Weather types accompanying very high pressure in Krakow in the period 1901–2000

Authors

  • Zuzanna Bielec-Bakowska,

    Corresponding author
    1. Department of Climatology, Faculty of Earth Science, University of Silesia, Bedzinska 60, Sosnowiec 41-200, Poland
    • Department of Climatology, Faculty of Earth Science, University of Silesia, Bedzinska 60 Street, Sosnowiec 41-200, Poland.
    Search for more papers by this author
  • Katarzyna Piotrowicz

    1. Department of Climatology, Institute of Geography and Spatial Management, Jagiellonian University, Gronostajowa 7, Krakow 30-387, Poland
    Search for more papers by this author

Abstract

The paper presents a classification of weather types observed in Krakow during the 20th century on days with particularly strong highs. The classification was based on daily values of a number of weather elements recorded at Krakow's Historic Weather Station during the period 1901–2000. Days with very high pressure were defined as those with an air pressure at 12 UTC equal to or greater than the 99th percentile of all the cases analysed (≥1037.5 hPa). A slightly modified version of a classification developed by Woś (1999) was used to determine weather types on each of the days identified. Very high pressure was found to have occurred solely during the cold half of the year (October–March). It was mostly accompanied by fairly frosty (9---) or moderately frosty (8---) weather types; subtype: sunny or with little cloud amount and very sunny (-02-) and very cloudy, without sunshine or with little sunshine (-20-); and weather class: without fog (---0). No significant annual or seasonal trends were found in the occurrence of days with very high pressure or in the various weather types. Copyright © 2010 Royal Meteorological Society

1. Introduction

Climatologists have recently given a strong focus to the climate changes observed during recent decades and to their effects. One of the aspects of these changes is the occurrence of specific weather types that are typically linked to the atmospheric circulation changes and to the occurrence of characteristic types of synoptic situation, the time of their occurrence (annual course) and persistence. An insight into the nature of change in weather types and their frequency of occurrence, especially of exceptional cases, in a given area is important both to improve the understanding and for applied purposes. Changing weather patterns force changes in a number of human activities, such as water management, agriculture, construction or transport; they strongly affect human health and well-being and require adaptation.

The term ‘weather type’ is used in various fields of climatology, including in complex climatology (a specialised area of climatology that evolved in the 1920s and the fundamentals of which were developed by Fiedorov and Czubukov; Kozłowska-Szczȩsna, 1965), bioclimatology and in synoptic meteorology and climatology (Olszewski, 1967; Niedźwiedź, 2003). In complex climatology, a weather type is defined as a 24-h course of meteorological elements or their complexes following a pattern. The earliest classifications of weather types so defined were proposed by Fiedorov (Kozłowska-Szczȩsna, 1965), Howe (1925), Nichols (1925, 1927) and Switzer (1925). In Polish literature, this type of classification was proposed by Woś (1977, 1999) and was successfully used by Kaszewski (1992), Marsz (1992), Lotko-Łozińska (1994), Kożuchowski (1996) and Ferdynus (2004). Bioclimatologists use weather types to assess bioclimate, as the frequency of weather situations has an influence on the human body and can be important for climatotherapy (e.g. classification by Fiedorow and Czubukow; Kozłowska-Szczȩsna, 1965; Błażejczyk, 1979, 2004). To a synoptic climatology, however, weather type is defined as a particular type of atmospheric circulation (Stefanicki et al., 1998; Littmann, 2000; Bissolli and Dittmann, 2001; Buchanan et al., 2002; Sheridan, 2002, 2003; Brown, 2004, 2005; Boé and Terray, 2008; Moron et al., 2008; Makra et al., 2009).

Most published research on the weather of a given area investigates one or a few meteorological parameters in terms of frequency of occurrence in certain synoptic situations (Trigo and DaCamara, 2000; Svensson et al., 2002; Brown, 2005; Twardosz, 2007). Other approaches involve an analysis of synoptic situations and finding values of selected meteorological parameters or indices on days when these situations occur (Niedźwiedź, 1981; Post et al., 2002). There is little research, however, that looks at the relationship between circulation types and weather types (Kaszewski, 1992; Bogucki and Woś, 1994).

