Mesocyclone Formation and Maintenance: A Review of Conceptual Models

Thomas Jones

School of Meteorology

University of Oklahoma

University of Oklahoma, Norman, OK 73019


Mesocyclones are the a definitive part of supercells and occur most often in the Great Plains of the United States. Mesocyclones are associated with various types of severe weather including high winds, large hail, and even tornadoes. Since mesocyclones are associated with these severe phenomena, research in the inner workings of mesocyclones is quite valuable. During 50 years of mesocyclone research, much has been learned about how they are formed and maintained, but many question still remain. This work describe the processes of mesocyclone formation and maintenance while acknowledging the many questions that still exist in this field of research.


A mesocyclone is typically defined as cyclostrophic storm scale rotation associated with a supercell thunderstorm (Lemon and Doswell III 1979). Typically the storm scale rotation is approximately 1 to 10 km in diameter and up to 12 km in height. Most significant mesocyclones rotate cyclonically, but it will be shown later that anticyclonic mesocyclones exist as well. The structure of either kind of mesocyclone is approximated by a rotating vertical column of air in which velocity in the perpendicular horizontal plane increases linearly outward from the center of the mesocyclone to a ring of maximum velocity, then the velocity decreases exponentially. This type of vortex is also known as the Rankine vortex (Fig.1). The ring of maximum horizontal velocities in the mesocyclone often in the range of 20 to 40 m s-1, but much higher values can exist.

Figure 1. The velocity profile of a Rankine vortex showing zero wind in the center increasing linearly outward to a ring of maximum velocity and decreasing exponentially thereafter.

Understanding the characteristics of mesocyclones and the conditions necessary for their formation is of great importance in meteorology. Mesocyclone research is important because mesocyclones are often the parent vortices of most tornadoes. Tornadoes kill dozens of people every year and do millions in property and crop damage. However, less than 30% of mesocyclones go on to produce tornadoes, which makes understanding why this occurs an important forecast issue with respect to the prediction of tornadoes (Stumpf et al. 1998). Mesocyclones are also responsible for other types of severe weather, such as large hail and high winds. Most observed mesocyclones have at least one of these types of severe weather associated with them (Doswell and Burgess 1993). Thus, knowing the structure of the mesocyclone can enable researchers to find correlations between certain mesocyclone characteristics and the various types of severe weather that they produce.

Research on mesocyclones began in earnest during the 1950ís with the first use of radar in the study of supercell thunderstorms, as often mesocyclones are not readily apparent to the visible observer. Early radar research revealed the hook echo associated with the supercell thunderstorms and mesocyclones. A hook echo is a "hook" seen in the reflectivity image of a radar usually on the southwest side of a rotating thunderstorm (Fig. 2). It is a result of precipitation being wrapped around the mesocyclonic circulation within a supercell thunderstorm; however, the strength of the "hook" depends on how much precipitation has been wrapped around the updraft, not necessarily on how strong the mesocyclone actually is itself (Stumpf et al.1998). Unfortunately, basic radars do not give any further information about a mesocyclone. Further information had to wait until the late 1960ís and early 1970ís for the development of Doppler radars which by detecting the Doppler shift of a returned radar pulse can deduce the radial velocity of the particles which reflected the radar pulse (Fig. 2).

Figure 2. Left side shows WSR-88D reflectivity of the supercell a tornado during the outbreak of May 3, 1999. Note the reflectivity "hook" on the south side of the storm. Right side is the radial velocity image of the same event. Note the center of the mesocyclone (where the upside-down triangle is located) is collocated with the reflectivity hook on the left image.

Instead of just the hook echo one could now determine the velocities inherent within the hook echo. With the advent of the Norman Doppler radar and its ability to scan a storm at many elevations, a 3D representations of thunderstorms as well as their mesocyclones were able to be found. Also during the 1970ís and early 1980ís, several field programs were funded to observe mesocyclones and tornadoes and the environmental conditions in which they form (Rasmussen et al. 1994). Much of the current theory of mesocyclone formation and maintenance is based on research done by those field programs (Rasmussen and Straka 1997). With the advent and nationwide deployment of WSR-88D radars in the early 1990ís, mesocyclone detection and documentation has increased exponentially. Also, algorithms have been developed that can find circulation in velocity data and give vital characteristics of the mesocyclone itself (Stumpf et al. 1998). All of these advances in mesocyclone research have allowed a greater understanding of their structure and for the first time have allowed forecasters the ability to issue tornado warnings based on detected mesocyclones characteristics.

