EAS 486 Lecture Content for Day 17: Non-Supercell Tornadoes

  1. Non-Supercell Tornadoes (Wakimoto and Wilson, 1989, MWR)
    1. Most of studies done on supercell storms and tornadoes
      1. Justification: Responsible for more than 90% of strong, 99% of violent tornadoes
      2. Have mesocyclone
      3. Latest thinking from individual VORTEX case(s)
        1. downdraft "blob" streamers of high reflectivity intersect ground in observed tornadoes
        2. Air diverges away from base of downdraft streamer
        3. Induces both cyclonic and anticyclonic rotation along rear-flank downdraft gust front
        4. As streamer moves horizontally along near gust front, gets pulled into updraft-downdraft junction around gust front bulge.
        5. Vortex-tube stretching converts streamwise vorticity into vertical vorticity
        6. Only possible with "warm downdraft" having theta-E the same as in base of updraft (?????????)
    2. History of non-supercell tornado research
      1. Bates (1968) showed tornadoes along "flanking line" of clouds
      2. Burgess and Donaldson (1979):
        1. Non-supercell tornadoes likely to be weaker, smaller, and have more brief duration than supercell tornadoes.
        2. Suggested a similarity to waterspout (forms in updraft stage of thunderstorm)
        3. Discussed "Cold Air Funnels" - small pendant cloud from towering cumulus in updraft stage which "never touch down"
      3. Bluestein (1980): Some of the non-supercell tornadoes appear to form along gust front (gustnado) or within inflow air itself
      4. Carbone (1982, 1983)
        1. Non-Supercell tornadoes forming along severe frontal rainband
        2. Mechanism: shearing instability
      5. Forbes and Wakimoto (1983)
        1. Found tornadoes forming along leading edge of outflows associated with bow echoes.
        2. Development seemed to be in response to local PBL vorticity generated in association with the horizontal and vertical wind shears on the edge of microbursts
        3. The two strongest tornadoes were of F2 or F3 intensity
        4. Several did not possess funnel clouds, nor were pendant from Cb
      6. Bluestein (1985)
        1. Coined landspout
        2. Tornado which generated under a line of rapidly growing cumulus towers, similar to the development of a waterspout
        3. No mesocyclone (environment was not favorable)
        4. Hypothesis: Vortex-tube stretching due to intersecting outflows
      7. Brady and Szoke (1988): Colorado landspout
      8. Wilson (1986)
        1. Single Doppler radar study
        2. Organization along non-precipitation-induced shear line
        3. No mid-level circulation
        4. Vertical extent was less than 3 km
        5. Would not have verified using original WSR-88D Tornado Vortex Signature algorithm
      9. Non-supercell tornadoes are assumed to be weak
        1. Fujita (1979) documented one fatality
        2. Forbes and Wakimoto (1983) found F3 damage
        3. Wakimoto (1984) found F4 damage
    3. This study resulted from CINDE (Convective Initiation and Downburst Experiment) experiment
      1. Purpose: Document causes of frequent dry microbursts to lee of Colorado Front Range during summer 1987
      2. Visually documented 27 cases of vortices without the presence of supercell thunderstorms during 47 days
        1. Dust devils were excluded (no case without a convective cloud overhead)
        2. Any radar circulation had to be of a diameter < 4 km (misocyclone)
        3. Circulations was surface-based and did not extend to cloud base (only 4 eventually were connected from the ground to cloud)
        4. Average lifetime: 7 minutes
      3. Large-scale conditions were unfavorable for tornado development, but local shear was forced by the presence of surface boundaries
      4. Radar characteristics
        1. Most formed in association with cyclonic circulations along radar-detected boundary or develop in response to collision of boundaries
        2. Almost all form at low-levels (0-2 km) and grow in vertical depth just prior to becoming visible
        3. 17 of 27 vortices were detected by radar
          1. Small size may not be detectable beyond 40 km range
        4. Radar vortex leads visual vortex by average of 14 min (longest 32 min)
          1. Vortex dimensions:
            1. Diameter <= 2 km
            2. At time of maximum shear, diameter was typically 130-1000 m
            3. Maximum depth: 0.8-7.9 km
          2. All smaller and shallower than mesocyclone signatures
          3. One radar vortex even displayed a "mini-hook echo"
          4. Detection problem
            1. Shear area moves under updraft cloud, so echo increases strength
            2. But then drops sharply in intensity
            3. May be radar problem as vortex shrinks and concentrates below radar resolution
        5. Characteristics of parent clouds
          1. Cumuli or Cumulus Congestus stage (become thunderstorms later)
          2. Developing stage:
            1. no downdrafts
            2. inferred strong updraft from appearance (flat, dark cloud base; increasing echo intensity)
        6. Low to modest instability (CAPE < 1100 J/kg)
        7. Weak wind shear (Bulk Richardson Number: 15-45)
          1. Combination not considered by Weisman and Klemp (1984)
        8. Working hypothesis
          1. Circulations develop along boundaries (Denver Vorticity Convergence Zone, outflow boundaries, etc.)
            1. Spun up by Kelvin-Helmholtz or other instabilities along shear line
            2. Theoretical spacing of K-H waves
              1. 5-7 km along gust front
              2. 13 km along cold front
            3. A vortex with an average pressure drop 3.0 mb could be created by a shear line with wind speed increase of 19.5 m s-1
          2. Move along boundary
          3. When developing cumulus moves from mountains or other source over circulation, tornado forms through vortex tube stretching

Last updated: April 7, 2009

Return to EAS 486 Page

Send comments to Bob Weisman