Science Seminar

Westmoore High School

12613 South Western Ave.

Oklahoma City, OK 73170





Supercell Thunderstorm, Updraft, Convection Bubble (Thermal), Convection Jet (Plume), Forward Flank Downdraft (FFD), Rear Flank Downdraft (RFD), Mesocyclone (MC), Anticyclonic Couplet (ACC), Tornadic Vortex Signature (TVS), Tornado, Wall Cloud, Funnel Cloud, Downburst, Microburst, Travelling Microburst (TMB), Gust Front, Hook Echo (HE), Bow Echo, Flanking Line, Tornadogenesis, Doppler Radar, Reflectivity, Bounded Weak Echo Region (BWER), Conservation of Angular Momentum (CAM), Dynamic Pipe Effect (DPE), Pressure Gradient Force (PGF), Thermal Gradient Field (TGF), Buoyancy Force, Evaporative Cooling, Shearing Deformation (SDF), Tilting, Stretching, Shearing Instability (SI),Convergence (CON), Divergence (DIV), Horizontal Vortex Tube (HVT), Inflow (IN), Windshear (WS), Veering Wind, Jetstream, Fujita (F) Scale





                It has long been known that most tornadoes (and all strong ones) develop between the updraft and the intense precipitation shaft or downdraft beneath a thunderstorm-scale, low-pressure system called the mesocyclone (1)(2).  Commonly located in the right (southern) half of the supercell thunderstorm (3), the clockwise circulation of the mesocyclone is created by the updraft penetrating a vertical windshear that increases in speed and changes direction (clockwise or veer) with height (4)(5)(6). What has eluded meteorologist so far is the connection between the mesocyclone and tornadogenesis, and what role, if any, the updraft and downdraft play in this process.

            Many meteorologists believe the downdraft, occurring along the right rear-flank of the supercell thunderstorm, might be the actual trigger or at least plays an integral part in the tornadogenesis process  (1)(7)(8).   This is based on field observation (9) by tornado chase teams, radar observation (1)(10), downburst damage swaths near tornado damage tracks (see figure 1) (11)(12)(13), computer simulations of the supercell thunderstorm (14)(15), and laboratory simulations of the supercell thunderstorm (8)(16).  

One tornadogenesis model (17) suggests this rear-flank downdraft (RFD) pulls the rotation occurring in the mesocyclone downward towards the ground by a vertical, shearing deformation field (SDF) (see figure 2a).  This pulling or stretching causes the radius of rotation to decrease, thereby increasing the spin of the descending column of air through the conservation of angular momentum. 

Another model (Shearing Instability) (11)(18)(19) suggests the outflow of the RFD causes the air at the surface to rotate in a cyclonic direction (see figure 2b).  This rotation is created by horizontal, directional windshear that forms along a gust front between the RFD outflow and the supercell thunderstorm’s inflow. Upward stretching, by updrafts located inside cumulus towers along this gust front or flanking line, intensify the rotation.  According to the model, when this cyclonic rotation drifts into and lines up vertically with the principle updraft and mesocyclone, tornadogenesis can occur.  By converging against the storm’s inflow, the RFD outflow can increase the cyclonic rotation created by the shearing instability.   This model may also explain why multiple tornadoes (not suction vortices) form simultaneously from the same thunderstorm.


Figure 1– Path of Bossier City, Louisiana tornado.  Note occurrence of downburst (microburst) with tornadogenesis (12).

A laboratory model (Tornado Machine) (8) that simulates the flow pattern of the supercell thunderstorm (3) has shown a high correlation between the occurrence of the RFD and tornadogenesis (see figure 3).  By simulating and altering the flow pattern inside the simulated supercell thunderstorm (16), vortexgenesis in the Tornado Machine did not occur until a downdraft of sufficient strength was created along the simulated rear-flank.  Even with a very intense updraft (including updraft pulse) (14) and the presence of a simulator mesocyclone, vortexgenesis did not occur until the simulator RFD was created in sufficient strength.

The above observation does not undermine the importance of having a relatively strong updraft and associated cyclonic rotation occurring within and below the mesocyclone.  In fact, during the Decaying Stage simulation of the Tornado Machine experiment, the intense vortex of the simulated Mature Stage dissipated when the updraft propeller was decreased to a low setting while keeping the RFD propeller at a moderate to high setting.

During vortexgenesis in the Tornado Machine, a vertical SDF existed between the updraft and downdraft propellers. From figure 4 (Mature Stage simulation), low-level convergence, as evidence from the possible bow echo like pattern in front of the vortex (11), was also present.  These observations in the Tornado Machine strongly support the theory that the vertical SDF created between the RFD and the updraft and shearing instability, created beneath the mesocyclone between the RFD outflow and supercell thunderstorm’s inflow (11)(18), contribute to tornadogenesis. 

