What we know so far about tornadoes in Europe?

Today I have submitted to Monthly Weather Review a manuscript written together with David M. Schultz and Fiona Lomas (University of Manchester) on European tornadoes. The manuscript, a synthesis of the European tornado datasets, is also a short history of tornado observations in Europe and a literature review of the major contributions on tornado research in Europe.  

There are three reasons for which we believe that this research was necessary. First, there is not a widespread recognition or the threat of tornadoes in Europe, and this lead to an underestimate of the importance of tornadoes in the current climate in Europe. Second, to better understand the climatology of tornadoes in Europe we need to understand the evolution of tornado databases for different European countries and to identifying the major influences on the development of these databases. Third,  when considering the possible influence of climate change on severe convective storms, the first step is to consider the
observational data. Without knowing what is occurring now, how many tornadoes occur each
year in Europe, and what is the current spatial distribution of tornadoes, all the analyses
of the influence of climate change on tornadoes are premature.

We hoped that this manuscript of the current knowledge on tornadoes in Europe [in a way, this is a companion paper to Groenemeijer & Kühne (2014)] will encourage further discussions and stimulate the interest of the scientific community, national meteorological services and the public on European tornadoes. We hope that this would result in increased awareness, in the identification of new data sources, or in the initiation of new databases that would allow a extension of the current European tornado database both in time and in space. 

You can read the draft of the paper here.

The first tornado climatology for Romania

The latest issue of the Monthly Weather Review (March 2015) contains the first tornado climatology for Romania. Our aim in developing this climatology was not only to show that tornadoes do occur in Romania, but also to contribute to the climatology of tornadoes in Europe by providing the spatial and temporal distribution of tornadoes over a region in Eastern Europe. In my previous post - Tornadoes in Romania: From dragons to radars, in which I have tried to tell the story behind this climatology - I haven't explained why I have included the dragon in the title. I am going to provide an explanation in this post. 

The first tornado report in Romania is from the beginning of the nineteenth century, from a period that coincides with the development of national and regional newspaper-type publications (e.g., the first Romanian newspapers, Albina Românească and Curierul Românesc, were first published in 1829) and also with the emergence of organized meteorological observations (e.g., the Prince Nicolae Şuţu included in his "Notiţii Statistice asupra Moldovei" (Statistical notes about Moldova) observations made between 1839–1840 at Iaşi). Obviously tornadoes were observed in Romania before the  nineteenth century, but without any written reports these observations were lost. One way to recover these observations is based on folklore sources.  Since tornadoes have a high impact on human communities, then they must have been represented in the Romanian folk mythology. In the paper we conjecture that in the Romanian folk mythology tornadoes are related to the figure of the dragon (balaur in Romanian) and the sorcerer (solomonar in Romanian). Andrei Oişteanu in his book "Ordine si Haos" (Order and Chaos, 2013) showed that 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, the Principle of Order, is a sorcerer that has the power to control the weather elements and to subdue the dragon. 

View fullsize

The Dragon and the sorcerer as seen by Marcel Olinescu in "Mitologie Românească" (Romanian Mythology) (source).

The description of the dragons in the folklore sources 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”.

Distribution of the folklore sources in which the tornadoes are mentioned as dragons (source).

Thus, we argue that 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 dragons. For southeastern Romania, a region were a large number of tornadoes are reported in the recent period, no folklore sources could be identified in which tornadoes are represented as dragons. 

I would end this post with a quote from "The Dragon" (1928) by Mihail Sadoveanu (1880–1961) which I think the illustrates perfectly our conjecture about tornadoes in dragons: 

“Over Moldova [i.e, northeast Romania], beyond the Bolandars hills, the sky was moving and bending as it was turning against the earth ; and a roar beyond measure, one never heard before, filled the valleys as it was getting near; and all those present, turning their faces and staring around saw the dragon coming in a spiralling whirlwind at a great speed. I saw it with my own eyes and I trembled. It was coming straight towards us. With a thin tail like a black scroll, it was touching the ground and its body was high up in the air and its mouth was gaping like that of a lioness in the clouds.
It was coming roaring and swinging its tail; its very breath was sucking throwing into the sky haystacks, house roofs and uprooted trees. Beneath the roar, a downpour of hail and water was unleashed as if the whole Moldova riverbed was taken to the sky and then knocked down on us.”