An example of this approach is provided by Petrovič (1968), who used a modified version of the classification by Fiedorov–Czubukov and circulation types developed by Czechoslovakian Meteorological Service. Maheras (1984, 1985, 1988) used the fundamentals of the Russian classification too. This strand of research is continued by Michailidou et al. (2009a, 2009b). The authors identified weather types by data clustering of two data types: meteorological and circulational, with the latter following the classification by Maheras et al. (2000, 2003). In Poland a similar approach was adopted by Niedźwiedź (1981, 1983), who applied his own typology of atmospheric circulation and the weather conditions. The author used a modified Czubukov classification and identified 16 weather types. Kaszewski (1984, 1992) followed a similar approach in his study of the city of Lublin and then of the whole of Poland using circulation types proposed by Lityński (1969) and weather types proposed by Woś (1977).

This paper continues in the footsteps of the aforementioned Polish climatologists looking at weather types and their relationships with synoptic situations. The objective of this study is to identify the weather types that occurred during particularly high pressure in Krakow during the 20th century. Particular attention was focused on the selection of the meteorological parameters best suited to reflecting the nature of the weather prevailing at a time of very high air pressure values. The research is supplemented with an analysis of multi-annual and annual variability of number of days with very high pressure and the weather types, subtypes and classes identified. This should help to assess the usefulness of the classification obtained, its level of detail as well as its applicability in climate research in Central Europe.

The research is primarily a methodological study designed for further application. It is connected with the fact that strong high-pressure areas are often accompanied by weather types that are both unfavourable to living organisms and can cause serious difficulties with transport, power and water supply. This makes this study potentially useful for agroclimatological and bioclimatological research. It may also be helpful in assessing the degree of preparedness of road or airport maintenance services.

2. Source materials

The study was based on daily values of selected meteorological elements recorded at the Historic Weather Station in Krakow (50°04′N, 19°58′E, 220 m a.s.l.) during the period 1901–2000. The station has been shown (Trepińska, 1997) to offer a good representation of climatic conditions across much of Central Europe below 500 m a.s.l. The following meteorological elements were selected for investigation:

  • (1)Sea level air pressure values at 12 UTC
  • (2)Average, maximum and minimum values of temperature at 12 UTC
  • (3)Average daily cloudiness (in %)
  • (4)Relative sunshine duration (in %) defined as effective sunshine duration (hours with the sun) divided by possible sunshine (hours between sunrise and sunset)
  • (5)Information on fog occurrence.

Days with exceptionally high pressure were defined as days when the air pressure value at 12 UTC was equal or greater than the 99th percentile of all the daily measurements (P99). In Krakow, this condition yielded the value of 1037.5 hPa (P99). The method adopted produced results similar to those obtained with the generally accepted climatological rules, whereas for an anticyclonic system to qualify as exceptionally strong, it must have pressure equal or greater than 1035 hPa (Kożuchowski, 1995). For comparison purposes the study also went on to determine the weather types on days when air pressure was equal to or greater than 1030.7 hPa (P95), which corresponded to the 95th percentile of the daily measurements.

The weather types considered in the study were identified using the meteorological elements mentioned above and a slightly modified classification by Woś (1999). Section 4 offers a detailed description of the work on the classification.

3. Days with very high pressure

During the study period, air pressures in Krakow ranged from 968.4 hPa (26 February 1989) to 1058.6 hPa (23 January 1907), and the mean sea-level pressure at 12 UTC was 1016.0 hPa. Usually both the maximum and minimum values were recorded in winter. It is connected with the strong anticyclones developing in winter to the north and east of Poland over the cooled European continent. Winter is also the time when the number of deep cyclones forming in the Atlantic-European area grows, especially the number of cyclones influencing European weather above 50°N, which are recorded: in the Greenland–Iceland region, a typical cyclogenesis area, in the north of the Scandinavian Peninsula (Trigo, 2006) and deep cyclones (mean pressure < 1000 hPa) formed over the Baltic Sea (Sepp, 2009). During the warm half of the year, because the land warms up strongly, values of atmospheric pressure are lower, and the pressure fluctuations were much lower at a maximum pressure range of 51.1 hPa in May, compared to 82.1 hPa in January (Table I). The regularities presented are also confirmed in other researches on pressure fluctuation in Poland (Fortuniak et al., 2000; Trepińska, 2007).