Figure 4 – Observed airflow in the Tornado Machine set to simulate the Third Stage of tornadogenesis.  An intense vortex is observed.  Notice possible bow echo in front (south) of the vortex, with fog streamlines flowing (front to back) ahead (east) of the bow echo (16).

Experiments in the Tornado Machine (16) also showed a high correlation between the strength of the RFD and the power of the tornado, providing the updraft remained relatively strong and the RFD outflow did not choke off the low-level, cyclonic rotation occurring below the principle updraft.  This experimental result, along with downburst damage swaths observed near tornado damage tracks, suggests the downburst, imbedded in the RFD and hook echo, could also be involved in the triggering process or at least, intensify an already existing tornado.

The downburst is defined (11) as a very strong, penetrative downdraft, which can produce an outburst of damaging winds up to F3 strength.  The microburst is a smaller scale version of the downburst, but can be just as strong and violent.

The question presented in this paper and science project is what triggers the RFD and downburst.  The answer to this question may reveal the trigger for tornadogenesis or tornado intensification.  There have been several models presented in the past (11) regarding the origination or trigger of downbursts. According to Fujita (11), further attempts (research, observations and experiments) are necessary to find the mechanisms by which some selected downdrafts descend to the ground as damaging downbursts. 




A new downburst origination model (Decreasing Updraft) is presented in this paper and science project.  It is based on radar observations by Lemon and Doswell (1), in which the Bounded Weak Echo Region (BWER) around the updraft in supercell thunderstorms disappeared just prior to tornadogenesis.  According to Lemon and Doswell, the BWER is created by the uploading of water (raindrops) and ice (hail) by the very strong updraft, and its disappearance is due to the release of the water and ice by the weakening updraft.  This weakening updraft has been observed during tornadogenesis.

It is also based on accidental observations during experiments with the Tornado Machine.  When hot water was poured over the dry ice to create a thick fog to visualize the vortex, a thermal of carbon dioxide and water was frequently observed to rise from the dry ice and water mixture.  Once the thermal cooled and weakened (20), a downward rush of air was soon afterwards observed to fall back onto the surface of the Tornado Machine.  It was believed that this downward rush of air was carbon dioxide, since at the same temperature it is heavier than air.

It is proposed that the down rush of water and ice released by the decreasing updraft could possibly create a downburst in supercell thunderstorms.  It is also proposed that the Decreasing Updraft downburst can trigger or strengthen the tornado by increasing the strength of the RFD and the SDF occurring between the RFD and updraft, and its outflow and shearing instability occurring below the mesocyclone between the outflow and the thunderstorms inflow.

The purpose of this science project is to test the Decreasing Updraft model as the trigger for downbursts in supercell thunderstorms.  A laboratory apparatus using a convective jet (plume) (20) was developed and used to simulate the decreasing updraft in this project.




A review of supercell thunderstorm models (1)(21)(22)(23) (see figure 5) presented over the past forty years show many similarities in their basic structure.  All models contain an intense updraft penetrating a vertical windshear that increases in speed and veers with height.  This windshear (WS), associated with the jetstream, creates a north-to-south, horizontal vortex tube (HVT) (2)(4)(5)(6)(7).  As shown in figure 6, a mesocyclone is formed on the right (south) flank of the thunderstorm when the intense updraft pushes upward through this vertical windshear, tilting the axis of the WS-HVT into the vertical (2)(4)(5)(6).  An anticyclonic couplet (ACC) is formed on the left (north) flank (2)(4)(5)(23)(24). 

The veering wind with height (5), the rightward movement of the thunderstorm relative to the environmental wind (6)(24), and the Coriolis effect (5) favor the development of the mesocyclone over the ACC.  Computer models of the supercell thunderstorm (5)(14) also created a dominant mesocyclone over the ACC with environmental winds that veer with height.

Figure 6 - Schematic diagram of how a typical HVT (created by a vertical windshear (WS) associated with the jetstream) changes its orientation by interaction with an updraft (4). Viewing into the page (east) the mesocyclone would be on the right (south) and the ACC would be on the left (north).

A similar configuration was achieved by Connell (25) using an upward jet through a strong crossflow (see figure 7).  Mid-level, counter-rotating vortices formed immediately downstream or in the wake of the penetrating updraft or jet. The cyclonic vortex (mesocyclone) formed on the right (south) side with the anticyclonic vortex (ACC) forming on the left (north) side.  An opposing, converging flow was created between the two vortices and a downdraft was created on the rear-flank, because of the blocking updraft.

