to annual cycle of thunderstorm occurrence. In winter, at all stations, days with thunderstorm occur sporadically, while in summer their distribution is similar to the annual one. Transitional periods of spring and autumn are characterised mainly by higher number of days with thunderstorms in spring, but also typical spatial distribution of the element in autumn: the highest number of days with thunderstorm is noted at the seaside, not in southern Poland. Special attention was paid to the multi-annual variability of days with thunderstorm occurrence. It was marked not only in annual and seasonal changes of the element in Poland and at the particular stations; also different types of its temporal and spatial patterns were found. No tendency was found in the multi-annual course of mean yearly number of days with thunderstorm in Poland, and the values usually do not differ much from the value typical for the climatic-geographical region in which Poland is situated. Periods with significantly higher number of days with thunderstorm are also difficult to find. It can only be stated that at the beginning of the analysed 50-year period the yearly numbers of days with thunderstorm were the smallest and the mean 50-year value was exceeded most often in 1960’s and 1970’s. The minimum value occurred in 1976 (17.5 days), and the maximum one in 1963 (31.3 days). The only season in which the number of days with thunderstorm increased was winter. This is in accordance with the tendencies observed in the whole Europe and most probably is due to the increase of the western circulation in winter. Multi-annual variability of the number of days with thunderstorm at particular stations differed significantly from the mean one for Poland, especially in southern part of the country (correlation coefficient < 0.20), while it was similar to it in northern Poland and in Wielkopolska Lowland (correlation coefficient > 0.70). Large differences in the number of days with thunderstorm at particular stations were found, even if the stations were located quite close to each other. Again, it is difficult to find periods when number of days with thunderstorm would be significantly larger than the 50-year value in the regions or at certain stations. Apart from the stations in Kielce and WIodawa no increasing or decreasing tendencies were found. Therefore the hypothesis that the last years of the 20th century were marked with exceptional increase of the number of days with thunderstorm cannot be confirmed. The annual variability of number of days with thunderstorm was used to distinguish the period of the highest thunderstorm activity (April-September), when 96% of all ca ses are noted. In Poland, on average, the number of days with thunderstorm reaches from 1.2 day in April to 5.6 days in July. However, more detailed analysis allowed to find out six types of its annual pattern. The types differ mainly in the occurrence of the monthly maximum in number of days with thunderstorm, which was noted from May to July in particular years, but also the element’s pattern in the months of the period of maximum thunderstorm activity was different. The analysis of the frequency of number of days with thunderstorm in particular days of the year proved large influence of the station’s location on the annual course of the thunderstorms occurrence. The smallest differences among stations are found in southern and south-eastern Poland, while the largest ones in north-western areas. In the North the period of the highest frequency of days with thunderstorm is by 15—30 days longer, and the decrease in autumn slower than in the South. The assessment of impact of atmospheric circulation on the occurrence of days with thunderstorm concerned variability of this index, both in space and time. In the study the types of atmospheric circulation conducive to the formation of thunderstorms were estimated. The differentiation of inter-air mass thunderstorm and thunderstorm connected with atmospheric front occurrence was also presented. Synoptic situations have less influence on the thunderstorm occurrence than it is usually suspected. The obtained results suggest that future research into the synoptic conditions of days with thunderstorm should be more concerned with air masses and values of other meteorological indicators. The most favourable are cyclonic situations with the advection of air masses from the South (D) or South-East (F) (probability of the thunderstorm occurrence is then 18% and 19.5%, respectively). The most unfavourable are situations with the high pressure center (G) over Poland (6%). There is little correlation between number of days with thunderstorm and the frequency of particular circulation types. Only for the seaside and southern regions there was an influence of eastern advection on the number of days with thunderstorm, while the influence of the western advection (dominating in mid-latitudes) was marked only in south-eastern regions. The analysis of relations between the variability of number of days with thunderstorm and the changes of the circulation indices proved the regularities already presented. The eastern advection has the largest influence on the occurrence of the days with thunderstorm, especially in southern Poland. In the Sudety Mts. southern advection is also favourable, while the increase of cyclonic situations causes larger number of days with thunderstorm only in north-eastern Poland. The results of the study let us divide Poland into 5 thunderstorm regions, which are characterised by similar variability of number of days with thunderstorm. The division was made using K-means method. The hunderstorm regions were distinguished with three criteria characterizing the occurrence of days with thunderstorm in Poland: multi-annual variability, yearly course of the days with thunderstorm and the frequency of synoptic situations accompanied by thunderstorms. Depending on the criterion, 5 or 6 regions were obtained. However, their limits are very similar to each other. The most significant differences among the divisions are the delimitation of Carpathian—Zamość Region and the division of the seaside into two smaller regions due to the analysis of the variability of number of days with thunderstorm in the multi-annual period. As the regions distinguished according to different criteria were similar, finally the territory of Poland was divided into 5 thunderstorm regions. The most important criterion used to classify a station to a certain region was yearly course of the days with thunderstorms, then the multi-annual variability of the index. The mentioned variability of number of days with thunderstorm was used to distinguish, within the South-Eastern Region, a Carpathian Region, which is characterized with significantly different number of days with thunderstorm and its smallest variability in the multi-annual period. Summarising, it should be stated that in Poland the number of days with thunderstorm is characterised by higher spatial differentiation than multi-annual variability. Apart from some regions, the types of atmospheric circulation do not control the variability of the examined element to a larger extent. Most probably only significant climatic changes could influence its pattern. The obtained results indicate also a great role of the local environment in the thunderstorm occurrence’s variability in Poland.