The following is the methodology used for this work. First, the atmospheric conditions necessary for supercells and thus mesocyclones will be discussed. Then, the processes involving the actual formation of the mid-level and low-level mesocyclones followed by processes of their maintenance will be discussed. Finally, a discussion about the current questions in current mesocyclone research will be held.


Supercells and thus mesocyclones are most prevalent in the Great Plains during the months from March to June (Brandes 1978). The reason is that the atmospheric conditions in the Great Plains during the spring season are often primed for the formation of supercell thunderstorms and mesocyclones. These atmospheric conditions can be categorized under two titles, thermodynamic and dynamic, the balance of which is key for mesocyclone development.

The first necessary thermodynamic atmospheric condition is the availability of large amounts of warm moist air near the surface (Rasmussen and Straka 1997; Lemon and Doswell III 1979). The location of the Gulf of Mexico is such that given a southeasterly surface wind, much of the Great Plains will be filled with warm moist air from the Gulf. A warm, moist airmass, for the Great Plains region, is defined by temperatures in the 70ís with dewpoints in the 60ís that extend over a depth of around 100mb above the surface. The amount of warm, moist air in an airmass is proportional to the instability of a parcel within that airmass. Thus, the greater the amount of moisture and the greater the temperature, the greater the instability. The Lifted Condensation Level (LCL), or the level of cloud base, of a mechanically lifted

Figure 3a. (top) and Figure 3b. (bottom). Figure 3a shows the formation of the horizontal circulations in a speed-sheared environment and the initial tilting of those circulations into the vertical due to an updraft. Figure 3b shows a much later period of development after the updraft has split apart resulting in two separate mesocyclones both rotating in opposite directions. The cyclonically rotating mesocyclones on the RHS is perferenced in the Northern Hemisphere for strengthening.

parcel is lower in an airmass with greater moisture. The lower the LCL, the lower the Level of Free Convection (LFC) can be. The LFC is the level at which mechanical lifting is no longer needed and the parcel accelerates upward due to its own buoyancy. This buoyancy is due the fact that warm, moist parcels of air are less dense than cooler dryer air at a particular level of the atmosphere. Since a parcel of less dense air will rise in an environment of cooler, dryer air, an atmosphere of this type is said to be unstable. Convective Available Potential Energy (CAPE) is a measure of the degree of instability in the atmosphere which is also directly related to the maximum updraft possible in a supercell. Thus, the greater the CAPE, the greater the instability and the greater the potential updraft. The strength of a convective updraft is essential to the possible formation of a mesocyclone (Rasmussen and Straka 1997; Lemon and Doswell III 1979).

However, in most cases, atmospheric thermodynamics are not enough for supercell and mesocyclone formation, so dynamical effects are required. These effects manifest themselves in the form of jet streams and shear profiles. First, jet streams are rivers of very fast moving air (>50 m s-1) in the upper levels (>500mb) of the atmosphere. Since the wind velocity is often greater at upper levels than lower levels, upper air divergence is created by the net mass transport of air which is greater at upper levels. This creates a void that must be filled by air from either above or below. Since the stratosphere above is stable, the air filling this void usually comes from below, inducing rising motion in the atmosphere below the jet stream. This mechanical lifting enables a parcel to take advantage of the thermodynamic conditions and form a thunderstorm, but still the ingredient for storm rotation has not been addressed.

Vertical wind shear is the necessary ingredient for storm rotation. Vertical wind shear is the degree of change with height in the shear vector on the environmental wind vector. When the southeasterly wind that brings up the warm moist air from the Gulf of Mexico is overlaid by a jet stream whose winds are generally westerly and much stronger, significant vertical wind shear (both speed and directional) is generated. This vertical wind shear creates a tube-like circulation parallel to the ground and tangent to the direction of the jet stream increasing low level vorticity (Lemon and Doswell III 1979). This becomes very important as updrafts begin to modify this "horizontal vortex tube" of rotating air (Fig. 3a). Other sources of low-level vorticity besides vertical wind shear include mesoscale baroclinic boundaries (outflow boundaries), the supercells own forward flank baroclinic region, and even baroclinicity due to the anvil shadow of another storm.

