Abstract

The first tornado climatology for Romania is presented based on datasets attained from three periods between 1822 and 2013. The historical period (1822–1944) contains 33 tornado reports originating from historical newspaper archives and publications of the Romanian Meteorological Institute. Evidence of tornado observations in Romania before the nineteenth century is found in the representation of tornadoes in the Romania folk mythology. The socialist period (1945–89) contains only seven tornado reports, likely because during this period it was believed that tornadoes did not occur in Romania. The recent period (1990–2013) contains 89 tornado reports that came from mass-media sources and eyewitness reports. Of the 129 tornadoes from the Romanian tornado database, 98 were reported between May and July with a peak in May (36 reports). Most of the tornadoes (28 reports) occurred during the afternoon hours 1500–1659 local time. Tornadoes were more frequently reported over eastern Romania compared with other regions of the country, with a maximum over southeastern Romania [0.37–0.45 (105 km2)−1 yr−1].

1. Introduction

Tornado climatologies are important for understanding the formation and characteristics of severe convective storms, and also for better quantifying the risks that tornadoes pose. The reported frequencies of tornadoes are, in general, lower in Europe compared to the United States. In his study on tornadoes and waterspouts in Europe, Alfred Wegener estimated that at least 100 tornadoes occur each year in Europe (Wegener 1917). More recently, Dotzek (2003) estimated that 329 tornadoes and waterspouts are observed each year in Europe, based on a survey among the participants of the Second European Conference on Severe Storms. Groenemeijer and Kühne (2014) showed, based on the data from the European Severe Weather Database (Dotzek et al. 2009) between 2006 and 2013, that the average annual number of tornadoes and waterspouts in Europe is 483, representing 4.8 (105 km2)−1. By comparison, the average annual number of tornadoes (no waterspouts) in the United States between 2006 and 2013 was 1228, representing 12.5 (105 km2)−1, based on the National Oceanic and Atmospheric Administration publication Storm Data (e.g., Smith et al. 2012; Thompson et al. 2012). Although the threat is apparently smaller in Europe compared with the United States, the true magnitude of the tornado threat in Europe is not known because of the lack of assembled datasets. Despite their importance, it was only recently that some of the European countries started a systematic documentation of tornado events and developed tornado climatologies. Tornado climatologies have been published for 14 (32%) of the 44 countries that have their capital city within Europe, covering approximately 2 958 988 km2 (30%) of the European surface area (9 930 054 km2). The European tornado climatologies mainly focused on northern, southern, and western Europe and, to a lesser extent, on eastern Europe (e.g., Szilárd 2007; Brázdil et al. 2012; Simeonov et al. 2013) (Fig. 1 and Table 1).

Fig. 1.

The spatial distribution of tornado climatologies in Europe. The countries for which tornado climatologies have been published (Table 1) are represented in green, and the climatologies for eastern Europe are labeled. Romania is represented in orange.

Fig. 1.

The spatial distribution of tornado climatologies in Europe. The countries for which tornado climatologies have been published (Table 1) are represented in green, and the climatologies for eastern Europe are labeled. Romania is represented in orange.

Table 1.

Climatologies of tornadoes for European regions. The European regions are based on the definition from the United Nations Statistics Division (available online at http://unstats.un.org/unsd/methods/m49/m49regin.htm). The tornado climatology for Turkey was also included since Turkey is a contiguous transcontinental country, located in western Asia and southeastern Europe.

Climatologies of tornadoes for European regions. The European regions are based on the definition from the United Nations Statistics Division (available online at http://unstats.un.org/unsd/methods/m49/m49regin.htm). The tornado climatology for Turkey was also included since Turkey is a contiguous transcontinental country, located in western Asia and southeastern Europe.
Climatologies of tornadoes for European regions. The European regions are based on the definition from the United Nations Statistics Division (available online at http://unstats.un.org/unsd/methods/m49/m49regin.htm). The tornado climatology for Turkey was also included since Turkey is a contiguous transcontinental country, located in western Asia and southeastern Europe.

For Hungary, Szilárd (2007) developed a synoptic climatology of damaging tornadoes (defined as tornadoes producing any type of damage) reported between 1990 and 2001. Before 1990, damaging tornadoes were estimated to occur less than once in a decade. Between 1990 and 2001, 36 tornadoes were reported, of which 18 were damaging tornadoes. The increase in the number of tornado reports after 1990 was attributed to an increase of vulnerability of society and industry, an increase in the public awareness and also to a “possible intensification of convective activity (among [those] years were record hot summers)” (Szilárd 2007, p. 264).

Brázdil et al. (2012) analyzed the spatial and temporal distribution of tornadoes in Czech land (recent Czech Republic) from 1119 to 2010. During this, period 307 tornadoes occurred in 264 tornado days (defined as the days in which a least one tornado was reported). Before 1500 a.d., a total of four tornadoes were reported, and between 11 and 16 tornadoes were reported for each century up to 1800. A maximum in the number of tornado reports was observed between 1931 and 1940 (44 tornado reports) and another maximum in 2001–10 (56 tornado reports). The recent increase in the number of tornado reports was attributed to the “availability of relevant sources as well as increased social awareness and advances in communication technology” (Brázdil et al. 2012, p. 193).

Simeonov et al. (2013) analyzed the tornado reports for Bulgaria and showed that 57 tornadoes occurred in 51 days between 1956 and 2010. The majority of tornadoes were reported after 1990 (45 reports). For a period of 35 yr between 1956 and 1990, only 12 tornadoes were reported in Bulgaria. During this period “most people in Bulgaria thought that tornadoes were exotic and not typical events for [the] country” (Simeonov et al. 2013, p. 62). The increase in the number of tornadoes after 1990 was attributed to the development of communications and the Internet, which made available the tornado reports collected by amateurs.

The aim of this article is contribute to the climatology of tornadoes in Europe by presenting the first tornado climatology for Romania, a country with a long history of meteorological observations in eastern Europe (the Romanian Meteorological Institute was founded in 1884) (Fig. 1). This article is structured as follows. Section 2 describes the Romanian tornado database. The spatial distribution of tornado reports is described in section 3. The monthly and diurnal distributions of tornado reports are discussed in sections 4 and 5, respectively. Finally, section 6 summarizes the results of this paper.