Table I. Air pressure values in Krakow in the period 1901–2000 (12 UTC)
PressureMonthSeasonYear
 JanuaryFebruaryMarchAprilMayJuneJulyAugustSeptemberOctoberNovemberDecemberApril–SeptemberOctober–March 
Minimum976.5968.4979.8988.7984.2993.6996.3992.0987.7985.7982.1974.3984.2968.4968.4
Average1018.71017.11015.91013.31014.51014.21013.91014.71017.01017.81017.11017.31014.61017.31016.0
Maximum1058.61045.61044.51038.01035.31030.21028.51028.31037.31039.51047.01048.91038.01058.61058.6
P991042.71040.41036.21030.21027.91026.11024.71025.51031.11035.41037.61040.51028.61039.71037.5
P951036.91034.81031.11025.21024.41022.81021.61022.51027.01030.61032.31034.71024.41033.71030.7
P901033.41031.41028.11022.51022.41020.81020.11021.01025.21028.11029.31031.71022.11030.51026.9
P851031.01028.31026.01020.81021.01019.61018.91019.81023.61026.41027.41029.31020.71028.11024.6
P801029.01026.11024.31019.41019.91018.71018.01018.91022.31025.01025.41027.31019.61026.31022.7

When determining criteria for days with very high pressure, the authors investigated the distribution of air pressure values corresponding to the 99th, 95th, 90th, 85th and 80th percentiles of all days considered monthly, seasonally and annually. A comparison of results reveals that differences between the calculated values (P99− P80) can exceed 10 hPa (Table I). It concerns especially the pressure values recorded in the cold half of the year. In summer, the difference is two times smaller. Similar regularities can be noticed when we compare differences between the highest and average pressure values (e.g. 39.9 hPa in January and 13.6 hPa in August) or P80 and average pressure values (e.g. 10.3 hPa in January and 4.1 hPa in July) in individual months. It is easy to notice that pressure values in the anticyclones during the warm half of the year are only slightly different from the average pressure values. This is caused by differences in the origins of strong anticyclones in winter and in summer. While the former develop typically as cold anticyclones over Russia and northern Europe, the strongest of the summertime anticyclones are linked with a northward movement of a tropical high-pressure zone. Both the stronger wintertime high-pressure areas and summer high-pressure areas are linked to weather conditions regarded as highly unfavourable for human health and for various activities, such as agriculture and transport. For this reason, it would be very interesting to compare weather types accompanying strong anticyclones identified separately in the warm and cool halves of the year, or even for individual seasons for that matter.

This study only looks at weather types that accompany the strongest anticyclones each year. To do this, as mentioned before, days on which air pressure values were equal to or greater than the 99th (P99) percentile of all days, i.e. 1037.5 hPa, were selected. The air pressure values were recorded in Krakow, which means that the actual air pressures at the centre of the systems might have been even higher. For this reason and because of the small number of such cases identified, the study also compared the results with weather types on days when the air pressure was equal to or greater than the 95th (P95) percentile of all days, i.e. 1030.7 hPa.

Weather in Krakow is seldom influenced by strong highs. During the 20th century, there were only 4 days on an average in any given year with systems where air pressure was equal to or greater than 1037.5 hPa (P99), but the numbers varied greatly from year to year. The highest number of 16 was recorded in 1932, but there were also 18 years without a single day with very high pressure (Figure 1). The days with very high pressure tended to mostly occur in the first half of the study period (60% of days), and especially during the period 1925–1935 (21%). After 1950, their occurrence was less frequent. An overall result of this pattern is that the number of days with very high pressure is falling, but not to the point of statistical significance.

Figure 1.

Number of days with strong highs in Krakow in the period 1901–2000

No seasonal or monthly trends were noted in the occurrence of days with very high pressure. They tended to concentrate in the winter months: January (38.2%), February (23.3%) and December (22.2%). No days with very high pressure were recorded between May and September (Table II).

Table II. Number of days with high-pressure values in Krakow in the period 1901–2000
Percentile/pressureMonth
 JanuaryFebruaryMarchAprilMayJuneJulyAugustSeptemberOctoberNovemberDecember
P99 (≥1037.5 hPa)1418620143582
P95 (≥1030.7 hPa)48934417725835155235373
P90 (≥1026.9 hPa)7985674091016519413171421500650
P85 (≥1024.6 hPa)1022746587190149763355371673707881
P80 (≥1022.7 hPa)1212914779301287166941475629159121079

