When most people think of tornadoes, images of destructive twisters careening across the plains of Oklahoma typically come to mind. So it may come as a surprise that England, the largest of the UK’s constituent countries, experiences more tornadoes per area per year than any country in the world, even the US. How can this be?
Tornadoes in the US are mostly concentrated east of the Rocky Mountains, particularly in the southern plains of “Tornado Alley.” By contrast, tornadoes occur throughout England, whose area is roughly that of Alabama. Between 1980 and 2012 more than 75% of tornadoes affecting the British Isles (defined here as England, Scotland, Wales, Northern Ireland, the Republic of Ireland, the Channel Islands, and the Isle of Man) occurred in England. The remainder were reported in the Republic of Ireland, Northern Ireland, Scotland, and parts of Wales. Tornadoes were most likely to occur in the south, southeast, and west of England, and in southern Wales.
A street in Moseley, a suburb of Birmingham, England, the day after a tornado struck the city on 28 July 2005. CREDIT: Liftarn
Comparing the absolute annual number of tornadoes, the US averages 1253 tornadoes per year compared with 34 per year in England. But even though the likelihood of a tornado on any day in England and the rest of the British Isles is less than in the US, tornadoes in some parts of the British Isles have a return period of 17 years.
Like their US counterparts, British Isles tornadoes threaten life and damage property. For example, the F2 tornado that hit Birmingham, England, in 2005 caused 19 injuries and £40 million in damage. How, then, are tornadoes formed in a country with a climate similar to that of the placid US Pacific Northwest instead of one similar to the that of the volatile Central Plains?
Determining the type of storm that begets tornadoes helps forecasters quickly spot potential tornadic storms. Using radar data from 2004–12, my thesis adviser David Schultz and I classified parent storms by date, time, and location. British Isles tornadoes were most commonly formed from linear storms, which can occur along cold fronts. In the US between 1998 and 2000, only 18% of tornadoes were formed from linear storms.
From a forecasting perspective, it is helpful to know how British Isles tornadoes form. But forecasters still need to know whether a particular linear storm will produce a tornado, and that is a trickier question.
One way to classify storm types, adapted from W. A. Gallus et al., Weather and Forecasting 23, 101 (2008).
Tornadoes require several preconditions for formation, so to distinguish a tornadic from a non-tornadic storm, scientists can tally conditions in an ingredients-based approach. First, deep, moist convection must occur, and this requires three ingredients: moisture, lift, and instability (that is, the rapid decrease of temperature with height). Other essential ingredients include a low cloud base and wind shear—that is, the change of wind speed and direction with height, which produces vorticity. (To learn more about the formation of tornadoes, see Paul Markowski and Yvette Richardson's article, which appeared on page 26 of the September 2014 issue of Physics Today.)
Using the ingredients-based approach with radiosonde data, Schultz and I compared tornadic versus non-tornadic environments in the British Isles. We also compared tornadic environments in the British Isles with those in the US.
Radiosondes are launched twice daily around the world and measure temperature, dewpoint, wind speed, and wind direction with height. Parameters from radiosondes that were useful for our tornado study include
- • the lifted condensation level (LCL) height, which approximates the height of the cloud base,
- • convective available potential energy (CAPE), which determines the buoyancy of a parcel that has been lifted from the ground, or instability, and
- • wind shear.
We analyzed two depths of wind shear: deep-layer (between 0−6 km above ground level) and low-level (between 0−1 km above ground level). Deep-layer shear promotes long-lived storms. When its magnitude is high, it can be used to distinguish supercell from nonsupercell environments. Low-level shear implies horizontal vorticity near the ground. If tilted upward, this horizontal vorticity can create a rotating updraft a few kilometers above the surface, thus classifying a storm as a supercell. Low-level shear has proven to be a good predictive discriminator between tornadic and non-tornadic storms.
Using data gathered by radiosondes within three hours and 180 km of tornado reports in the British Isles, as well as reports of non-tornadic thunderstorms, Schultz and I looked for any clear threshold values in LCL height, CAPE, and wind shear to help distinguish a tornadic from a non-tornadic environment. We then compared those values to their equivalents determined from US data.
Mean LCL heights in the British Isles turned out to be significantly lower for tornadic cases than for non-tornadic cases. The finding indicates that lower cloud bases and more moisture are available in tornadic cases. In the US during 1997−99, 75% of significant tornadoes (tornadoes F2 or greater in intensity) and less than 50% of non-tornadic cases occurred in environments with LCL heights lower than 1200 m. In the British Isles, 90% of tornadic and 88% of non-tornadic LCL heights were lower than 1200 m. In other words, the US threshold of LCL heights was not relevant in the British Isles, where LCL heights are lower in both tornadic and non-tornadic environments.
Less unstable, less low-level shear
Mean CAPE in the British Isles was significantly higher in tornadic cases compared with non-tornadic cases. Tornadic environments were less stable than non-tornadic environments, as the ingredients-based approach would suggest. However, there were some cases in the British Isles in which tornadoes occurred in zero CAPE and other cases where no tornado occurred in non-zero CAPE. CAPE in the US can exceed 4000 J/kg whereas the highest tornadic CAPE value in the British Isles during 1980−2012 was only 1388 J/kg. Overall, the air that gives rise to tornadoes in the British Isles is less unstable than is the case in US. Consequently, because tornadoes can occur in zero CAPE conditions, other ingredients must be considered in forecasting tornadic storms.
In the British Isles, we found no significant difference in low-level wind shear between tornadic and non-tornadic conditions. However, deep-layer shear was significantly lower under tornadic than under non-tornadic conditions. Those findings seem contrary to the ingredients-based approach, which presumes that high wind shear promotes tornadic storms. However, when Schultz and I separated significant tornadoes and tornado outbreaks (days during which three or more tornadoes occurred), the amount of wind shear was higher than both non-tornadic and all tornadic cases.
A 1992 survey of soundings taken close to tornadoes in the US found that the median deep-layer wind shear for non-tornadic, non-supercellular storms (those with a rotating updraft) was 10.8 m/s; for tornadic supercells, the deep-layer wind shear was 19.1 m/s. Although the median values of deep-level wind shear of British Isles tornadic supercells and US tornadic supercells were similar, the US has much less non-tornadic deep-level shear than does the British Isles. High-shear environments in the British Isles are common. Even though the difference between non-tornadic and tornadic environmental shear was significant in the British Isles, the difference between the mean (and median) values was small enough that it would be difficult to set a threshold for forecasters trying to warn for tornadoes.
The results from our research help clarify the types of storms and environments from which tornadoes form in the British Isles. The ingredients necessary to produce tornadoes are the same worldwide, but the magnitudes of the ingredients are different in the British Isles and the US, as are the type of storms most commonly producing tornadoes. The next big questions are why and how tornadoes are being produced from linear storms in the British Isles. Then we must derive useful tools for forecasters to predict tornadoes along linear storms in real time. Those questions will form the focus of my future research.
Kelsey Mulder is a graduate student at the University of Manchester’s Centre for Atmospheric Science. To learn more about her research on tornadoes, see K. J. Mulder, D. M. Schultz, Monthly Weather Review 143, 2224 (2015).