2. Data

The definition of a tornado used in this article has been adopted from the Glossary of Meteorology (Glickman 2000). Thus, a tornado is defined as “a violently rotating column of air, in contact with the ground, either pendant from a cumuliform cloud or underneath a cumuliform cloud, and often (but not always) visible as a funnel cloud” (Glickman 2000, p. 781). In this article, the definition was extended by considering all the waterspouts that hit the land as tornadoes, consistent with the tornado definitions used in other European countries (e.g., Rauhala et al. 2012). This definition of a tornado was used by the Romanian National Meteorological Administration (RNMA) since 2005. The intensity of all tornadoes in the Romanian tornado database was assessed following the approach of Rauhala et al. (2012), based on (i) the Fujita scale [F scale; Fujita (1981)] and (ii) guidance for assigning tornado damage to buildings [Table 4 in Minor et al. (1977); appendix C in Bunting and Smith (1993)].

The climatology of tornadoes in Romania was divided into three periods. The first period, comprising the historical database, starts in 1822 and ends in 1944 when Romania became a socialist country (section 2a). The second period contains only seven tornado reports for an interval of 55 yr between 1945 and 1989. The third period contains the tornado reports between 1990 and 2013, the period during which the RNMA has been collecting and analyzing tornadoes reports in Romania (section 2c).

a. Historical tornado reports

The first tornado report in Romania is from the beginning of the nineteenth century, but tornadoes have been observed before, as is shown by the Romanian folk mythology related to the figure of the “dragon” (balaur in Romanian) and the “sorcerer” (solomonar in Romanian). For the folk mentality, the dragon is the Principal of Disorder, which disturbs the order of nature and human communities by bringing thunderstorms and hail. The “solomonar,”1 the Principle of Order, is a sorcerer that has the power to control the weather elements and to subdue the dragon (Oişteanu 2013). In the folklore of other countries, high winds and severe storms also had supernatural representations or were interpreted as divine judgment [e.g., Janković (2000), for United Kingdom]. We conjecture that tornadoes have been represented in the Romanian folk mythology as “balauri” [“ale,” or “hale” in southern Romania; Haşdeu (1887)]. The description of the dragons varies from one region to another, but with some common characteristics. Thus, the dragon has a long tail “swinging when it is up into the cloud” (representing the funnel cloud) and “slapping with a loud noise when it is touching the ground” (representing the tornado itself); the dragon’s head is either the head of a crocodile or the head of a horse (representing the anvil of the cumulonimbus cloud); the dragon’s breath “is so cold that [it] is freezing the water in the clouds” thus producing large hail (sometimes associated with tornadic events); the dragon is also able to “lift people up into the clouds” (Marian 1878a,b; Pamfile 1915, 1916; Gherman 1928; Rezuş 1972). Sometimes the dragon takes the form of a winged white horse, which the solomonar rides through the clouds (Ioniţă 1982). These winged horses ridden by sorcerers with meteorological powers are also mentioned in Serbo-Croatian mythology [“graboncijaš dijak”; Jagić (1877)]. Thus, tornadoes were not unknown events in Romania before the nineteenth century, as shown by the geographical distribution of the folklore sources in which the tornadoes are mentioned as balauri (Fig. 2). For southeastern Romania, no folklore sources could be identified in which tornadoes are represented as dragons,2 despite the fact that a large number of the tornadoes in the recent period are reported in this region.

Fig. 2.

The topography of Romania and the spatial distribution of the folklore sources in which the tornadoes are mentioned as balauri (yellow circles). The major cities in Romania (with populations greater than 280 000) are represented by the black circles. Other cities referenced in this article are represented by white circles.

Fig. 2.

The topography of Romania and the spatial distribution of the folklore sources in which the tornadoes are mentioned as balauri (yellow circles). The major cities in Romania (with populations greater than 280 000) are represented by the black circles. Other cities referenced in this article are represented by white circles.

The first historical report of a tornado in Romania is from 4 June 1822 and occurred in Banat (near Timişoara in Fig. 2). The severe storm, described as a whirlwind, hit several villages and destroyed houses and churches and uprooted or snapped large trees, the damages being evaluated at approximately 90,0000 florins [approximately $1 million (U.S. dollars) in 2014] (Dudaş 2006, p. 45). Most of the historical tornado reports came from newspaper archives. To find these reports, newspapers between 18293 and 1944 were studied using archives from the Digital Library of Bucharest (http://www.dacoromanica.ro) and the Digital Library of the Central University Library of Cluj-Napoca (http://dspace.bcucluj.ro/). National (e.g., Adevărul, Albina Românească, Curierul Românesc, Epoca, Scânteia) and local (e.g., Albina Carpaţilor, Clujul, Gazeta de Transilvania) newspapers were studied using keyword searches for “trombă” (from the French trombe or the Italian tromba word for tornado) [e.g., Hepites (1887) used the term “trombă” to describe the 9 June 1886 Bucharest tornado], “uragan/orcan” (from the French ouragan or the German orkan for strong windstorms) [e.g., Hepites (1904) used the term uragan to describe the 29 June 1904 Moscow tornado], “tornadă” (tornado), and “vârtej” (whirlwind). This keyword search resulted in approximately 2690 archive entries, which were then analyzed individually to identify tornado events based on the information from the newspaper article (e.g., description of an event and its damage, eyewitness reports, location, and time of occurrence). Thus, 16 tornado reports (48.5%) of the 33 reports from the historical dataset are from newspaper articles (Fig. 3). The most significant tornadic event retrieved from the newspaper archives is a tornado ranked as a category 3 (F3) event on the Fujita scale that occurred on 13 May 1912. The tornado killed six people and injured more than 50 others, and caused extensive damage to five villages near Dej, central Romania (Fig. 2).

Fig. 3.

The distribution of tornado reports per decade between 1822 and 2013. The first decade includes 1822–29 and the last decade includes 2010–13.

