CONVERGENCE-ZONE TORNADOES - A BETTER UNDERSTANDING
By TONY GILBERT
Tornado Site Investigation Coordinator and Director to the Tornado and Storm Research Organisation
15 Jessie Road, Gosport, Hampshire, PO12 3QL, UK.
E. Mail Anthony.gilbert@torro.org.uk
Abstract: The author attempts to highlight the startling differences in synoptic and atmospheric conditions associated with the development of convergence zone tornadoes by making a comparison to the development of tornadoes in a vertically sheared environment. His purpose is not to isolate the phenomena but to allow better forecasting by recognition. The risk of such vortices is rarely included in severe weather forecasting but can nevertheless pose a significant threat to life and property. The author witnessed just such an event in August 2001 causing damage and personal injury local to his home in Gosport. An immediate investigation ensued whereby damage details were recorded and relevant weather data saved. No tornado watch or warnings were available at the time of the incident. He has continued his efforts in this field of research since. His active involvement with the ‘Tornado and Storm Research Organisation’ has provided an excellent source for new information. Weather data is rarely archived in such a way as to be accessible for private research. Only by prompt notification to tornadic events can the necessary data be intercepted and saved for reference purposes. By collating this information a picture unfolds as to associated conditions for each report. Categorisation for the varied and differing conditions conducive to development will prove to have some importance. The convergence zone tornado has shown that upper air condition have little or no relevance to what may be occurring in the lower layers of the atmosphere. His findings amongst the many other associated conditions include modest surface based C.A.P.E values which are almost always present upward of 150 j/kg with a negative Lifted Index. An abundance of moisture, 80 % humidity minimum between surface and 800mb, a slight capping inversion with dry colder air directly above this level. A steep lapse rate below this level often produces a small scale 'loaded gun scenario'. A combination of explosive localised convective development over a convergence zone may be responsible for the short-lived weaker variety of tornado we often see in the summer months throughout the UK and Ireland. Delving deeper into the source of rotation using wind profile data it was concluded that horizontal shear and not vertical shear is responsible. A pre-existing surface confluence or swirl of air bought about by either an outflow boundary or more often sea breeze convergence zones. The updraft merely encounters this wind field and ingests with the usual vortex stretching that would be associated with tornado development.
KEY WORDS
Tornado
Landspout
Convergence
Spout
Non-mesocyclone
INTRODUCTION
The purpose of this paper is to present and initiate discussion on convergence zone tornadoes thus avoiding over scrutinising of technical data, which could easily detract from the main issues of theoretical debate. The concept of a tornado developing out of the accepted and commonly associated weather parameters may seem an unusual hypothesis. It is only in recent years that we have begun to accept that tornadoes can develop under a wide spectrum of atmospheric, synoptic and topographic conditions (Wakimoto and Wilson 1989). Many early textbook definitions appear to restrict the whole idea to the ‘classic supercell’ spawned or vigorous cold front scenario’, although now we do not have to look far into recent pioneering research to see how much advancement there has been in this field of knowledge. This paper examines the concept of the ‘Convergence Zone Tornado’, alternatively known to some as the ‘Landspout’ (Prof Howard Bluestein 1979, 1985, 1999). This may not seem important but the acceptance of its existence certainly is. There are many severe-weather textbooks in circulation that make no, or very little, reference to the concept in its entirety. By ignoring or disregarding the available evidence so far is to miss an opportunity to forecast such an event and ultimately to issue tornado watches, which could in every sense of the word help avoid a serious risk to life and property (Gilbert 2003).
Convergence zone tornadoes rarely last more than a few minutes. The TORRO data base for Britain and Europe show eyewitness accounts that typically suggest an average ten minutes with often one or more visible at any one time but not always in direct contact with the ground. Descriptions tend to favour the slender narrow slow-moving funnel devoid of any wall cloud formation or broad cloud-base rotation, many of which have been seen to develop off shore and on shore near coastal environments. These are particularly hazardous to small sea vessels and coastal habitats (whether temporary or fixed) of which occupy such areas. Some earlier references have been made here in the UK (Meaden 1985) as to the varying degree of characteristics shown by the different reports TORRO had received. Eyewitness accounts and site investigation have also shown such vortices to have a strong upward velocity derived by a strong inward ‘Pressure Gradient Force’ PGF. Debris tends to stay within the rotation longer rather than being centrifuged outwards. Much is dependent on the forward motion speed of the vortex.