Since the beginning of the 20th century, the number of days with very high pressure in Krakow has been highest in winter (December–February) at 83.7%. Examples include the winter of 1992–1993 with 15 days, of 1928–1929 (one of the coldest) and 1931–1932 with 14 days. Special attention should be paid to 1959 when all 13 such days in that year, including 10 days in a row, occurred in February. Very strong high-pressure systems were much less frequent in the transitional seasons where autumn (September–November) accounted for a greater share of days at 10.6% than spring (March–May)—5.7%. Springtime highs were particularly sporadic, with the last such occurrence in the 20th century being one single day in 1992. Similar long-term trends of change are visible when looking at the number of days with high pressure of 1030.7 hPa or more (P95). This number also drops in the second half of the 20th century, although the drop is not statistically significant (Figure 1). The greatest number of 43 such days was recorded in 1932, while the lowest number of just 2 days was recorded in both 1923 and 1966. There is a greater variation in the annual distribution; the days thus defined were clearly more frequent during the warm half of the year than for the P99 criterion and the period in summer when such high-pressure values were not recorded in Krakow only extended from June to August inclusive (Table II). The proportion of days with high pressure occurring in wintertime dropped from 83.7% (P99) to 65.5% (P95) of all days.

4. Method of classification of weather type

This classification of weather types employed a method proposed by Woś (1999) with slight modifications (Table III). It is based on methods used in complex climatology and has a broad-scale character. The method has had numerous applications in Polish climatological research ranging from analyses of weather types in the whole country, in its regions and at individual stations, to finally being employed in characterising the climate of the Island of Spitsbergen (Kaszewski, 1992; Kożuchowski, 1996; Ferdynus, 2004). It was also expanded with the addition of wind speed (Marsz, 1992; Lotko-Łozińska, 1994).

Table III. Classification of weather types and frequency [%] of weather types on days with high pressure in Krakow in the period 1901–2000
SymbolsPartition ( °C)Weather typesFrequency (%)
   P99P95
  1. For example, 3010—very warm, sunny or with little cloud amount, moderately sunny, without fog.

Air temperature
33---tmean > 25.0; tmin and tmax > 0.0Hot
3---tmean 15.0–25.0; tmin and tmax > 0.0Very warm0.1
2---tmean 5.0–15.0; tmin and tmax > 0.0Moderately warm1.410.4
1---tmean 0.1–5.0; tmin and tmax > 0.0Cool4.111.3
4---tmean > 5.0; tmin ≤ 0.0 °C; tmax > 0.0Ground-frost, moderately cool0.7
5---tmean 0.1–5.0; tmin ≤ 0.0 °C; tmax > 0.0Ground-frost, very cool7.616.1
6---tmean − 5.0–0.0; tmin ≤ 0.0 °C; tmax > 0.0Ground-frost, moderately cold18.416.9
7---tmean < − 5.0; tmin ≤ 0.0 °C; tmax > 0.0Ground-frost, very cold1.10.4
8---tmean − 5.0–0.0; tmin and tmax ≤ 0.0Moderately frosty15.514.4
9---tmean − 15.0 to − 5.0; tmin and tmax ≤ 0.0Fairly frosty44.726.8
0---tmean < − 15.0; tmin and tmax ≤ 0.0Very frosty7.32.9
Cloudiness
-0--N ≤ 20%Sunny or with little cloud amount24.718.9
-1--N 21–79%Cloudy40.938.3
-2--N ≥ 80%Very cloudy34.442.8
Relative sunshine duration
--0-Sr ≤ 33%Without sunshine or with little sunshine51.257.4
--1-Sr 34-67%Moderately sunny27.623.3
--2-Sr ≥ 68%Very sunny21.219.3
Fog
---0Day without fogWithout fog64.564.6
---1Day with fog (morning, noon or night)With fog35.535.4

In the weather types classification by Woś (1999), three meteorological elements were considered: air temperature, cloudiness and daily precipitation totals. Air temperature is the climate element that has the strongest and clearest differentiating effect on daily weather conditions. For this reason, air temperature has been accepted as the basis for a majority of classifications of weather type. Because of a high degree of temperature fluctuation, its role in the shaping of physical processes and its impact on living organisms, air temperature is normally broken down into several ranges. The breakdown proposed by Woś (1999) and employed in this study is very universal, including its usefulness for the identification of hot days, days with ground-frost and frosty days (Table III).