Fig. 3.

The distribution of tornado reports per decade between 1822 and 2013. The first decade includes 1822–29 and the last decade includes 2010–13.

Tornado reports were also collected from Annals of the Romanian Meteorological Institute (14 reports), marginalia [2 reports; Dudaş (2006)], and memoirs [1 report; Michelet (1916)]. The Annals, published between 1886 and 1915, contained monthly summaries of meteorological data and descriptions, as well as case studies of significant weather events. The first case study of a tornado in Romania, describing the 9 June 1886 tornado that hit Bucharest, was published in the second issue of the Annals (Hepites 1887). The tornado occurred close to the Meteorological Institute, thus allowing detailed observations. At least three houses were completely destroyed, and one person was killed by the tornado that occurred between 1700 and 1800 LT (1500–1600 UTC). At the Meteorological Institute, situated approximately 9 km from the area most affected by the tornado, the wind changed from easterly to northeasterly and reached a maximum speed of 15 m s−1 during the passage of the storm, which also produced hail with a diameter of 2 cm. A damage survey was conducted by Ştefan Hepites, the first director of the Meteorological Institute, wherein the damages were estimated at 200,000 francs [approximately $800,000 (U.S. dollars) in 2014].

There are a series of limitations with the historical tornado dataset. Thus, not all the newspapers (in particular, local newspapers) published between 1822 and 1944 were available for analysis. Compared with a tornado damage surveys, a tornado report retrieved from a newspaper contains a limited description of the event. This could result in location and time errors. Also, tornado reports were collected by the Meteorological Institute between 1884 and 1915. In 1916, Romania declared war on Austria–Hungary, thus formally entering World War I, and the Meteorological Institute ceased all activities due to shortage of staff. The activity of the Meteorological Institute was fully restored in 1920, but without any official publications until the 1950s.

b. Tornado reports during socialist period

For the socialist period (1945–89), only seven tornado reports were included in the database (Fig. 3). Two reports came from the monthly meteorological bulletin published by the Meteorological Institute. In the 1960s, two tornadoes were documented in the meteorological observers notebooks, but because they were considered erroneous observations, they were not included in the official reports. The keyword search in the digital newspaper archives between 1945 and 1989, using the same methodology as for the historical period, resulted in only two tornado reports. The lack of tornado reports during the socialist period can be a result of the fact that in the 1970s and 1980s, the word tornado was forbidden in the official meteorological reports and in the mass media reports, despite the previous observations of tornadoes in Romania. During this period, the senior meteorologists considered that tornadoes do not occur in Romania because the country is situated too far north (approximately 45°N), and “thus the Coriolis effect will not allow the formation of tornadoes,” which are “confined to the tropics” (Lemon et al. 2003).4 All the tornadic events during this period were described as high-wind events, thus not recognizing the threat of tornadoes. This situation is what Doswell (2003) described as a self-fulfilling prophecy: denying the existence of tornadoes in Romania resulted in no record keeping for such events, and when tornadoes were observed, they were not reported. A similar situation was described by Setvák et al. (2003) for the Czech Republic. Another source of tornado reports for this period is the eyewitness reports received at the Meteorological Institute after 2002. Thus, one report was included in the database based on the interview with the eyewitness.

c. Recent tornado reports

After the fall of the Iron Curtain and the Romanian Revolution in 1989, there was no formal recognition that tornadoes can occur in Romania. This situation changed in August 2002 when an F3+ long-track tornado crossing through southeastern Romania was responsible for at least three fatalities and the destruction of 33 houses mainly in the village of Făcăeni (Lemon et al. 2003). Although this was not the first F3 tornado documented in Romania, it was the first F3+ tornado whose parent storm circulation was observed with the new Romanian Doppler radar network installed in 2002. The radar network comprised five Weather Surveillance Radar-1998 Doppler (WSR-98D) S-band radars and three existing C-band radars (Ioana et al. 2004). The data provided by the Romanian radar network and the radar-data algorithms for mesocyclones (Mitchell et al. 1998) and tornado detection (Stumpf et al. 1998) were successfully used in 2005 to issue the first tornado warning in Romania (Teittinen and Schultz 2008). After 2005, a systematic documentation and analysis of tornado reports was started at the RNMA with the aim of developing a tornado database for Romania.

Of the 89 tornadoes reported between 1990 and 2013, 85 reports were obtained from mass-media sources and 4 reports were eyewitness reports submitted to the RNMA. A mass-media report was confirmed if (i) a photo or a video of the tornado was available (40 reports, 45.0% of all recent reports), (ii) a photo or a video showing typically tornado damage and interviews with the credible eyewitnesses were available (26 reports, 29.2%), or (iii) a damage survey or a cases study was conducted by the RNMA (19 reports, 21.3% of all recent reports). The eyewitness reports were submitted by professional meteorologists (two reports, 2.2% of all recent reports) and severe weather spotters (two reports, 2.2% of all recent reports). Radar and satellite data were analyzed for all the tornado reports after 2002 to confirm that the radar imagery showed radar echoes or the satellite imagery showed cloudiness during or after the time of the event. Even with a verification system in place, there are limitations to the correctness of the tornado reporting. Thus, not all the recent tornado reports were well correlated with the cell locations from the radar data. These differences were associated with locations errors (e.g., location of the nearest village was provided instead of the actual location of the tornado) and time errors (e.g., only an estimate of the actual time of the tornado was provided).