Also, the low level vertical wind shear can be influenced by the low level jet (LLJ) which forms around the 850 mb level during the late evening from the influence of the temperature gradient between the higher elevations that cool quickly and the lower elevations that cool slowly. The LLJís effect on wind shear is usually to increase speed shear in this critical level of the atmosphere where many mesocyclones form. Thus, the vertical wind shear allows for the updraft of a developing storm to rotate and this rotation is the beginning of a mesocyclone.



The origin of a mid-level mesocyclone begins with the formation of the first towering cumulus cloud in an environment like the one described above. The consolidation of surface convergence and rising motion over a particular location is usually due to dryline forcing, frontal forcing, and mesoscale or mircoscale boundaries near the surface which have a number of possible origins (Rasmussen et al. 1994). Whatever the reason for the initial upward lift once a surface based updraft exceeds the LFC, it becomes buoyant and will accelerate upward until it reaches the tropopause. As a parcel accelerates upward, moisture inherent in the surface parcel is condensed out forming towering cumulus clouds and eventually thunderstorms. On either side of the sustained updraft are downdrafts induced by the initial updraft (Fig. 4) (Lemon and Doswell III 1979). The downdraft on the down-shear side or "forward flank" of the updraft is induced by falling precipitation, while the cause of the downdraft on the up-shear side or "rear flank", is much more complicated. However, it is this down-shear downdraft, also called the Rear Flank Downdraft (RFD), that is vital to the formation and maintenance of the low-level mesocyclones and also any tornado produced by those mesocyclones (Lemon and Doswell III 1979; Rasmussen et al. 1994; Rasmussen and Straka 1997).

The storm relative winds (environmental winds surrounding the updraft) must be strong enough to blow away rain from the updraft; thus, decreasing the possibility that the updraft will be weakened by rainfall and collapse (Rasmussen et al. 1994). Then, updrafts begin to rotate at mid-levels as a result of

Figure 4. Plan view of a supercell thunderstorm showing the locations of the rear flank downdraft (RFD) and forward flank downdraft (FFD) relative to the structure of the supercell. Note the location of the RFD in left rear quadrant of the supercell just on the rear (southwest) side of the main updraft.

tilting caused by the updraft, stretching of the positive low-level vorticity, and positive vorticity produced by baroclinic interactions (Rasmussen and Straka 1997). The stretching and baroclinic generation of vorticity acts to reduce the diameter of the circulation while increasing wind velocities within the vortex (Lemon and Doswell III 1979). Thus, the horizontal vortex tube described above is tilted into the vertical by the updraft and strengthened by the other two processes. This process creates two vertical circulations on either side of the updraft: an anticyclonic circulation on the left side and a cyclonic one on the right side both of which are stronger in magnitude than was circulation of the horizontal vortex tube (Fig. 3b).

In the Great Plains, the wind shear profile usually consists of wind veering and increasing with height. This vertical wind structure favors the cyclonic circulation in an updraft. Thus, the updraft splits due to a downdraft induced by precipitation between the updrafts. The anticyclonic updraft usually turns lefts of the mean flow and weakens, while the cyclonic circulation turns to the right of the mean flow and strengthens (Lemon and Doswell III 1979). This is valid since the low-level environmental vorticity that becomes tilted by the updraft usually has a large streamwise component with respect to the storm relative winds. In numerical simulations, it has been shown that an isolated updraft with those atmospheric conditions will result with a net cyclonic rotation (Doswell and Brooks 1993). In cases where the low-level vorticity is crosswise with respect to the storm relative wind, both anticyclonic and cyclonic portions of the updraft can persist for an extended period of time, which would lead to an updraft with no net rotation. When the cyclonic updraft becomes independent of other updrafts, the thunderstorm produced by that updraft becomes a supercell and the rotating updraft becomes a mesocyclone. However, at this stage of development, the storm rotation is only significant in the mid-levels of the atmosphere (3 to 7 km) (Brooks et al., 1994). The intense stretching of environmental parcels drawn into the updraft creates significant amounts of positive vorticity that allows for the maintenance and strengthening of the mid-level mesocyclone (Wakimoto et al. 1997).



The low-level mesocyclone is, which is crucial for tornadogenesis, forms from different mechanisms than does the mid-level mesocyclone though both eventually will merge to form a single mesocyclone (Rasmussen et al. 1994). In the lowest one or two kilometers of the atmosphere, the formation of low-level rotation below the mid-level mesocyclone awaits the formation of a rain-cooled downdraft which is known as the rear flank downdraft or RFD. The RFD is a downdraft that usually develops in the rear side of the updraft near the top of a supercell (8-10 km) and spirals down on the outside of the updraft (Rasmussen and Straka 1997). The exact cause of the RFD remains unclear, but may be due in part to a downward directed, non-hydrostatic pressure force created by increased cyclonic rotation of the mesocyclone (Fig. 5) (Lemon and Doswell III 1979). Whatever the cause, the sinking air gives rise to a cooler, dryer outflow near the rear base of the mid-level mesocyclone.