Woś (1999) also considers cloudiness to account for the ‘visible’ aspect of the weather, i.e. cloudy versus fair days. This is, however, a limited measure, as mere cloudiness without the type of clouds fails to reflect the full characteristics of the weather conditions. Indeed, even a fully overcast sky may provide much solar radiation as long as the clouds are of the high-level type. A day like this could be classified as fair in terms of insolation. For this reason, as an additional element, relative insolation has also been used in this study (Table III).

The one element taken into account by Woś (1999) that this study excluded is daily precipitation totals. On the days studied, precipitation was very rare in Central Europe, just as in other regions (Ustrnul and Czekierda, 2001). Days without precipitation or with trace precipitation (<0.1 mm in 24 h) accounted for more than 80% of the total days studied and the highest precipitation recorded did not exceed 20 mm (precipitation of 10.1–20.0 mm was recorded only twice).

In our research, to provide a more detailed characterisation of days with very high-pressure fog data were also included. Fog was recorded on 35% of such days and it often lasted for just a few hours per day (Table III). Wind speed was also considered at some stage, but was finally excluded, as the Krakow long-term record of wind speed was non-homogeneous.

In the end this study included air temperature, cloudiness, relative insolation and fog occurrence to identify weather types (Table III). The ranges of temperature and cloudiness adopted remained the same as in the original classification by Woś (1999). Three ranges of equally probable insolation were adopted, and fog was recorded as present or absent (Table III).

In the next step, each day with high air pressure was then coded with a weather type. A four-digit code refers to (1st) air temperature (or first two digits ‘33’ in the case of hot weather), (2nd) cloudiness, (3rd) relative sunshine duration (2nd and 3rd refer to the weather subtype) and (4th) fog occurrence (weather class) (Table III). This approach allows the use of several concurrent meteorological elements to describe the state of the atmosphere. As a result, a total of 198 weather types were identified (198 = 11 temperature classes × 3 cloudiness × 3 relative sunshine duration × 2 fog), but only 68 accompanied very strong highs (Figure 2).

Figure 2.

Frequency [%] of weather types on days with strong highs (P99 and P95) in Krakow in the period 1901–2000

5. Very high pressure and weather types

Very high pressures were normally accompanied by the fairly frosty weather type (9---, 44.7 or 26.8%; Table III), especially:

  • (1)Sunny or with little cloud amount, without fog—9020 (8.7% for P99 and 5.7% for P95)
  • (2)Very cloudy, without sunshine or with little sunshine, without fog—9200 (6.8% for P99 and 2.7% for P95) Sunny or with little cloud amount, moderately sunny, without for—9010 (5.1% for P99 and 3.4% for P95; Figure 2)

This type mostly occurred during winter months and accounted for between 32.6% (February) and 63.4% (December) of very high-pressure days during the study period for P99 and from 29.4% (February) to 45.4% (January) for P95 (Table IV). Other weather types encountered under wintertime strong high-pressure systems mainly included ground-frost, moderately cold (6---) and moderately frosty (8---) (Figure 2, Table IV). They constituted 18.4% for P99 and 16.9% for P95 during 6--- weather type and 15.5% for P99 and 14.4% for P95 during type 8---. Warmer weather occurred less frequently (1--- and 2---; Figure 2), but even in January and February (in 1918 and 1990, respectively) isolated cases of this type of moderately warm weather (2---) occurred. Their frequency rose when days with P95 were also taken into account bringing the total to 5 days in January (4 days in 1918 and 1 in 1983) and 10 days in February (1938, 1943, 1950, 1989, 1990 and 1998). On such days, cloudiness was most frequently low, the relative insolation was greater than 68% and there was no fog; in other words, the 2020 weather type prevailed. When a strong high-pressure area of this type persisted for several consecutive days, as in January 1918 (4 days) and in February 1938 (3 days), then the first or second day with the 2020 or 2021 weather type was followed by the 2210 or 2211 weather. There were slightly more days with cool weather types recorded (1---; 15 days for P99 and 208 for P95). Most frequently, the days were very sunny, with little cloud amount, without fog (1020, P99 − 2.7%, P95 − 4.7%), and in the P95 the days were also often foggy (1021–3.2%; Figure 2). The very frosty days with average daily temperature lesser than − 15 °C (0---) are worth mentioning as well. They constituted as much as 7.3% for P99 and 2.9% for P95 (81 days in total) and most frequently they were very cloudy and without fog—0200 (Figure 2). An exception to the rule was recorded between 3 and 19 February 1959 (17 days), when the air pressure in Krakow exceeded 1030.7 hPa (P95), including more than 1037.5hPa (P99) on 10 of those days. The predominant weather types were 9011 and 8021. On 14 February 1959, the weather started to warm up and its type changed from ground-frost to very cold (7---) and cool (1---).