Altogether, 129 tornadoes that occurred on 112 days have been reported in Romania between 1822 and 2013. The majority of the tornadoes occurred over land (121 reports) and eight tornadoes occurred first over water and then hit the land. Certainly, this dataset is incomplete, as shown for example by the low number of tornado reports during the socialist period. The Romanian tornado dataset is clearly dominated by recent events, with an increase in the number of reports after the year 2000 (Fig. 3). This increase in the number of tornado reports was observed in other European countries, too. For example, for Finland, Rauhala et al. (2012) showed an increase from 50 tornado reports between 1990 and 1999, to 130 reports between 2000 and 2007. For Romania, the recent increase in the number of tornado reports can be attributed to

  1. increased public awareness after the Făcăeni F3+ tornado in 2002;

  2. implementation of the WSR-98D radar network in 2002 (Ioana et al. 2004) helped in defining tornado locations considerably, especially in underpopulated areas, by detecting the larger circulation pattern in which the tornado was embedded (i.e., the mesocyclone);

  3. a rapid increase in cellular telephone subscriptions per 100 inhabitants from 11.2 in 2000 to 105.0 in 2012, along with a rapid increase in the percentage of individuals using the Internet from 3.7 in 2000 to 50.0 in 2012;5

  4. volunteer severe weather spotters, some of them trained by the RNMA (e.g., the Association for Monitoring Severe Weather Phenomena was founded in 2010); and

  5. the “Twister effect” hypothesized by Rauhala et al. (2012) in which the movie Twister (released in 1996) and documentary reality television series about tornadoes (e.g., Storm Chasers that premiered in 2007) resulted in an increased awareness among the public about tornadoes.

3. Spatial distribution

The spatial distribution of tornadoes reports in Romania is shown in Fig. 4. The distribution of tornado reports during the historical period reflects the availability of documentary sources rather than the true distribution. Thus, the majority of tornadoes (29 reports representing 87.9% of all historical reports) were reported over southern and eastern Romania, a region that between 1881 and 1913 was the Romanian Kingdom. The region of Romania bounded to the east and south by the Carpathian Mountain Range was a part of the Austro–Hungarian Empire before 1918, and no official reports for this region were available before 1920 (Fig. 4a).

Fig. 4.

Spatial distribution of (a) historical tornado reports (33 reports between 1822 and 1944), (b) tornado reports during the socialist period (7 reports between 1945 and 1989), and (c) recent tornado reports (89 reports between 1990 and 2013) in Romania. Tornadoes were classified according to their intensity on the F scale for weak tornadoes (F0 or F1) (yellow) and significant tornadoes (F2 and F3) (red). Tornadoes for which an estimation of the intensity was not possible are represented in blue.

Fig. 4.

Spatial distribution of (a) historical tornado reports (33 reports between 1822 and 1944), (b) tornado reports during the socialist period (7 reports between 1945 and 1989), and (c) recent tornado reports (89 reports between 1990 and 2013) in Romania. Tornadoes were classified according to their intensity on the F scale for weak tornadoes (F0 or F1) (yellow) and significant tornadoes (F2 and F3) (red). Tornadoes for which an estimation of the intensity was not possible are represented in blue.

Figure 5 shows the distribution of all tornado reports between 1822 and 2013 by F scale, in which the F-scale estimate is the minimum that can be retrieved from the description of the event. From the 23 historical reports for which an estimation on the F scale was possible, 11 reports were for weak tornadoes [F0 or F1; Hales (1988)] and 12 reports were for significant tornadoes (F2 or F3). The large percentage of significant tornadoes during the historical period compared with the other periods is because strong tornadoes have a large impact on society and are therefore more likely to be reported than weak tornadoes (Brooks and Doswell 2001; Verbout et al. 2006).

Fig. 5.

Number of tornado reports during the historical period (1822–1944) (light gray, bar chart), the socialist period (1945–89) (dark gray, bar chart), and the recent period (1990–2013) (solid line) based on estimated intensity on the F scale.

Fig. 5.

Number of tornado reports during the historical period (1822–1944) (light gray, bar chart), the socialist period (1945–89) (dark gray, bar chart), and the recent period (1990–2013) (solid line) based on estimated intensity on the F scale.

During the socialist period, the annual average number of tornado reports decreased from 0.44 (105 km2)−1 (yr)−1 between 1879 and 1913 to 0.06 (105 km2)−1 (yr)−1 between 1945 and 1989. All the tornadoes reported during the socialist period, with one exception (the 12 June 1961 tornado from Cluj), occurred over eastern and southern Romania (Fig. 4b). The lack of tornado reports is an artifact of the socialist period and not the result of climatological factors. For example, Iliescu (1989) showed, based on cloud-to-ground (CG) lightning data between 1966 and 1980, gathered using CG lightning counters, that the annual average number of thunderstorm days (days in which at least 15 CG lightning flashes were detected) varies from 35 to 40 thunderstorm days over most parts of Romania to 25–30 days over southeastern Romania [Fig. 21 in Iliescu (1989), p. 87]. Geicu and Cândea (2008), using data from the Romanian surface observation network collected between 1961 and 2000, showed that the annual average number of thunderstorm days over Romania varies from 30 to 40 days. Thus, the annual average number of thunderstorm days is not lower during the socialist period compared with the average between 1961 and 2010.

For the recent period, the annual average number of tornado reports increased compared with the previous periods to 1.55 (105 km2)−1 (yr)−1 between 1990 and 2013. The tornado reports are most frequent for the low elevations over southern and eastern Romania with a maximum over southeastern Romania (Fig. 4c). From the 41 recent reports for which an estimate of the F scale was possible, 68.3% (28 reports) were for F0 tornadoes, 24.4% (10 reports) were for F1 tornadoes, and 7.3% (3 reports) for F2 or F3 (Fig. 5). The small percentage of significant tornadoes during the recent period can be attributed to a more efficient collection of tornado reports (particular those for weak tornadoes) and to an increased awareness among the public. Similarly, Rauhala et al. (2012) showed that the 29% of the tornado reports in Finland between 1796 and 1996 were for significant tornadoes, and only 4% of the reports between 1997 and 2007 were for significant tornadoes.

Figure 6a shows the average number of tornadoes per 105 kilometers squared per year based on the tornado reports between 1822 and 2013 using kernel density estimation (KDE). The KDE for a spatial point pattern (i.e., tornado locations) assumes that the spatial pattern has densities at every location, rather than only at the locations where the events were reported (e.g., Dixon et al. 2011; Brooks et al. 2003a). In this article, a Gaussian kernel with a bandwidth of 50 km (comparable with the area for which the RNMA issues severe weather warnings) was used. The 50-km KDE, calculated on a 10-km output grid to produce a smooth map, shows that the average number of tornadoes is between 0.30 and 0.45 (105 km2)−1 (yr)−1 [approximately 1.5–2.25 (105 km2)−1 every 5 yr] over northeastern and southeastern Romania, and is lower than 0.22 (105 km2)−1 (yr)−1 (approximately 1.1 tornadoes every 5 yr) over most of Romania (Fig. 6a).