Most models listed above contain two downdrafts, one downstream (east to north) of the updraft (forward-flank downdraft or FFD) and the RFD located upwind (west to south) of the updraft (see figure 8).

All of the models above agree that the FFD originates with mid-level, outside air circulating around the storm along the right-flank and south of the updraft and mesocyclone.  The counter-clockwise (cyclonic) circulation associated with the mesocyclone pulls this outside air into the thunderstorm.  With precipitation falling downstream (26)(27) into this relatively drier outside air, the air becomes cooler and denser (due to evaporative cooling) than the air inside the thunderstorm and sinks toward the ground with the falling precipitation.  Once the FFD reaches the ground, it spreads outward, creating the forward gust front, associated self-cloud, and a thermal gradient field (TGF) with the warmer air south and east of the FFD (26).

Figure 7 - Upward Jet through a Crossflow model (25).  (a) Side view of a supercell thunderstorm updraft imbedded in a crossflow.  The plus and minus signs denote pressure excess and deficiency regions, respectively.  (b) Top view shows generated, low-pressure vortex pairs.  Bottom vortex would be the mesocyclone and the top vortex the ACC.  (c) Top view of the vortex pair with the resulting streamline flow occurring within the thunderstorm. 


With a low-level pressure gradient force (PGF) created below the rotating updraft and mesocyclone (6)(7), a low-level inflow (IN) is created south and east of the updraft.  As the inflow moves toward the updraft and parallel or streamwise (2) to the TGF, the inflow (since warmer air rises and cooler air sinks) begins rotating about a horizontal axis (right hand rule (6) with thumb pointed toward the updraft), creating a new HVT (7)(15)(24).   When the updraft moves over the IN-HVT, the updraft tilts (right hand thumb tilted upward) and stretches the IN-HVT upward (2) (see figure 2c).  This process is similar to the upward tilting and stretching of the WS-HVT.  The added cyclonic spin of the tilted and stretched IN-HVT causes the rotation of the mesocyclone to increase even more.  This increase spin causes the vertical velocity of the updraft to increase due to an increase in the upward, pressure gradient force (PGF) (14), creating the Dynamic Pipe Effect (DPE) (28)(29) (see figure 2d), associated Tornadic Vortex Signature (TVS) (30), and the wall cloud.  If the upward vertical velocity, cyclonic rotation or spin, and upward PGF is strong enough, the DPE could possibly reach the surface with the required intensity for tornadogenesis (28)(29).

One disparity in the thunderstorm models is the origination of the RFD.  In Browning (22) and Klemp (21) models (see figure 5a and 5d), the FFD is eventually wrapped around the north side of the updraft by a strengthening mesocyclone.  It eventually reaches the rear flank of the thunderstorm creating the hook echo.  In Lemon and Doswell (1) model (see figure 5c), the RFD originates as outside air impinges the upper, rear flank side of the updraft.  With the updraft acting as a barrier to this flow, it is forced downward (31) until it reaches the ground along the rear flank.  In Heymsfield (23) model (see figure 5d), the RFD originates as mid-level, outside air circulating around the storm’s left (north) flank.  It is pulled into the thunderstorm by the ACC, then rotated counter-clockwise (cyclonic) around to the rear flank by the mesocyclone.

Figure 8- Schematic plan view of a typical tornadic, supercell thunderstorm (1).  The thick line encompasses radar echo or reflectivity.  An encircled “T” shows the location of the tornado.

As stated earlier in the Introduction, the RFD can pull the rotation occurring in the mesocyclone downward toward the ground by a vertical SDF.  This pulling or stretching increases the cyclonic rotation within the mesocyclone, intensifying and lowering further the TVS.  The ensuing RFD helps create and intensify the shearing instability (SI) occurring below the mesocyclone and between the RFD outflow and the thunderstorm’s inflow.




The downburst or microburst (smaller scale) is an intense, penetrative downdraft, which spreads outward in all directions upon reaching the surface (see figure 9) (11)(32).  The lateral scale of the downburst is approximately 2-3 miles with the microburst one half to one mile.  The life cycle of a downburst, as it descends within the rain shaft, can be as long as 10-15 minutes with the maximum surface winds lasting 2-4 minutes.  The vertical speed of the downdraft can be as strong as 6000 feet per minute (approx. 60 knots) with surface winds reaching 45 knots for stationary storms (32) or F-3 strength for travelling microbursts (TMB) (11)(32) (see figure 10). 

Except for the Barrier Downburst model, all of the downburst origination models (11) presented below point to the area east to north of the updraft as the possible origination point.  They also agree that this origination occurs in the middle to upper levels of the supercell thunderstorm. Similar to the RFD, the downburst would be circulated by the mesocyclone counter-clockwise from the origination point to the rear flank of the thunderstorm.  Here, it interacts with drier outside air impinging the storm from the west.  Due to evaporative cooling, the downburst becomes more negatively buoyant, increasing its downward momentum.