Table IV. Frequency [%] of weather types on days with high pressure in Krakow in the period 1901–2000 (annual course)
Strong highs and weather typesJanuaryFebruaryMarchAprilMayJuneJulyAugustSeptemberOctoberNovemberDecember
High pressure—percentile 99 (P99)
(33---) hot
(3---) very warm
(2---) moderately warm0.71.210.025.0
(1---) cool4.33.525.011.41.2
(4---) ground-frost, moderately cool
(5---) ground-frost, very cool5.09.320.025.011.44.9
(6---) ground-frost, moderately cold15.620.945.0100.025.025.79.8
(7---) ground-frost, very cold4.7
(8---) moderately frosty14.918.65.020.014.6
(9---) fairly frosty50.332.615.031.563.4
(0---) very frosty9.29.25.06.1
Sum100.0100.0100.0100.0100.0100.0100.0
High pressure—percentile 95 (P95)
(33---) hot
(3---) very warm25.0
(2---) moderately warm1.02.916.932.075.091.451.07.70.8
(1---) cool5.710.210.78.018.720.912.3
(4---) ground-frost, moderately cool0.33.48.02.91.9
(5---) ground-frost, very cool10.016.028.848.05.723.918.712.3
(6---) ground-frost, moderately cold16.421.520.34.02.622.616.9
(7---) ground-frost, very cold0.21.21.1
(8---) moderately frosty15.115.18.51.319.120.6
(9---) fairly frosty45.429.49.60.611.134.0
(0---) very frosty6.13.50.62.9
Sum100.0100.0100.0100.0100.0100.0100.0100.0100.0100.0100.0100.0

It is worth noting in Table IV that the very high pressures (P99) in March or April were mostly accompanied by the ground-frost, moderately cold (6---) weather type. However, it was difficult to determine any patterns in the occurrence of the accompanying weather types because very strong highs in these months were few and far between. The same is true for October when as many as four weather types (1---, 2---, 5---, 6---) accompanied very high pressures in Krakow at the same level of probability (one case each) (Table IV).

No days with P95 were recorded between June and August (Table IV) in any year during the century analysed. Very high pressure in spring (March–May) was mostly accompanied by ground-frost, very cool (5---) and moderately warm (2---) weather types. A similar pattern prevailed in autumn (September–November), but with a greater frequency of moderately warm weather. It seems that where strong high-pressure systems occur during the intermediate seasons, they bring the weather type: ground-frost, very cloudy, without sunshine and fog (5200) or moderately warm, very cloudy, without sunshine and fog (2200); in autumn, there is in addition moderately warm, very sunny weather without fog or with fog (2020 or 2021) and the moderately warm, cloudy, moderately sunny with fog (2111) weather type.

The study went on to check whether there was any variation in the time when the weather types occurred. Only the main weather types were taken into account, as the various subtypes and classes were not numerous enough. The frequency of the most often occurring subtypes did not exceed 8% (weather types: 9200–7.6%; 9110–7.3% and 9010–6.0%).

A long-term analysis of weather types accompanying very high pressure in Krakow did not yield any discernible annual or seasonal trends. This is further confirmed by the two most frequent weather types: fairly frosty (9---) and moderately frosty (8---), which combined account for 60.2% of days (Figures 3 and 4). In both cases, a slight fall in the number of days with a particular weather type can be noticed after 1966; however, it is connected with a simultaneous fall in the number of days with strong highs (Figures 1 and 3). No clear change in the occurrence of the examined weather types during a year can also be observed. They usually occurred from December to February (Figure 4) and were more frequent towards the end of autumn only at the beginning and at the end of the study period. It is also worth noting that these two types also occurred near the end of the 20th century, despite the exceptionally warm winter seasons recorded in that period.

Figure 3.

Number of days in selected weather types on days with strong highs (P99 and P95) in Krakow in the period 1901–2000

Figure 4.