Fig. 6.

(a) KDE analysis for the tornado reports in the recent dataset (1990–2013) showing the annual average number of tornado reports [(105 km2)−1 yr−1, shaded according to the scale]. (b) Population density of Romanian counties (people per km2, shaded according to the scale) and the urban areas (blue) derived from the 2002–03 MODIS data at 1-km resolution (Schneider et al. 2003).

Fig. 6.

(a) KDE analysis for the tornado reports in the recent dataset (1990–2013) showing the annual average number of tornado reports [(105 km2)−1 yr−1, shaded according to the scale]. (b) Population density of Romanian counties (people per km2, shaded according to the scale) and the urban areas (blue) derived from the 2002–03 MODIS data at 1-km resolution (Schneider et al. 2003).

The high number of tornadoes over northeastern Romania can be attributed to the population density across this region, with high population density resulting in more tornadoes being reported. Figure 6b shows, based on data from the Romania National Institute of Statistics, that the population density in 2003 over northeastern Romania was greater than 100 people per kilometer squared. The high number of tornadoes over southeastern Romania (maximum over Romania) cannot be entirely attributed to the population density. Compared with northeastern Romania, the population density over southeastern Romania is lower (70–80 people per kilometer squared) with fewer urban areas, derived from the 2002–03 Moderate Resolution Imaging Spectroradiometer (MODIS) data at 1-km grid spacing (Schneider et al. 2003) (Fig. 6b).

The high number of tornadoes over southeastern Romania can be attributed to the mesoscale environments over this region that are more favorable for tornadoes compared with other regions. Brooks et al. (2003b) used the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) global reanalysis dataset (Kalnay et al. 1996) for the years between 1997 and 1999 to obtain vertical profiles that resemble radiosonde profiles (Lee 2002). The profiles were used as proximity soundings to develop relationships between the environmental variables (e.g., CAPE, 0–6-km shear, 0–1-km shear, and 2–4-km lapse rate) and severe convective storms in the United States. These relationships were then applied to make estimates of the distribution of favorable severe convective storms and tornado environments over Europe. For Romania, the average number of days with favorable tornado environments is between 4.5 and 6.0 over southeastern Romania and is lower than 3.0 days over most of Romania (Fig. 16 in Brooks et al. 2003b, p. 89). Thus, the results from the present study are consistent with the results obtained by Brooks et al. (2003b) for the distribution of favorable tornado environments over Romania.

4. Monthly distribution

From the 126 tornadoes reported in Romania between 1822 and 2013 containing information on the occurrence month, 125 tornadoes were reported between March and September and one tornado was reported during wintertime on 24 January 2006. The tornadoes occurred on 110 tornado days, defined as the days in which at least one tornado was reported. The earliest start of the tornado season in Romania occurred on 29 March 2006, when a tornado was reported over Târgu-Jiu (southern Romania; Fig. 2), and the latest end of the tornado season occurred on 23 September 1913, when a tornado was reported over Tuzla (eastern Romania; Fig. 2). The majority of tornadoes have been observed during May–July, when 98 tornadoes (77.8% of the reports containing information on the occurrence month) were reported in 83 days (75.5% of the tornado days) (Fig. 7). The peak month for tornado reports is May, with 36 reports representing 28.6% of the reports containing information on the occurrence month. The large number of tornado reports includes a tornado outbreak from 7 May 2005, when eight tornadoes were reported over southeastern Romania. The peak month for tornado days is June with 31 days representing 28.2% of all tornado days (Fig. 7).

Fig. 7.

The percentage of the annual total of tornado reports (bar chart) and tornado days (solid line) occurring in each month between 1822 and 2013.

Fig. 7.

The percentage of the annual total of tornado reports (bar chart) and tornado days (solid line) occurring in each month between 1822 and 2013.

As in Romania, tornadoes in the neighboring countries occur most frequently between May and August, with a peak in late spring or early summer (May–June) [Szilárd (2007) for Hungary; Simeonov et al. (2013) for Bulgaria]. The peak occurs later in the summer (July–August) over western Europe [Holzer (2001) for Austria; Dotzek (2001) for Germany; Paul (2001) for France] and northern Europe [Tooming (2002) for Estonia; Marcinoniene (2003) for Lithuania; Rauhala et al. (2012) for Finland; Peterson (2000) for Sweden; Holden and Wright (2004) for the United Kingdom; Tyrrell (2003) for Ireland]. For southern Europe, the peak occurs in late summer and autumn (August–October) [Gayà (2011) for Spain; Leitão (2003) for Portugal; Giaiotti et al. (2007) for Italy; Matsangouras et al. (2014) for Greece].

5. Diurnal distribution

The diurnal distribution of tornadoes in Romania into 2-h bins in local time (LT = UTC + 2 h), is shown in Fig. 8. The diurnal distribution is based on 95 tornado reports (75.4% of all reports) between 1822 and 2013 that contained information on the occurrence time. The majority of tornadoes (88 reports, 92.6% of all cases) were reported between 0900 and 2059 LT, with a peak in the afternoon between 1500 and 1659 LT (28 reports, 29.4% of all cases). The reporting of only seven tornadoes between 2100 and 0859 LT may be because of the difficulties associated with spotting tornadoes at night (sunset is approximately at 2100 LT during June–August) or because they occur when the public tends to be asleep (Ashley et al. 2008). The diurnal distribution of tornado reports in Romania is similar to those observed in the neighboring countries. For Bulgaria, Simeonov et al. (2013) showed that tornadoes tend to occur between 1400 and 1800 LT (80% of all reports) with a peak around 1600 LT. Tornadoes in Hungary occur most frequently between 1500 and 1900 LT (72% of all reports) (Szilárd 2007).