Figure 9 - Vertical cross-section of the evolution of the downburst or microburst (32).  T is the time of initial divergence at the surface.

(a)                                                                         (b)

Figure 10 - Schematic diagram of the Traveling Microburst (TMB) accompanied by an extreme wind (up to F3 strength) near the ground (11).  (a) Outburst Stage.  (b) Foreburst Stage.  Note movement of extreme wind from point A in Outburst Stage to point B in Foreburst Stage.


Four downburst origination models (11) are presented below (see figure 11):


Barrier Model

This is similar to the RFD model by Lemon and Doswell (1) (see figure 5c and 11a).  The thunderstorm’s principle updraft (acting as a barrier to the flow) forces outside air (impinging the storm from the west) downward to the surface along the rear flank (31).  The downburst is created from evaporative cooling when precipitation, falling from the rear (upwind) part of the anvil, mixes with the downdraft, thereby increasing its negative buoyancy and downward momentum.  It should be noted that with the formation of the DPE around the updraft, the updraft could block any radial flow impinging on it (29).  This is due to the cyclostrophic balance between the air rushing in (pressure gradient force) and the rotating air pushing outward (centrifugal force).  The DPE could possibly be a blocking updraft.  Barnes (6) also showed how an excess of hydrostatic pressure associated with the detrainment layer in the upper half of the thunderstorm could also act as a blocking updraft.


Converging Downflow Model

            This is similar to the RFD model by Heymsfield (24) (see figure 5d and 11b). The downburst originates as mid-level, outside air circulating around the storm’s right and left flank.  Outside air along the left (north) flank is pulled into the thunderstorm by the ACC where it collides with outside air being pulled in from the right (south) flank by the mesocyclone.  This collision, occurring in the middle to upper parts of the thunderstorm, forces the air downward.  The primary mechanism forcing the outside air into the thunderstorm is counter-rotating vortices created in the wake of a rigid updraft (possibly DPE, excess hydrostatic pressure or both), which is penetrating the strong crossflow of the jetstream (25). The mesocyclone pulls the descending air around the north side of the updraft to the rear flank. 

            Both models can be explained using Connell’s Upward Jet through a Crossflow model (25) (see figure 7).  For the Barrier model, the excess pressure created against the updraft by the collision with the impinging southwest airflow would force the colliding air downward along the rear-flank (31).   For the Converging Downflow model, the rigid updraft would also force the air around it, creating the two vortex pair as illustrated in figure 7.  This intensifies the mesocyclone further, but more importantly, the ACC, since it is initially suppressed by the veering environmental winds, Coriolis effect, and the rightward movement of the thunderstorm relative to the environmental winds.


Overshooting Top Model

Cumuli-form type clouds have been frequently observed overshooting the anvil of severe thunderstorms (33)(34).  These overshooting towers (see figure 11c) were associated with very intense updrafts that penetrate the very stable and dry, stratospheric air located above the thunderstorm.  Eventually the rising air of the overshooting updraft or tower becomes cooler  (adiabatic cooling) than the surrounding air above the thunderstorm and descends.  As the tower collapses into the thunderstorm, it pulls the drier, stratospheric air into the thunderstorm.   If the collapsing overshooting top is to reach the surface as a powerful downburst, it must entrain and evaporate cloud droplets and hydrometeors from inside the thunderstorm, or otherwise the descent would stop (adiabatic warming) near the top of the thunderstorm.  According to Squires (35), the returning downdraft from the overshooting top could penetrate the “cumulus” (i.e. several km) by mixing in the cloud droplets and hydrometeors.  Interaction with the updraft will subject both to attrition, but it is possible the downdraft could reach the surface, especially if it remained unsaturated.  Even after saturation, the descending parcel or downdraft would continue to subside or descend to the surface, especially in regions of relatively drier air.  Near saturated RFD’s have been observed in supercell thunderstorms (36). 


Decreasing Updraft Model

One observation by Lemon and Doswell (1) showed that prior to tornadogenesis the updraft increases, thereby increasing the water (raindrop) and ice (hail) content surrounding the updraft in the upper region of the thunderstorm (see figure 11d).  This is seen on radar as a Bounded Weak Echo Region (BWER).  During tornadogenesis, the updraft weakens and the surrounding water and ice disappear (on radar) with an increase in the RFD.  This disappearance is believed due to the rapid downward rush, from the pull of gravity, of the rain and hail released by the weakened updraft. Once released, the mesocyclone swings the downward rush of rain and hail to the rear-flank of the thunderstorm.