Number of days with fairly frosty (9---) and moderately frosty (8---) weather types on days with strong highs (P99) in Krakow in the period 1901–2000

6. Discussion and conclusions

This study proposes an analysis of the relationship between an important element in the circulation—very strong high-pressure systems and the weather defined using a number of meteorological elements. The results obtained may be helpful in understanding the change in regional circulation and in determining its impact on climate change in the study area (Anagnostopoulou et al., 2009; Cahynová and Huth, 2009).

The study has shown that during the 20th century the number of days with very high air pressure in Krakow was gradually falling. The lowest numbers of such days were recorded from around 1960 to the late 1980s. The drop was statistically insignificant, but it confirmed the change in the number of anticyclonic circulation types observed in winter in the Czech Republic since the 1960s and the frequency of the Central European high (HM according to Hess–Brezowsky) over Europe in the 20th century (Kyselý, 2008; Cahynová and Huth, 2009). On the other hand, however, studies of the circulation change in Europe in the second half of the 20th century, using both objective and subjective (Hess–Brezowsky classification of weather types) methods, suggest that there is an increase in the number of days with anticyclonic types in the winter season (Stefanicki et al., 1998; Kyselý and Huth, 2006). Although this does not normally lead to extremely high air pressure values, it is responsible for the influx of very cold air masses into Central Europe.

The study has found that very high air pressure values occur predominantly in the cool half of the year (October–March). In the warm half of the year (April–September), the pressure values in the anticyclone systems were much lower (differences between pressure values in January and July for particular values of percentile can serve as an example; Table I); and in summer in most cases differences between the pressure in the anticyclone systems and the average pressure values in a given month were lower than 8.6 hPa (from 4.1 hPa in July for P80 to 8.6 hPa in June for P95). That is the reason why in the analysis only the days on which air pressure values were equal or greater than the 99th (P99) and 95th (P95) percentile of all days were used. Over the entire study period, days with air pressure equal to or greater than 1037.5 hPa (P99) did not occur between May and September, and days with air pressures equal to or greater than 1030.7 hPa (P95) were not recorded from June to August. These days were accompanied by 68 and 115 different kinds of weather in eight main types, respectively. The two most frequently occurring weather types were fairly frosty (9---) and moderately frosty (8---), which accounted for 60.2% of all cases. The most frequent subtype was very cloudy, without sunshine or with little sunshine (-20-) at 30.6%, and the most frequent class was without fog (---0) at 7.6%. The annual distribution of the weather types identified reveals that the winter season is dominated by frosty weather, whereas the ground-frost weather type dominates in spring and autumn. In autumn, warm and warm with fog weather types grow in frequency on days, with the air pressure equal to or greater than 1030.7 hPa (P95). A long-term variability analysis produced no annual or seasonal trends in either days with very strong highs or in the accompanying weather types.

It is noteworthy that despite the increase in the number of mild winters recorded in Central Europe since the late 1980s, there has been no noticeable drop in the frequency of incidence of frosty weather. This may be accountable to the increase in the number of strong anticyclones recorded in the 1990s and their greater persistence, which has been reflected in the growing duration of cold waves (Kyselý, 2008).

The proposed classification is a compromise between the desire to minimise the number of weather types, subtypes and classes and the need to provide the maximum possible detail of the weather conditions on a given day (Ustrnul, 2000/2001). The large number of weather types eventually proposed in the classification is a good reflection of the actual weather conditions produced on days with very high pressure in Krakow. It also facilitates an application of the classification to more detailed research whereby one of the meteorological elements is selected as the lead element. On the other hand, it would be possible to merge subtypes and classes, thus boosting the sample size for statistical purposes, if significant relationships were to be found between the weather and an event studied. The study also confirms that, at moderate latitudes, the main differentiating meteorological element is air temperature (Anagnostopoulou et al., 2009).

The proposed classification may be used to determine weather types on any day of the year, and not just on days with very strong highs. For this purpose, however, additional meteorological elements should be considered, e.g. precipitation or wind speed.

The method used to classify very strong highs does not take into consideration seasonal variations in air pressure. As a result, high-pressure systems occurring in Krakow during the warm half of the year did not come out in the classification. These systems may display slightly lower values, but their impact on the weather is equally significant (heat waves, droughts, etc.). For this reason, a future classification exercise should be considered covering very strong pressure systems and related weather in all seasons of the year.

Acknowledgements

This study was supported by a grant from the Ministry of Science and Higher Education (N306 049 32/3237).

Ancillary