The low-level mesocyclone first develops in this region of evaporatively cooled, sinking air where intense temperature and vorticity gradients form due to the proximity of warm, rising air to cool, sinking air. The low-level circulation begins to form when the sinking cooler air traveling parallel to much warmer air just outside the rain cooled downdraft begins to rotate due to the positive (cyclonic) vorticity generated by buoyancy torques (Rasmussen et al. 1994, Brooks et al., 1994). Buoyancy torques occur because of the tendency for an air parcel in an environment of a large temperature gradient (a baroclinic zone) to begin to rotate about the horizontal axis of that gradient since the warm side is more buoyant (less dense) than the cool side. Since this is occurring on the left (or rear side) of the mid-level mesocyclone, the upper part of this new low-level circulation begins to become entrained into the rear side of the mid-level mesocyclone. The entrainment process stretches the low-level circulation, increasing its vorticity, and thus rotational velocity. At this point, the low-level circulation and mid-level circulation become one and a full-fledged mesocyclone is born.

Figure 5. Enhanced view of the rear flank region of a supercell thunderstorm. Shown is the mesocyclone and associates updraft (rings). Also the downward directed PGF is shown by the black and white arrow on the left.


In stronger mesocyclones, the updraft is tilted up-shear to the storm relative environment (i.e. the updraft is tilted against the storm relative wind flow). This creates a situation in which the kinetic energy of the storm relative to the environment can be transferred to the mesocyclone, increasing its kinetic energy. This allows for strengthening of the mid-level mesocyclone which increases the updraft, increasing positive vorticity tilting and stretching, acting as a positive feedback mechanism for the maintenance of the mesocyclone. If the updraft becomes tilted down-shear, then this process is reversed and kinetic energy is taken from the mesocyclone and transferred to the environment, weakening the mesocyclone.

The strength and maintenance of many mesocyclones is highly dependent on the low-level balance of baroclinic positive vorticity with outflow development (Brooks et al. 1994). It has been shown that the vertical velocity of the updraft associated with a mesocyclone is directly proportional to the low-level vorticity produced by the baroclinic zone near the rain-cooled downdraft. However, if the outflow were to become to strong, the cool air embedded in the outflow would cut off the mid-level mesocyclone; thus, cutting off its surface updraft. This would lead to weakening of the mesocyclone since low-level vorticity production would be pushed far clear of the updraft and not allow it to be entrained into the mid-level mesocyclone. The balance between instability and wind shear that allows a mesocyclone to form is also essential to its survival. If the supercell containing the mesocyclones moves to a slightly different airmass where either the shear or the instability is weaker, the maintenance of the updraft is compromised and with it the maintenance of the mesocyclone.

A mesocyclone weakens significantly whenever the updraft is perturbed. If the updraft is destroyed all together, the mesocyclone loses the energy source required to maintain the rotation, and the rotation fills in as the remaining updrafts turns in to a rain- dominated downdraft. Since the balance of atmospheric conditions necessary for supercells and mesocyclones are very fragile, its duration is usually very limited, on the order of one to two hours (Stumpf et al. 1998). However, there are rare times when that atmospheric balance can be maintained much longer allowing for much longer lived and stronger mesocyclones. It is during these times that major tornado outbreaks like the events of April 10, 1979 and May 3, 1999 occur.



Over the past 50 years, meteorologists have learned details about mesocyclones, but many other details are still unknown. One current topic of research involves the formation of low-level mesocyclones. As stated previously, current theory suggests that low-level mesocyclones are a result of the evaporatively cooled downdraft on the rear flank of the mid-level circulation. This implies that the amount of moisture available at low levels affects amount of precipitation generated which affects the potential for evaporative cooling could play a significant role in the formation of low-level mesocyclones (Brooks et al. 1994). Current research suggests that the more moisture available at low-levels (though not up to saturation), the more cyclonic low level vorticity is generated and thus the more likely it becomes a low-level circulation may form. Unfortunately, very little quantitative work has been done on the amount of moisture necessary for low-level mesocyclone formation.