Parts of the released water and ice would fall to the north and east of the updraft in the area of the FFD.  This should strengthen the FFD, but since the air in this area of the thunderstorm is already near saturation, a downburst would not be created here.  This is due to the lack of evaporative cooling, which is needed to increase the negative buoyancy of the descending air (34)(35)(37).  This is not to be confused with the initial formation of the FFD.  As discussed earlier, drier outside air is pulled into the thunderstorm along the south flank by the mesocyclone.  This drier outside air mixes with the more moist air inside the thunderstorm and descends (due to evaporative cooling) creating the FFD.

In all four models, the contribution of dry air impinging the supercell thunderstorm along the rear-flank is vital to the downward acceleration (from evaporative cooling) of the downburst toward the surface (34)(35)(37).  This is especially true if the downburst is near saturation upon reaching and mixing with this drier air (35).  Perhaps this would explain the frequent occurrence of tornadoes and severe weather events in supercell thunderstorms developing along drylines and frontal dryslots (38)(39)(40).

With the cyclonic circulation of the mesocyclone, the downburst or microburst reaches the surface along the right-rear flank within or near the hook echo. With the supercell thunderstorm moving with or to the right of the environmental winds, the downburst would be similar in structure to the Traveling Microburst model or TMB (11) (see figure 10).  The resulting eastward outflow of the downburst creates a bow echo to the right (south) of the mesocyclone and updraft (see figure 12).  This outflow increases the shearing instability occurring below the mesocyclone, which could possibly trigger the tornado or strengthen an already existing one. Overtime this stronger outflow would eventually choke off the thunderstorm’s inflow.  This choking weakens the updraft in the lower part of the supercell thunderstorm, thus shutting down the generation of rotation (DPE and upward tilting and stretching of the IN-HVT and WS-HVT) needed for tornadogenesis. 




A Plume Apparatus has been developed to test the Decreasing Updraft model for downburst origination.  Even though there is still some discussion (41) whether the updraft is a mass of rising bubbles (thermal) or an upward jet (plume), all of the Supercell models (1)((21)(22)(23), the Upward Jet through a Crossflow model. (25), and the Dynamic Pressure Distribution model (6) in figure 13 agree the updraft in supercell thunderstorms has a jet-like or plume characteristic instead of thermal one.  As a result, the simulated updraft for this experiment should be a plume.

The Plume Apparatus was designed after observing a down rush of water from decreasing plumes created in a water organ during the Dancing Waters show.  The Plume Apparatus was similarly built by projecting a water hose upward using a wooden base.  A down rush or downburst of water should be created in the apparatus by momentarily pinching the water hose.  This pinching restricts the flow of water and its upward pressure, thereby releasing the excess water supported upward by the previously strong water plume due to gravity.   Again, this down rush of water should simulate the downward rush of water and ice released by the decreasing updraft (upward jet).

If downbursts can be created in the Plume Apparatus, it is believed this would validate the radar observation by Lemon and Doswell (1) and support the Decreasing Updraft model as a possible mechanism for downburst origination in supercell thunderstorms.

Figure 13Dynamic Pressure Distribution model.  Used to estimate dynamic pressure distribution associated with jet-like updraft (6).  This model neglects the effects of rotation and interactions with ambient flow.  The bottom-shaded area is the surface, while the top is the level at which the updraft is arrested due to negative buoyancy.



The Plume Apparatus, using a water hose oriented vertically, was used to simulate the Decreasing Updraft model for downburst origination.  The method or procedures used to create the decreasing updraft and resulting downburst are given below:

1.       Conduct experiment with no ventilation or air currents.

2.       Turn water on to create a three-foot plume.

3.       Momentarily (less than 0.5 seconds) pinch the water hose supplying the plume.

4.       Observe and note results.

All of the above steps for this experiment were tested four times (trials) to support the results of this paper and science project.




                 In each of the four trials during the experiment, a plume of water was observed to upload water at the top of the plume (see figure 14a).  Once the upward flow of water (plume or jet) was momentarily restricted and weakened by pinching the water supply hose and cutting it from its source, a rush of water, released by the once stronger water plume, was observed to fall to the ground (see figure 14b).

            The results of this experiment validate the radar observation by Lemon and Doswell (1) and support the Decreasing Updraft model as the trigger of the downburst in supercell thunderstorms.




                The above results suggest that an intense updraft (possibly updraft pulse) (14), uploading water and ice in the upper regions of the supercell thunderstorm, initially sets the stage for downburst development.  This uploading is seen on radar as a BWER.  As the updraft weakens, the extra water and ice being held aloft by the previously strong updraft is released and the BWER disappears.  The ensuing down rush of water and ice follows the same path as the RFD and gains momentum from evaporative cooling along the rear flank.