Thursday 9th August 2001 produced just such an event to investigate. Less than a mile from his home on the south coast of England the author witnessed the formation of a funnel cloud and then developing tornado (Gilbert 2002). A site investigation was initiated within minutes. The damage survey found several flipped caravans and serious personal injury. No severe weather advisories or watches were available from any meteorological source. Later that day temporary data was saved to his computer for evaluation and further research into the causes commenced. Temporary meteorological data was archived for subsequent evaluation and analysis into the causal conditions.
TYPICAL ATMOSPHERIC AND SYNOPTIC CONDITIONS
Some notable and important detail became evident regarding the conditions that were in effect on the 9th August 2001. It is in many ways similar to ‘the classic waterspout’, (not vertically sheared type). Though the convergence zone tornado may not necessarily always be as a result of sea breeze. A typical synoptic outlook would see weak forcing near the surface associated with a stalled or slow moving cold front or trough. Upper and lower winds will often be weak in strength and may be unidirectional in nature. At this point we can understand why some forecasters could easily overlook the subtle risk of severe weather. Attention needs to be paid to the finer detail in a localised sense. The existence of an upper trough has relevance by way of altering the overall lapse rate of the atmosphere and for introducing some forcing mechanism, but whilst being a necessary ingredient, it has little or no other relevance. Warm air advection at the surface looks to be the key. This may take the form of continental warm air, solar heating or sea breezes moving inland (usually late summer onwards for the latter). A steeper lapse rate below 800mb is common with surface C.A.P.E often 150 j/kg or more.
At a glance positive “Surface to 500mb Lifted Index” could be misleading to the potential in the lowest layers. Figs. 1 and 2 show a succession of funnel clouds off Salcombe, Cornwall, UK, some of which became waterspouts. These events surprisingly took place on a morning with a recorded surface to 500mb Lifted Index of +10. The strongest dynamic forcing has also shown to be within the lowest levels, often well below 700mb. Using Vertical Velocity (V.V) data at 700mb only, could once again be misleading. VV will typically show between just –5 to -15 hpa /h. This is not critical but the height with which one look for this within the atmosphere is: - Relative humidity plays an important role with over 80% between 900-750mb (critical) and is also preferential below this level. Skew.t-p soundings show an abrupt layer of very dry air anywhere between 800mb and 500mb. The lack of upper shear all but diminishes the effects of upper diffluence or confluence as a contributing factor. So too, excludes the existence or effects of PVA (Positive Vorticity Advection) or NVA (Negative Vorticity Advection). Use of such indices has always normally been imperative to forecasters as a tool for predicting most types of severe weather, and in particular vertically sheared tornadoes. The final and to all intensive purposes the most important factor for identifying ‘risk zones’, is locating convergence (stronger wind moving into a climate of weaker winds) or confluence (the meeting of winds from opposite directions) at surface. It is these two factors, which show us regions most at risk. Models can predict land sea breeze or nocturnal sea breeze confluence with some reasonable accuracy. It is this last but critical factor that is required for such development. A classic misconception has been to try to predict such spout activity using vertical shear calculation. Associated wind profile data has shown quite the opposite, and that it is not vertical shear but horizontal shear in the majority of cases is the source of rotation (Grazulis 2001). A clashing of boundaries produce waves along the periphery resulting in vertical swirls or eddies (Fig 3). It is these invisible eddies near the surface that are merely ingested into any approaching strong updraft (Fig 4), with the usual vortex stretching process, taking place and producing a visible funnel. Using this explanation of events we can see that the funnel is actually developing from the ground upwards, which may for some go directly against many previous attempts to define such development. (Fig 5) shows an idealistic and typical sounding for a day on which convergence zone tornadoes have occurred. Note the abrupt layer of very dry air ‘B’ directly above very moist air and a relatively steep lapse rate ‘A’. ‘B’ depicts a small capping inversion creating the climate for localised and sudden vertical cloud growth where the cap is breached whilst ‘C’ represents a continued steep lapse rate up to mid levels. Low-level wind vectors show conflicting flow suggesting the existence of convergence/ confluence.
CONCLUSION
Tornado related weather data is rarely archived in the United Kingdom in such a way as to be accessible for private research and development. Only by prompt notification of tornadic events can the relevant data be intercepted and archived for reference purposes. By collating this information categorisation for the varied and differing conditions conducive to such development will prove to have some importance. TORRO maintains a central fulcrum for severe weather reporting within the UK. Its many contacts via members, public and media have proven to be very beneficial to TORRO’s ongoing research papers, projects and database. It is by benefiting from this status that TORRO can refine its search into all aspects of tornado development, and primarily toward a better understanding of the convergence zone tornado which may very well make a larger portion of our database than we currently realise.