— Mihail Sadoveanu, Excerpt from “The Dragon” (from Ancuta’s Inn, collection of short stories)

(It is interesting to note that this collection of short stories was listed as a compulsory reading for high school students during a period in which was considered that tornadoes do not occur in Romania.) 

Tornadoes in Romania: From dragons to radars

Our paper, Tornadoes in Romania - a contribution to the study of tornadoes in Eastern Europe, a region were there is a lack of tornado reports (as I showed in my previous post) - was accepted for publication in Monthly Weather Review (Mon. Wea. Rev.) [update 14 Oct. 2014: the paper is now available as an early online release]. This blog post is the story of Tornadoes in Romania. 

I begin collecting tornado reports, with the aim of developing a tornado database for Romania, in 2004 few months after I started to work as a short range forecaster for Romanian National Meteorological Administration (RNMA). At that time the general opinion was that tornadoes cannot occur in Romania, despite the previous observations of tornadoes in this country. For example, on August 2002 a F3+ long-track tornado that occurred over southeastern Romania was responsible for at least three fatalities (Lemon et al. 2003). So, why tornadoes cannot form in Romania? Lemon et al. (2003) provide this explanations:

“It has been stated publicly by senior meteorologists that the latitude of Romania (∼45° north) was too far north to permit tornadoes. Beyond that, little explanation has been given except in newspapers where it was stated that tornadoes are ‘confined to the tropics’.”

— Lemon et al. (2003, p. 392)

This situation resulted in what Doswell III (2003) described as a self-fulfilling prophecy: since the existence of tornadoes in Romania was denied, then no records were keep for such events, and when tornadoes do occurred they were not reported.

This situation began to change after 2004 due: a) the efforts of a newly form severe weather forecasting group (lead by Aurora Bell), b) the implementation of the WSR-98D radar network in 2002 (which allowed the detection of the larger circulation in which the tornado is embedded, i.e., the mesocyclone) and c) increased public awareness (the Facaeni tornado and other tornadoes featured extensively in mass media). This change culiminated with the first verified tornado warning for Romania (and one of the few tornado warnings in Europe). In their paper on severe thunderstorms and tornado warnings in Europe, Rauhala and Schultz (2009) provide a description of this forecast:

“Romania issued their first tornado warning on 28 May 2005. Forecasters were alerted to the developing scenario by a convergence line in satellite imagery. Later, radar-data algorithms for mesocyclones and tornado detection were used as guidance. Because this warning was the first tornado warning issued in Romania, the mesoscale forecaster [ link], who is responsible for warnings in the Center of Operational Forecasts at the National Meteorological Administration, and the synoptic forecaster, discussed with the deputy director and with a severe-weather expert before issuing the warning. The warning was issued for an area approximately 170 km2 . Ten minutes after the tornado warning was issued, a local TV station reported a funnel cloud ”

— Rauhala and Schultz (2009, p. 373)

Despite the accumulated evidence from mass-media, from the general public, and from the damage surveys conducted by a small team from RNMA (that included Aurora Bell, Sorin Burcea, Daniel Carbunaru, and Carolina Oprea among others) the general view that tornadoes cannot occur in Romania persisted. To provide more convincing arguments (in 2005 the Romanian tornado database comprised only few cases) we needed to show that tornadoes were observed in the past in Romania.  