The mesocyclone continues to pull the downburst around to the right-rear flank of the thunderstorm where it reaches the surface near or within the hook echo.  With the forward motion of the thunderstorm, the downburst takes on a TMB configuration (see figure 10), creating a bow echo along the right (south) flank of the mesocyclone.  TVS strengthening and lowering may occur within the mesocyclone as the stronger downburst strengthens the vertical SDF occurring between the principle updraft and RFD. 

Tornadogenesis or tornado strengthening may occur as the outflow or bow echo intensifies further the cyclonic, directional windshear and low-level convergence (shearing instability) occurring below the mesocyclone between the bow echo and the thunderstorm’s inflow.  




Overshooting Top

The Overshooting Top model (figure 11c) may be the cause for the decreasing updraft in the upper levels of the supercell thunderstorm.  As the overshooting top or tower penetrates the very stable and dry air located above the thunderstorm’s anvil, the air inside the tower becomes cooler than the surrounding air and descends.  The descent or return of the air inside the tower now opposes the air rising upward within the updraft (35).  This weakens the updraft in the upper part of the thunderstorm.  Any extra water and ice being held upward by the previously strong updraft is now released, and the downburst is created.   This may explain the frequent observation of severe weather events including tornadogenesis occurring with the collapse of overshooting tops and descending cloud tops (42)(43)(44).


Dynamic Pipe Effect

            The DPE may be the cause for the decreasing updraft in the upper level of the supercell thunderstorm.  Close examination of the DPE model (see figure 11d) shows it resembles the description of the converging updraft by Scorer and Ludlam (20).  As the air in the jet or updraft accelerates upward, it converges, thereby aiding the formation of the DPE.   For the updraft above the DPE, a diverging jet (20) more closely fits the computer results of the DPE model (29).  This would also agree with the Dynamic Pressure Distribution model (6) given in figure 13.  According to Scorer and Ludlam (20), the vertical velocity in a diverging jet increases toward the center of the jet or updraft, but the velocity decreases with height.  As the upward velocity increases, it eventually becomes so large that the laminar flow of the jet disintegrates into “meandering lumps” or eddies.  These “meandering lumps” decrease the lifting capacity of the updraft above a strengthening DPE.  The now weakened updraft releases any extra water and ice being held upward by the previously strong, laminar flowing updraft, and the downburst is created.  This model agrees with measurements of penetrating aircraft (41), where the updraft was observed to have a smooth and uniform appearance near the bottom or cloud base but a turbulent, bubbly structure near the top or anvil.


Overshooting Top/DPE

            It is quite possible that a combination of the Overshooting Top and DPE models could create the Decreasing Updraft downburst.  With the formation of the DPE, the upward velocity of the updraft will increase, creating the overshooting top and the uploading of water and ice (BWER) in the top part of the thunderstorm (see figure 15a).  As the DPE intensifies and updraft increases, the critical speed is finally reach that causes the laminar flow of the updraft to disintegrate into eddies or “meandering lumps”.  This weakens updraft in the upper part of the thunderstorm, with the resulting collapsing top weakening the updraft further (see figure 15b).  The speed and laminar flow of the updraft in the lower part of the thunderstorm would not be affected (20).  This could explain observations by aircraft flights that indicated the updraft near cloud base was smooth and uniform, but turbulent and bubbly in the upper portions of the thunderstorm (41).  This Overshooting Top/DPE combination would increase the amount of water and ice released by the decreasing updraft, and add to the power and violence of the downburst and possibly the tornado.




It has long been known that most tornadoes develop between the updraft and the intense downdraft beneath the mesocyclone.  The role of the updraft in this process is the generation of rotation (mesocyclone and TVS) through the upward tilting and stretching of horizontal vortex tubes (HVT), created by the jetstream or vertical windshear (WS) and the thermal gradient field lying streamwise with the storm’s inflow (IN).  As the rotation of the mesocyclone increases, the upward PGF and updraft velocity are increased, which increases the spin even more, creating the DPE and TVS.

With the rotating updraft setting the stage for tornadogenesis, field observations by tornado chase teams and frequent observations of downburst or microburst damage patterns near tornado damage tracks point to the RFD (including downburst) as the possible trigger for tornadogenesis. Experiments with the Tornado Machine support these observations.  In that science project, the role of the RFD and the downburst was shown to be two-fold. First, the RFD concentrates and intensifies rotation occurring with the mesocyclone (TVS strengthening and lowering) through the vertical SDF.  Second (through shearing instability), it creates and intensifies low-level cyclonic rotation through a directional windshear and convergence occurring near and below the mesocyclone between the RFD outflow or bow echo and the thunderstorm’s inflow. Through upward stretching, the principle updraft increases this cyclonic rotation further.

