ESSL [Research - Tornado climatology]

ESSL Research

Tornado climatology, 2000-2007 and long-term

This page will be updated annually, or when significant new data or results are available. The ESWD reporting criteria and thresholds are outlined in the ESWD data format description (ESSL Tech. Rep. 2006-01) available under "Publications".

Tornado incidence in Europe

Tornado incidence as a primary measure of hazard

This preliminary figure is based on ESWD reports of tornadoes over land. The 8yr-period covered is 2000-2007 for which a reasonably homogeneous pan-European data coverage has been achieved. The quantity depicted is incidence, i.e. number of ESWD reports per year per unit area. On a 1° × 1° latitude-longitude grid spacing, tornado reports per year per 10 000 km² has been computed from 2000 to 2007. Due to the short, sub-climatological time period, the peak values in the graph are likely exaggerated compared to the "true" climatology. The relevant range of numbers (about 1 to 2 reports per year and per 10 000 km² covers a large portion of Europe, however, confirming an estimate by Koschmieder (1946).

Tornado incidence as a primary measure of hazard
This preliminary figure is based on ESWD reports of tornadoes over land. The 8yr-period covered is 2000-2007 for which a reasonably homogeneous pan-European data coverage has been achieved. The quantity depicted is incidence, i.e. number of ESWD reports per year per unit area. On a 1° × 1° latitude-longitude grid spacing, tornado reports per year per 10 000 km² has been computed from 2000 to 2007. Due to the short, sub-climatological time period, the peak values in the graph are likely exaggerated compared to the "true" climatology. The relevant range of numbers (about 1 to 2 reports per year and per 10 000 km² covers a large portion of Europe, however, confirming an estimate by Koschmieder (1946).

Significant tornado (F2+) incidence in Europe

Significant tornado incidence as a further measure of hazard

This preliminary figure is based on ESWD reports of tornadoes over land with an intensity rating of F2 or higher (F2+) on the Fujita scale. The 8yr-period covered is 2000-2007 for which a reasonably homogeneous pan-European data coverage has been achieved. The quantity depicted is incidence, i.e. number of ESWD reports per year per unit area. On a 1° × 1° latitude-longitude grid spacing, tornado reports per year per 10 000 km² has been computed from 2000 to 2007. Due to the higher likelihood on a significant tornado being reported, the spatial distribution might already approximate the "true" climatology, although the range of numbers (up to about 1 report per year and per 10 000 km² is certainly not yet well-represented in general.

Significant tornado incidence as a further measure of hazard
This preliminary figure is based on ESWD reports of tornadoes over land with an intensity rating of F2 or higher (F2+) on the Fujita scale. The 8yr-period covered is 2000-2007 for which a reasonably homogeneous pan-European data coverage has been achieved. The quantity depicted is incidence, i.e. number of ESWD reports per year per unit area. On a 1° × 1° latitude-longitude grid spacing, tornado reports per year per 10 000 km² has been computed from 2000 to 2007. Due to the higher likelihood on a significant tornado being reported, the spatial distribution might already approximate the "true" climatology, although the range of numbers (up to about 1 report per year and per 10 000 km² is certainly not yet well-represented in general.

2000-2007 tornado intensity distribution

Based on the data shown on the above maps, the following intensity distribution over F-scale results:

F-scale	F-2	F-1	F0	F1	F2	F3	F4	F5	Sum	All reports	% rated	Avg. reports per year
Tornadoes over land	-	7	224	502	254	47	-	-	1034	2048	50.5%	256

Climatological tornado intensity distribution

Comparison of long-term tornado intensity distributions in Europe and the USA

The figure shows Tornado intensity distributions over F-scale: for Europe from 1880-2007 (1919 of 3500 reports, diamonds) with Weibull fit (red), and for the USA from 1920-1999 (45761 reports, boxes) with Weibull fit (blue). The long period of the US reports includes decades in which mainly strong (F2, F3) or violent (F4, F5) tornadoes have been reported, up to the 1990s in which detection efficiency had become so high that the vast majority of reported tornadoes were weak (F0, F1, cf. Dotzek et al., 2005). This mixture of reporting efficiencies in the USA between 1920 and 1999 resembles the currently still inhomogeneous tornado reporting in Europe. Therefore, a certain degree of similarity between the two would not be surprising.

The above figure illustrates that the present intensity distribution of tornadoes in Europe is indeed very similar to that in the USA, except for the European F0-tornadoes for which a strong underreporting still appears to persist. First, the consistency of both intensity distributions is in line with the global analyses by Dotzek et al. (2003, 2005) and Feuerstein et al. (2005) who fitted the tornado intensities by Weibull distributions. Second, judging from the US-experience over the last five decades (Verbout et al., 2006), we can expect the number of reported F0-tornadoes to strongly rise in the future, as public awareness levels and reporting standards will become more homogeneous all over Europe.