Fig. 8.

The percentage of the daily total of tornado reports between 1822 and 2013 occurring in 2-h bins starting at the indicated local time (e.g., 1500 LT indicates the period between 1500 and 1659 LT).

Fig. 8.

The percentage of the daily total of tornado reports between 1822 and 2013 occurring in 2-h bins starting at the indicated local time (e.g., 1500 LT indicates the period between 1500 and 1659 LT).

6. Conclusions

This study summarizes the tornado climatology of Romania between 1822 and 2013 based on a dataset comprising 129 tornadoes reported on 112 days. The tornado climatology was divided into three periods: (i) the historical period (1822–1944) during which 33 tornadoes were reported, (ii) the socialist period (1945–89) containing only 7 reports, and (iii) the recent period (1990–2013) with 89 tornado reports. The increase in the number of reports during the recent period can be attributed to increased public awareness (e.g., after the Făcăeni F3+ tornado from 2002, or due to movies and documentary reality television series on tornadoes), increased access to information and communication technology (e.g., mobile phones, Internet), and volunteer severe weather spotters. The main conclusions of the study are as follow.

  1. The spatial distribution of tornado reports shows that tornadoes are more frequently reported over eastern Romania, with a maximum over southeastern Romania [approximately 1.5–2.25 (105 km2)−1 every 5 yr]. We speculate that the large number of tornadoes over southeastern Romania can be attributed to the mesoscale environments over this region that are more favorable for tornadoes compared with other regions of the country, but further studies are necessary to confirm this hypothesis.

  2. The distribution of tornado reports on the F scale shows that the historical dataset is dominated by significant tornadoes (F2 or F3) (12 of 23 reports for which an estimation of the intensity was possible). This bias toward significant tornadoes can be explained by the fact that strong tornadoes have a large impact on society and thus are reported more frequently compared with weak tornadoes (F0 and F1). With a more efficient collection of tornado reports and an increased level of awareness among the public, the recent dataset is dominated by weak tornadoes (38 of 41 reports for which an estimation of the intensity was possible).

  3. The monthly distribution of the 126 tornado reports containing information about the occurrence month shows that tornadoes are reported more frequently in May–July (98 reports), with a peak in May (36 reports).

  4. The majority of tornadoes were reported between 0900 and 2059 LT (88 of the 95 tornado reports containing information about the occurrence time) with their peak between 1500 and 1659 LT (28 reports).

The Romanian tornado database developed in this study represents a contribution toward a pan-European tornado database that will provide the basis for better understanding the tornado threat in Europe.

Acknowledgments

We thank David M. Schultz from the University of Manchester for his helpful suggestions and advice on many aspects of this research and on improving the manuscript. This work was initiated when both authors were at the Romanian National Meteorological Administration, and we thank all our colleagues in the Laboratory of Nowcasting Techniques and Severe Weather Forecasting for their contributions to the development of the Romanian tornado database. We also thank Editor Pam Heinselman and anonymous reviewers for their comments that improved our article. Partial funding for BA comes from Grant NE/H008225/1 from the U.K. Natural Environment Research Council (NERC) to the Tropopause Folding, Stratospheric Intrusions and Deep Convection (TROSIAD) project at the University of Manchester and from an AXA Research Fund postdoctoral grant.