Experimental results in the Tornado Machine also showed a direct correlation between the strength of the RFD and the power of the tornado with the downburst creating the most powerful tornadoes.  This providing the updraft remains relatively strong and the RFD outflow does not choke off the cyclonic rotation being created by the updraft.

The question presented in this paper and science project was what triggers the downburst and the RFD.  By using the Plume Apparatus, a downburst was created by restricting the upward force of the jet or plume.  The results of this experiment support the Decreasing Updraft model as a possible trigger for downbursts in supercell thunderstorms and possibly the trigger for powerful tornadoes. 

It was proposed that a descending Overshooting Top, formation of eddies or “meandering lumps” above a strengthening DPE, or a combination of both could be the cause of the Decreasing Updraft in the upper part of the supercell thunderstorm.




                I extend my sincere appreciation to my father, Lloyd Tidwell, who as my Qualified Scientist (MS in Meteorology, College of Engineering, University of Oklahoma) and Designated Supervisor guided me through this research project and edited this manuscript.  As my dad, he helped me build the two Laboratory Apparatus and the display board for my project.  I also extend a special thanks to Mr. Bradley Brauser for being my Science Seminar Teacher and Adult Sponsor.







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

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

3.         Doswell, C. A., A. R. Moller, and R. Przybylinski, 1990. A unified set of conceptual models for variations on the supercell theme. Preprints, 16th Conf. Severe Local Storms, Kananaskis Park, Alberta Canada, Amer. Meteor. Soc., 40-45.

4. Rotunno, R., 1981: On the evolution of thunderstorm rotation.  Mon. Wea. Rev., 109, 577-586.

5. Klemp, J. B and R. Wilhelmson, 1978: Simulations of right- and left-moving storms produced through storm splitting. J. Atmos. Sci., 35, 1097-1110.

6. Barnes, S. L., 1970: Some aspects of a severe, right-moving thunderstorm deduced from mesonetwork rawinsonde observations.  J. Atmos. Sci., 27, 634-648.

7. Klemp, K. B. and R. Rutunno, 1983: A study of the tornadic region within supercell thunderstorms.  J. Atmos. Sci., 40, 359-377.

8. Tidwell, L. G., 1982: A laboratory simulation of the observed airflow within the 8 June 1974 Harrah, Oklahoma tornadic thunderstorm.  Preprints, 12th Conf. Severe Local Storms, San Antonio, Texas, Amer.  Meteor.  Soc., 314-317.

9. Golden, J. H. and D. Purcell, 1978: Life cycle of the Union City, Oklahoma tornado and comparison with waterspouts.  Mon. Wea. Rev., 106, 3-11.

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

11.        Fujita, T. T., 1981: Tornadoes and downburst in content of generalized planetary scales.  J.Atmos. Sci., 38, 1511-1534.

12.   Abbey, R.F., T.T. Fujita, and J.R. McDonald, 1982: Meteorological and engineering analysis of the Bossier City, Louisiana, tornado of 3 December 1978. Preprints, 12th Conf. Severe Local Storms, San Antonio, Texas, Amer.  Meteor.  Soc., 135-138.

13.   Forbes, G. and R. Wakimoto, 1983: A concentrated outbreak of tornadoes, downbursts and microburst, and implications regarding vortex classification. Mon. Wea. Rev., 111, 220-235.

14.   Wicker, L. J. and R. B. Wilhelmson, 1995: Simulation and analysis of tornado development and decay within a three-dimensional supercell thunderstorm.  J. Atmos. Sci., 52, 2675-2703.

15.   Trapp, R. J and B. H. Fiedler, 1995: Tornado-like vortexgenesis in a simplified numerical model. J. Atmos. Sci., 52, 3757-3778.

16.   Tidwell, A. E., 2000: The role of the downdraft in tornadogenesis.  Oklahoma Junior Academy of Science Project.  32pp.

17.   Davies-Jones, R. P., 1982: A new look at the vorticity equation with application to tornadogenesis. Preprints, 12th Conf. Severe Local Storms, San Antonio, Texas, Amer.  Meteor.  Soc., 294-252.

18.   Wakimoto, R. M. and J. W. Wilson, 1989: Non-supercell tornadoes. Mon. Wea. Rev., 117, 1113-1140.

19.        Brandes, E. A., 1977: Gust front evolution and tornado genesis as viewed by Doppler radar. J. Appl. Meteor., 19, 333-338.

20.   Scorer, R. S. and F.H. Ludlam, 1953: The bubble theory of penetrative convection. Quart. J. R. Met. Soc., 79, 96-103.