In addition, the formative processes of the RFD remains somewhat unclear. The basic theory of the non-hydrostatic pressure gradient force has been around for over 20 years (Lemon and Doswell III 1979). However, more recent research has led to the conclusion that other factors, such as interaction between storm relative winds and the mesocyclone and evaporative cooling at upper-levels, have an impact on the formation of the RFD (Brooks et al. 1994; Rasmussen and Straka 1997)

Another big question that has yet to be answered is why some mesocyclones produce tornadoes and others do not. The Verification of the Origins of Rotation in Tornadoes Experiment (VORTEX) of 1994 and 1995 leads to a better understanding of this phenomenon (Rasmussen et al. 1994). Unfortunately, no definitive answers concerning tornadogenesis have emerged from work carried out during VORTEX, though the project was considered a great success (Rasmussen and Straka 1997).

Radar detection techniques of mesocyclones continue to improve. NSSL recently completed work on the Mesocyclone Detection Algorithm (MD which was the first storm-scale vortex algorithm that used exclusively WSR-88D data during its development (Stumpf et al. 1998). Still the MDA has many limitations in the types of circulations it can detect. The MDA only detects cyclonic circulations and perfectly circular rotation, where in the real atmosphere, anticyclonic and asymmetric mesocyclones do exist. (Stumpf et al. 1998). With the limitations inherent in the WSR-88D velocity itself, algorithm detections of mesocyclones and their associated attributes still have a long way to go. The development of the phased array radar in Norman, OK over the next few years should overcome many of the limitations associated with the WSR-88D, improving radar derived information of mesocyclones dramatically. Still, the fundamental problem with lack of reflectivity around some mesocyclones remains, which calls into question how much of a mesocyclone can be detected using radiometric processes (Wakimoto et al. 1997).


Bluestien, H.B., 1999: Tornado alley: monster storms of the great plains., Oxford Uni. Press, New York.

Brandes, E. A., 1978: Mesocyclone evolution and tornadogenesis: some observations., Mon. Wea. Rev., 106, 995-1011.

Brooks, H. B., C. A. Doswell III, and R. B. Wilhelmson, 1994: The role of midtropospheric winds in the evolution and maintenance of low- level mesocyclones., Mon. Wea. Rev., 122, 126- 136.

Brooks, H.B., C. A. Doswell III, and J. Cooper, 1994: On the environment of tornadic and nontornadic mesocyclones.,Wea. Forecasting, 10, 606-618.

Davies-Jones, R. P., 1984: Streamwise vorticity: The origin of updraft rotation in supercell storms., J. Atmos. Sci., 41, 2991-3006.

Davies-Jones, R. P., C.A Doswell III, and H. B. Brooks, 1994: Initiation and evolution of updraft rotation within an incipient supercell thunderstorm., J. Atmos. Sci., 51, 326-331.

Doswell, C. A., and D. W. Burgess, 1993: Tornadoes and tornadic storms: A review of conceptual models. The Tornado: Its Structure, Dynamics,

Prediction, and Hazards., No. 79, Amer. Geophys. Union, 161-172.

Lemon, L. R., and C. A. Doswell III, 1979: Severe thunderstorm evolution and mesocyclone structure as related to tornadogenesis., Mon. Wea. Rev., 107, 1184-1197.

Rasmussen, E. N., J. M. Straka, R. P. Davies-Jones, C.A. Doswell III, F. H., Carr, M. D.Eilts, and

D.R. MacGorman, 1994: The Verifications of the Origins of Rotation in Tornadoes Experiment: VORTEX., Bull. Amer. Meteor.Soc., 75, 997-1006.

Rasmussen, E. N., and J. M. Straka, 1997: Tornadogenesis: a review and a new conceptual model, submitted to Mon. Wea. Rev.

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.

Trapp, R. J., 1998: Observations of nontronadic low- level mesocyclones and attendant tornadogenesis failure during VORTEX,Accepted for publication in Mon. Wea. Rev.

Wakimoto, R. M., C. -H. Liu, and H. Cai, 1997: The Garden City, Kansas storm during VORTEX95. Overview of the storm's life cycle and mesocyclogenesis. Accepted for publication in Mon. Wea. Rev.

Wakimoto, R. M., and C. -H. Liu, 1997: The Garden City, Kansas storm during VORTEX-95. Part II: The wall cloud and tornado. Accepted for publication in Mon. Wea. Rev.

OU Home | Disclaimer | Copyright | Equal Opportunity | OU Web Policy