REFERENCES

REFERENCES
Ashley
,
W. S.
,
A. J.
Krmenec
, and
R.
Schwantes
,
2008
:
Vulnerability due to nocturnal tornadoes
.
Wea. Forecasting
,
23
,
795
807
, doi:.
Brázdil
,
R.
,
K.
Chromá
,
P.
Dobrovolný
, and
Z.
Černoch
,
2012
:
The tornado history of the Czech Lands, AD 1119–2010
.
Atmos. Res.
,
118
,
193
204
, doi:.
Brooks
,
H. E.
, and
C. A.
Doswell
III
,
2001
:
Some aspects of the international climatology of tornadoes by damage classification
.
Atmos. Res.
,
56
,
191
201
, doi:.
Brooks
,
H. E.
,
C. A.
Doswell
III
, and
M. P.
Kay
,
2003a
:
Climatological estimates of local daily tornado probability
.
Wea. Forecasting
,
18
,
626
640
, doi:.
Brooks
,
H. E.
,
J. W.
Lee
, and
J. P.
Craven
,
2003b
:
The spatial distribution of severe thunderstorm and tornado environments from global reanalysis data
.
Atmos. Res.
,
67–68
,
73
94
, doi:.
Bunting
,
W. F.
, and
B. E.
Smith
,
1993
: A guide for conducting convective windstorm surveys. NOAA Tech. Memo. NWS SR146, Scientific Services Division, Southern Region, Fort Worth, TX, 44 pp.
Dixon
,
P.
,
A. E.
Mercer
,
J.
Choi
, and
J.
Allen
,
2011
:
Tornado risk analysis: Is Dixie Alley an extension of Tornado Alley?
Bull. Amer. Meteor. Soc.
,
92
,
433
441
, doi:.
Doswell
,
C. A.
, III
,
2003
:
Societal impacts of severe thunderstorms and tornadoes: Lessons learned and implications for Europe
.
Atmos. Res.
,
67–68
,
135
152
, doi:.
Dotzek
,
N.
,
2001
:
Tornadoes in Germany
.
Atmos. Res.
,
56
,
233
251
, doi:.
Dotzek
,
N.
,
2003
:
An updated estimate of tornado occurrence in Europe
.
Atmos. Res.
,
67–68
,
153
161
, doi:.
Dotzek
,
N.
,
P.
Groenemeijer
,
B.
Feuerstein
, and
A.
Holzer
,
2009
:
Overview of ESSL’s severe convective storms research using the European Severe Weather Database (ESWD)
.
Atmos. Res.
,
93
,
575
586
, doi:.
Dudaş
,
F.
,
2006
: Catastrofe naturale în Transilvania (Natural Disasters in Transilvania). Editura de Vest, 176 pp.
Fujita
,
T. T.
,
1981
:
Tornadoes and downbursts in the context of generalized planetary scales
.
J. Atmos. Sci.
,
38
,
1511
1534
, doi:.
Gayà
,
M.
,
2011
:
Tornadoes and severe storms in Spain
.
Atmos. Res.
,
100
,
334
343
, doi:.
Geicu
,
A.
, and
I.
Cândea
, Eds.,
2008
: Clima României (The Climate of Romania). Administraţia Naţională de Meteorologie, Editura Academiei Române, 366 pp.
Gherman
,
T.
,
1928
: Meteorologie Populară—Observări, Credinţe şi Obiceiuri (Popular Meteorology—Observations, Beliefs and Traditions). Tipografia Seminarului Teologic Greco-Catolic, 174 pp.
Giaiotti
,
D. B.
,
M.
Giovannoni
,
A.
Pucillo
, and
F.
Stel
,
2007
:
The climatology of tornadoes and waterspouts in Italy
.
Atmos. Res.
,
83
,
534
541
, doi:.
Glickman
,
T. S.
, Ed.,
2000
:
Glossary of Meteorology
. 2nd ed. Amer. Meteor. Soc., 855 pp.
Groenemeijer
,
P.
, and
T.
Kühne
,
2014
:
A climatology of tornadoes in Europe: Results from the European Severe Weather Database
.
Mon. Wea. Rev.
,
142
,
4775
4790
, doi:.
Hales
,
J. E.
, Jr.
,
1988
: Improving the watch/warning program through use of significant event data. Preprints, 15th Conf. on Severe Local Storms, Baltimore, MD, Amer. Meteor. Soc., 165–168.
Haşdeu
,
B. P.
,
1887
: Etymologicum Magnum Romaniae—The Dictionary of the Historical and Folk Language of Romanians. Socec and Teclu, 2224 pp.
Hepites
,
S. C.
,
1887
: Trombe à Bucharest. Ciel et Terre, 1 March 1886–15 February 1887, 234–237.
Hepites
,
S. C.
,
1904
:
Uraganul de la Moscova (The Moscow hurricane)
.
Analele Institului Meteorologic al României
(Ann. Rom. Meteor. Inst.),
22
,
179
.
Holden
,
J.
, and
A.
Wright
,
2004
:
UK tornado climatology and the development of simple prediction tools
.
Quart. J. Roy. Meteor. Soc.
,
130
,
1009
1021
, doi:.
Holzer
,
A. M.
,
2001
:
Tornado climatology of Austria
.
Atmos. Res.
,
56
,
203
211
, doi:.
Iliescu
,
M.
,
1989
: Manifestări Electrice Atmosferice pe Teritoriul României (Atmospheric Electrical Activity in Romanian). Editura Academiei Republicii Socialiste România, 144 pp.
Ioana
,
M.
,
V.
Ivanovici
,
E.
Cordoneanu
,
D.
Banciu
,
A.
Apostu
, and
B.
Ford
,
2004
: SIMIN–The integrated system for meteorological surveillance, forecast and alert in Romania. 20th Int. Conf. on Interactive Information and Processing Systems (IIPS) for Meteorology, Oceanography, and Hydrology, Seattle, WA, Amer. Meteor. Soc, 11.2. [Available online at https://ams.confex.com/ams/pdfpapers/70954.pdf.]
Ioniţă
,
M.
,
1982
: Cartea vâlvelor. Legende din Apuseni (The Book of Spirits. Legends from Apuseni). Dacia, 239 pp.
Jagić
,
V.
,
1877
:
Die südslavischen volkssagen von dem grabancijas dijak und ihre erklärung (The southern Slavic grabancijas dijak folk tales and their explanations)
.
Arch. Slavische Philologie
,
2
,
437
481
.
Janković
,
V.
,
2000
: Reading the Skies: A Cultural History of the English Weather, 1650–1820. University of Chicago Press, 272 pp.
Kahraman
,
A.
, and
P.
Markowski
,
2014
:
Tornado climatology of Turkey
.
Mon. Wea. Rev.
,
142
,
2345
2352
, doi:.
Kalnay
,
E.
, and Coauthors
,
1996
:
The NCEP/NCAR 40-Year Reanalysis Project
.
Bull. Amer. Meteor. Soc.
,
77
,
437
471
, doi:.
Lee
,
J. W.
,
2002
: Tornado proximity soundings from the NCEP/NCAR reanalysis data. M.S. thesis, School of Meteorology, University of Oklahoma, 61 pp.
Leitão
,
P.
,
2003
:
Tornadoes in Portugal
.
Atmos. Res.
,
67–68
,
381
390
, doi:.
Lemon
,
L. R.
,
A.
Stan-Sion
,
C.
Soci
, and
E.
Cordoneanu
,
2003
:
A strong, long-track, Romanian tornado
.
Atmos. Res.
,
67–68
,
391
416
, doi:.
Marcinoniene