In 2006, the only historical report that we had in the database was of a tornado that occurred in Bucharest on 9 June 1886. This tornado was discovered in  Alfred Wegener's  book on tornadoes and waterspouts in Europe. The reference provided by Wegener for the Bucharest tornado was a paper by Stefan Hepites (the founder of the Romanian Meteorological Institute) published in Ciel et Terre (a Belgium journal devoted to astronomy and meteorology). Hoping that we will be able to find references to other tornadoes that occurred before 1886, we ask for a copy of this paper from the Société Royale Belge d'Astronomie, de Météorologie et de Physique du Globe. Unfortunately, the paper, an excellent case study of the event, contained a single reference to a paper describing the same event published in the Annals of the Romanian Meteorological Institute. Fortunately, the library of RNMA had copies of the Annals published between 1885 and 1915. Thus, I have spend almost one year going through the entire collection of the Annals, learning about the history of meteorology in Romania, reading the studies of the Romanian meteorologists and discovering new tornado cases. At the end of 2007 we had the first version of the Romanian tornado database and for the next 6 year, together with my colleagues from RNMA we developed the database adding new tornado reports and trying to find historical reports.

In 2013, with more than 100 tornado reports in the database (from 1886–2013, with only six reports between 1945–1989 corresponding to the period during which Romania was socialist country) we decided to write the first climatology of tornadoes in Romania and thus to provide the definite proof that tornadoes do occur in Romania (Fig. 1).

Figure 1. Tornadoes reported in Romania between 1822 and 2013. (click on the image to explore the details) 

The paper was submitted to Mon. Wea. Rev. and during the review process I have discovered through an excellent on line resource more historical tornado reports. Thus, the version of the study that will be published in Mon. Wea. Rev. is based on a database extending back to 1822. Trough this article we hope that we have provided the evidence that tornadoes do occur in Romania, and we hope that we have made a small contribution toward a pan-European tornado database that will provide the basis for understanding the tornado threat in Europe.

Reaching the end of this post you may ask were are the dragons from the title ? I can tell you, but I think is better to let you discover by reading the paper. 

---

On the history of tornadoes observations in Europe

In my previous post I have shown the spatial distribution of tornadoes in Europe based on tornado reports between 2004–2013. I have chosen this interval because this is the interval during which the European Severe Weather Database has been actively collecting tornadoes and waterspouts reports in Europe (the operational phase of  ESWD started in 2006). The efforts of the ESWD were not concentrated only on the recent events, but also on collecting historical tornado and waterspout reports. To show the history of tornado observations in Europe, I have animated the spatial distribution of tornadoes and waterspouts in five year intervals between 1800–2013 using data from ESWD.

View fullsize

The spatial distribution of tornadoes and waterspouts in Europe in five years intervals between 1800–2013 based on data from the European Severe Weather Database. The new tornado and waterspout reports are represented in orange and the previous reports in blue.

The animation starts from zero in 1800. However, this does not signify that tornadoes were not reported before the XIX century, but only that between 1800–1804 we do not have any records of tornadoes or waterspouts in Europe. In fact,  tornadoes have been reported in Europe since the beginning of the XI century. The earliest tornado reported in Europe occurred at Rosdalla (near Kolbeggan) in Ireland on 30 April 1054.  A systematic documentation of tornadoes and waterspouts began in the XIX century. Before the XIX century there are only three notable contributions to the study of European tornadoes. The first is a study by the French theologian François Lamy in which "physical conjectures" on the formation of tornadoes (in particular on the formation of a tornado that occurred near Reims, France on 10 August 1680) are presented. The second is a "physical-mathematical dialogue" by the Italian astronomer Geminiano Montanari describing a tornado that occurred on 29 July 1686 over Mantova, Padova and Verona (northern Italy). The third is a study by Roger Joseph Boscovich of a tornado that occurred in Rome (Italy) on the night between 11 and 12 June 1749.  

If we return now to the animation and we focus only on the reports before the 1930, we observe that the majority of these reports came from France and Germany. This is not surprising if we take into account the research on tornadoes in this two countries. 

In France, Jean Charles Athanase Peltier published in 1840 a study entitle Météorologie: Observations et recherches expérimenta les sur les causes qui concourent à la formation des trombes" (Meteorology: Observations and experimental research on the causes that contribute to the formation of tornadoes ).  Peltier, who began his career as watchmaker and watch dealer, later become interested in experimental physics (electrodynamics in particular, see Peltier effect), atmospheric electricity and meteorology. In his book, Peltier does not only critically discuss the major theories from the XIX century on the formation of tornadoes, but also does an excellent job of collecting and analysing tornado reports from Europe. Thus, his study contains what is probably the first tornado climatology for Europe.