21.   Rasmussen, E. N., J. Straka, R. Davies-Jones, C. Doswell, F. Carr, M. Eilts, and D. MacGorman, 1994: Verification of the Origins of Rotation in Tornadoes EXperiment: VORTEX.  Bull. Amer. Meteor. Soc., 75, 997-1006.

22.   Browning, K. A., 1964: Airflow and precipitation trajectories within severe local storms which travel to the right of the winds.  J. Atmos. Sci., 21, 634-639.

23.   Heymsfield, G. M., 1978: Kinematic and dynamic aspects of the Harrah tornadic storm analyzed from dual-Doppler radar data.  Mon. Wea. Rev., 106, 233-254.

24.   Rotunno, R. and J. Klemp, 1985: On the rotation and propagation of simulated supercell thunderstorms. J. Atmos. Sci., 42, 271-292.

25.   Connell, J., 1975: A non-thermal mechanism for forcing cumulonimbus cloud.  J. Appl. Meteor., 14, 1406-1410.

26.   Barnes, S. L., 1978: Oklahoma thunderstorms on 29-30 April 1970.  Part I: Morphology of a tornadic storm.  Mon. Wea. Rev., 106, 673-684.

27.   Knupp, K. R. and W. R. Cotton, 1985: Initiation within precipitation convection clouds. Preprints, 14th Conf. Severe Local Storms, Indianapolis, Indiana, Amer.  Meteor.  Soc., 171-174.

28.   Smith, R. K. and L. M. Leslie, 1979: A numerical study of tornadogenesis in a rotating thunderstorm.  Quart. J. R. Met. Soc., 105, 107-127.

29.   Trapp, R. J. and R. Davies-Jones, 1997: Tornadogenesis with and without a dynamic pipe effect.  J. Atmos. Sci., 54, 113-133.

30.   Brown, R. A., L. R. Lemon, and D. W. Burgess, 1978: Tornado detection by pulsed Doppler radar.  Mon. Wea. Rev., 106, 29-38.

31.   Newton, C.W. and H. R. Newton, 1959: Dynamical interactions between large convection clouds and environment with vertical shear. J. Meteor., 16, 483-496.

32.   Aeronautical Information Manual, 2000: Federal Aviation Administration (FAA), Washington D.C., XXpp.

33.   Fujita, T.T. and F. Caracena, 1977: Analysis of three weather-related aircraft accidents.  Bull. Amer. Meteor. Soc., 58, 1164-1181.

34.   Fujita, T.T.,and H.R. Byers, 1977: Spearhead echo and downburst in the crash of an airliner.  Mon. Wea. Rev., 105, 129-146.

35.   Squires, P., 1958: Penetrative downdraughts in cumuli. Tellus, 10, 381-389.

36.   Nelson, S. P., 1977: Rear flank downdraft: A hailstorm intensification mechanism. Preprints, 10th Conf. Severe Local Storms, Omaha, NE, Amer.  Meteor.  Soc., 521-525.

37.   Srivasatava, R. C., 1985: A simple model of evaporatively driven downdraft: Application to microburst downdraft.  J. Atmos. Sci., 42, 1004-1023.

38.   Tidwell, L., 1975: A synoptic and sub-synoptic study of the June 8, 1974 severe thunderstorm and tornado outbreak in Oklahoma. M.S. Thesis, University of Oklahoma, 65pp.

39.   McCarthy, J. and S. Koch, 1982: The evolution of an Oklahoma dryline.  Part I: A meso- and subsynoptic-scale analysis. J. Atmos. Sci., 39, 225-236.

40.   Smith, D. R., E. Agee and T. Klingler, 1985: Severe (and non-severe) weather development in the upper Midwest on 27 April 1984. Preprints, 14th Conf. Severe Local Storms, Indianapolis, Indiana, Amer.  Meteor.  Soc., 305-30.

41.   Barnes, S. L. 1976: Severe local storms: Concepts and understanding.  Bull. Amer. Meteor. Soc., 57, 412-419.

42.   Burgess, D.W., 1982: Lifecycle of the Wichita Falls tornado storm. Preprints, 12th Conf. Severe Local Storms, San Antonio, TX, Amer.  Meteor.  Soc., 441-443.

43.   Alder, R. F. and D. D. Fenn, 1981: Satellite-observed cloud-top height changes in tornadic thunderstorms.  J. Appl. Meteor., 20, 1369-1375.

44.   Heymsfield, G. M., R. H. Blackmer, and S. Schotz, 1983: Upper-level structure of Oklahoma tornadic storms on 2 May 1979.  I: Radar and satellite observations.  J. Atmos. Sci., 40, 1740-1755.