Successful computer simulations of spout formation have already been achieved (Lee and Wilhelmson 1996) and have gone a long way in compounding the whole concept. The definition ‘Spout’ seemingly secures a noteworthy reference to this portion of the tornado spectrum. ‘Convergence zone’ or ‘Spout’ vortices are tornadoes in every sense of the word, but develop under a very different set of parameters. By recognising these parameters we can then enhance better forecasting of, what is for all intents and purposes a potentially deadly hazard. The goal is not to isolate the phenomena but to facilitate better forecasting by recognition. The risk of such vortices is rarely included in severe weather forecasting but can nevertheless pose a significant threat to life and property. The assumption that such events are not worthy of a forecast or do not fit the criteria for severe weather is somewhat misplaced in the meteorological world. It is only by including the risk potential within our general forecasting that we can truly hope to better understand this type of phenomena.
ACKNOWLEDGEMENTS
The author would like to thank TORRO directors, Terence Meaden, Robert Doe and Jonathan Webb for proof reading and suggestions. Also Angel Demitrove (TORRO representative for Bulgaria) for technical advice and to W.C. Gilbert for use of photo’s.
REFERENCES
BLUESTEIN, H. B. (1979) A Mini Tornado in California. Mon. Weather Review. 107, 1227-1229.
BLUESTEIN, H. B. (1985) The Formation of a ‘Landspout’ in a Line Squall in
Oklahoma. Preprint 14th Conference on Severe Local Storms.
America Meteorology Society. 267-70
BLUESTEIN, H. B. (1999). Tornado Alley. Oxford University Press, New York, USA. 4: 100. 7: 154-156
BRADY, R. H., and E. J. Szoke (1989) A Case Study of non-mesocyclone Tornado Development in North East Colorado.
GILBERT, A,. (2002). Tornado at Gosport, Southern England 9th August 2001. J. Meteorology, 27, 197-203.
GILBERT, A,. (2003). Note on the Theory of Tornadoes of the ‘Landspout’ Variety 9th August 2002. J. Meteorology, 28, 11-14.
GOLDEN, J, H,. (1974). ‘On the Life Cycle of Florida Keys Waterspouts I’. Journal of Applied Meteorology 13:676-92.
GOLDEN, J, H,. and D, PURCELL (1978). ‘Life Cycle of the Union City, Oklahoma Tornado and Comparison with Waterspouts’. Monthly Weather Reviews 106: 3-11
GRAZULIS, P. T. (2001). Landspouts and Waterspouts. The Tornado. (Natures Ultimate Windstorm) Oxford University Press, New York, USA 4: 69.
LEE, B, D., and R. B. WILHELMSON (1996) The Numerical Simulation of Non Supercell Tornadoes. 18th Conference on Severe Local Storms. America Meteorology Society, San Francisco. 408-11.
MEADEN, G. T. (1985). The tornado's most characteristic features. Tornadoes in Britain, page 15 (2.2) and page 16 (2.3). Published for HM Nuclear Installations Inspectorate, Bootle.
ROGER WAKIMOTO and JAMES WILSON (1989). Non-Supercell Tornadoes. Monthly Weather Review 117: 1113-40.
CAPTIONS
Fig 1. Tuba funnel cloud developing under low-topped cumulonimbus early Aug 2003 just south of Salcombe, Cornwall, UK. Copyright W. C. Gilbert.
Fig 2. A series of funnel clouds and waterspouts occurring (same date and location). Copyright W. C. Gilbert.
Fig 3. Simplified plan view diagram showing the ideal surface wind vector components (B) producing confluence/ convergence zone (A). Note vertical swirls or eddy formation (C).
Fig 4. Simplified plan view diagram of cumulonimbus (E) updraft (D) encountering eddy swirls at surface (C) along a line of confluence/convergence (A)
Fig 5. Idealistic sounding representing typical atmospheric conditions associated with convergence zone tornadoes. (A) Saturated low level air with steepest lapse rate. (B) Abrupt layer of dry air directly above moist. (D) Slight capping inversion producing localised loaded gun scenario. (C) Continued steep lapse rate.