,
I.
,
2003
:
Tornadoes in Lithuania in the period of 1950–2002 including analysis of the strongest tornado of 29 May 1981
.
Atmos. Res.
,
67–68
,
475
484
, doi:.
Marian
,
S. M.
,
1878a
:
Balaurii (The dragons)
.
Albina Carpaţilor
,
2
,
22
23
.
Marian
,
S. M.
,
1878b
:
Solomonarii (The sorcerers)
.
Albina Carpaţilor
,
4
,
54
56
.
Matsangouras
,
I. T.
,
P. T.
Nastos
,
H. B.
Bluestein
, and
M. V.
Sioutas
,
2014
:
A climatology of tornadic activity over Greece based on historical records
.
Int. J. Climatol.
,
34
,
2538
2555
, doi:.
Michelet
,
J.
,
1916
: Principatele Dunărene: Maria C. A. Rosetti în Revoluţia Română (Danubian Principalities: Maria C. A. Rosetti and Wallachian Revolution of 1848). Lui C. A. Rosetti: La o sut de ani de la naterea sa (1816–1916), E. Costinescu, Ed., Democraţia, 99–112.
Minor
,
J. E.
,
J. R.
McDonald
, and
K. C.
Mehta
,
1977
: The tornado: An engineering-oriented perspective. NOAA Tech. Memo. ERL NSSL-82, National Weather Service, Norman, OK, 196 pp.
Mitchell
,
E. D.
,
S. V.
Vasiloff
,
G. J.
Stumpf
,
A.
Witt
,
M. D.
Eilts
,
J. T.
Johnson
, and
K. W.
Thomas
,
1998
:
National Severe Storms Laboratory tornado detection algorithm
.
Wea. Forecasting
,
13
,
352
366
, doi:.
Oişteanu
,
A.
,
2013
: Ordine şi Haos: Mit şi Magie în Cultura Tradiţională Românească (Order and Chaos: Myth and Magic in Romanian Traditional Culture). 2nd ed. Polirom, 624 pp.
Pamfile
,
T.
,
1915
: Mitologie Românească (Romanian Mythology). Socec, 401 pp.
Pamfile
,
T.
,
1916
: Văzduhul (The Sky). Socec, 216 pp.
Paul
,
F.
,
2001
:
A developing inventory of tornadoes in France
.
Atmos. Res.
,
56
,
269
280
, doi:.
Peterson
,
R. E.
,
1998
:
A historical review of tornadoes in Italy
.
J. Wind Eng. Ind. Aerodyn.
,
74–76
,
123
130
, doi:.
Peterson
,
R. E.
,
2000
: Tornadoes in Sweden. Preprints, 19th Conf. on Severe Local Storms, Minneapolis, MN, Amer. Meteor. Soc., 89–92.
Rauhala
,
J.
,
H. E.
Brooks
, and
D. M.
Schultz
,
2012
:
Tornado climatology of Finland
.
Mon. Wea. Rev.
,
140
,
1446
1456
, doi:.
Rezuş
,
P.
,
1972
: Dochiţa împărătiţa. Basme şi Poezii Populare din Ţara de Sus (The Empress Dochiţa. Folk-Tales and Poems from Ţara de Sus). Minerva, 435 pp.
Schneider
,
A.
,
M.
Friedl
,
D.
McIver
, and
C.
Woodcock
,
2003
:
Mapping urban areas by fusing multiple sources of coarse resolution remotely sensed data
.
Photogramm. Eng. Remote Sens.
,
69
,
1377
1386
, doi:.
Setvák
,
M.
,
M.
Salek
, and
J.
Munzar
,
2003
:
Tornadoes within the Czech Republic: From early medieval chronicles to the “internet society.”
Atmos. Res.
,
67–68
,
589
605
, doi:.
Simeonov
,
P.
,
L.
Bocheva
, and
I.
Gospodinov
,
2013
:
On space-time distribution of tornado events in Bulgaria (1956–2010) with brief analyses of two cases
.
Atmos. Res.
,
123
,
61
70
, doi:.
Smith
,
B.
,
R.
Thompson
,
J.
Grams
,
C.
Broyles
, and
H.
Brooks
,
2012
:
Convective modes for significant severe thunderstorms in the contiguous United States. Part I: Storm classification and climatology
.
Wea. Forecasting
,
27
,
1114
1135
, doi:.
Stumpf
,
G. J.
,
A.
Witt
,
E. D.
Mitchell
,
P. L.
Spencer
,
J. T.
Johnson
,
M. D.
Eilts
,
K. W.
Thomas
, and
D. W.
Burgess
,
1998
:
The National Severe Storms Laboratory mesocyclone detection algorithm for the WSR-88D
.
Wea. Forecasting
,
13
,
304
326
, doi:.
Szilárd
,
S.
,
2007
:
A systematic approach to synoptic tornado climatology of Hungary for the recent years (1996–2001) based on official damage reports
.
Atmos. Res.
,
83
,
263
271
, doi:.
Teittinen
,
J.
, and
D. M.
Schultz
,
2008
:
Severe thunderstorm and tornado warnings in Europe
.
Atmos. Res.
,
93
,
369
380
, doi:.
Thompson
,
R.
,
B.
Smith
,
J.
Grams
,
A.
Dean
, and
C.
Broyles
,
2012
:
Convective modes for significant severe thunderstorms in the contiguous United States. Part II: Supercell and QLCS tornado environments
.
Wea. Forecasting
,
27
,
1136
1154
, doi:.
Tooming
,
H.
,
2002
: Strong tornadoes in Estonia. Preprints, Second European Conf. on Severe Storms, Prague, Czech Republic,
Czech Hydrometeorological Institute
, 33.
Tyrrell
,
J.
,
2003
:
A tornado climatology for Ireland
.
Atmos. Res.
,
67–68
,
671
684
, doi:.
Verbout
,
S.
,
H.
Brooks
,
L.
Leslie
, and
D.
Schultz
,
2006
:
Evolution of the U.S. tornado database: 1954–2004
.
Wea. Forecasting
,
21
,
86
93
, doi:.
Wegener
,
A.
,
1917
: Wind- und Wasserhosen in Europa (Tornadoes and Waterspouts in Europe). Vieweg, 301 pp.

Footnotes

Publisher’s Note: This article was revised on 25 August 2015 to include the CCBY license that was missing when originally published.

This article is licensed under a Creative Commons Attribution 4.0 license.
1

The term is quite recent and is documented approximately between 1650 and 1750 in Transylvania, central Romania (Oişteanu 2013).

2

“Volbură” is another term used to described whirlwind events over southeastern Romania (Gherman 1928), but this term is used in the folklore sources to describe dust whirls (i.e., dust devils) and not tornadoes.

3

The first Romanian newspapers, Albina Românească and Curierul Românesc, were first published in 1829.

4

This explanation can also result from the confusion between the word “uragan” used to describe tornadoes in the historical database and the use of the word in modern Romanian to describe tropical cyclones.

5

Source is the World Telecommunication/Information and Communication Technologies Indicators Database (2013) (available online at http://www.itu.int/en/ITU-D/Statistics/Pages/stat/default.aspx).