Alfred Lothar Wegener, a German meteorologist and polar researcher, mainly remembered today for advancing the theory of continental drift, continued Peltier's work of collecting tornado and waterspout reports in Europe.  While recuperating in a military hospital in Berlin from an injury he suffered as a German soldier during World War I, Wegener developed his comprehensive study of tornadoes and waterspouts in Europe. Published in 1917, Wind- und Wasserhosen in Europa (Tornadoes and waterspouts in Europe) is a classic of tornado research literature. Based on tornado and waterspout reports between 1456 and 1916 (258 reports), Wegener estimated that at least 100 tornadoes and waterspouts are observed each year in Europe. In my previous post I have estimated, using tornado reports from ESWD, that in average 233 tornadoes were reported each year (between 2004–2013) in Europe. In a recent study, Groenemeijer and Kühne estimated using data from ESWD, that on average approximately  480 tornadoes and waterspouts are reported across Europe each year (between 2006–2013).

If we return again to the animation and focus this time on the reports after 1930, we observe that most of these reports still came from Western Europe and that there is a lack of tornado reports over Eastern Europe. This lack of tornado reports is associated with non-meteorological factors. For example, in some socialist countries from Eastern Europe (e.g., Romania) the word tornado was forbidden in the official meteorological reports and in the mass media. After 1990, there is an increase in the number of reports over entire Europe, due to increased data collections efforts and increased public awareness.  

In my next posts, I will discuss in more details the early contributions to tornado research in Europe. If you are aware of any historical efforts on collecting and analysing tornado reports in your country I will be interested to here from you. 

---

Tornadoes in Europe

We know that approximately 1000 tornadoes are observed each year in the United States, but how many tornadoes are observed each year in Europe?  As part of my current research project (funded by AXA Research Fund) and trying to provide an answer to this question in order to assess the threat of severe convective storms (those producing tornadoes, large hail, severe wind gusts and lightning) over Europe. 

Tornado databases are maintained by few European countries and thus is difficult to evaluate the number of tornadoes that occur each year in Europe. Recently, a new pan-European tornado database has become available that will allow a step-change in our ability to observe and understand tornadoes in Europe. Thus, in my research I am using tornado data from the European Severe Weather Database (ESWD), a unique database of severe-weather maintained by European Severe Storm Laboratory (ESSL). The ESWD is a joint effort between National Meteorological and Hydrological Services and voluntary observers. Also, the public can contribute with observations (submitting a severe weather report to ESWD is very easy and it takes around 5 min. and you can make a contribution to science).

Based on the data from ESWD, 2338 tornadoes were reported between 1 January 2004 and 31 December 2013 in Europe (EU countries and Norway, Belarus, Ukraine, Moldova, Serbia, Bosnia and Herzegovina, Cyprus and Turkey). Thus, approximately 233 tornadoes are observed each year in Europe. 

Next, we can ask what is the spatial distribution of tornadoes in Europe? To answer this question, I have plotted the annual average number of tornado per square kilometre at NUTS 3 level for EU countries and at the national level for non-EU countries (Fig. 1).  Thus, most of the tornadoes are reported over Northern and Central Europe.

Figure 1. The annual average number of tornadoes per square kilometres (shaded according to the scale) at NUTS 3 level for EU countries and at the national level for non-EU countries. (click on the image to explore the details) 

The spatial distribution of tornadoes in Fig. 1 depends not only on the meteorological factors associated with the tornado occurrence (e.g., Markowski and Richardson 2014)  but also on non-meteorological factors like the population density. Figure 2 shows the average population density between 2004 and 2013 for Europe based on the data from EUROSTAT.

Figure 2. The average population density (inhabitants per square kilometre, shaded according to the scale) between 2004 and 2013 based on the data from EUROSTAT. (click on the image to explore the details)

The high population density over Northern and Central Europe, Italy or United Kingdom may result in more tornadoes begin reported since more people are living in those areas. Thus, there is a population influence on tornado reports in Europe. In my next post I will discuss this influence and how to account for the population bias.