At approx 1800hrs 3/10/01 a heavy shower commenced here at Leopold, the shower  and occasional thunder was of a cold air type that had drifted from southwest of Geelong over the previous 20 minutes. As this CB moved north east of Leopold a flanking line came into view which extended westward from the main rain area.  This flanking line showed massive and rapidly growing updrafts and ragged lowerings...all back lit by the low sun.  At approx 1810 the very rear of the flanking line and immediately under a very strong updraft showed what I would describe as weak rotation, this area of rotation rapidly intensified and formed a small funnel. The funnel then moved just north of Leopold and increased in length and actually touched down at approx 1814hrs in a grass paddock. (no debris seen) Then lifted and continued to move northeast as a funnel initially weakening and then intensifying. By 1820 the funnel now 3 or 4 klm northeast of Leopold became large and touched down again, the tornado was occurring immediately under a massive and persistent updraft and was maintained for at least another 10 minutes, and finally dissipated over Port Phillip Bay at about 1832, however the wall cloud continued to show rotation and was rather large, at about 1840 another smaller funnel developed along the infeed flanking line but only lasted about 3 minutes.

The video sequence lasts about 5 minutes and shows the tornado near to its most intense phase with another view lasting about 39 seconds showing an almost wedge shaped tornado. The early stages of this tornado were unable to be videoed due to water in the camera!!! and the auto dew cut out was activated I need not tell any ASWA member what that made me feel like........ The passage of this storm across the Geelong area seems to have coincided with the passage of a thermal trough although I have not checked the soundings I believe the 500hpa temp was close to -24c surface temps were not high being around 13c however the low to mid layers were very moist.

Harald Richter (Bureau of Meteorology) - 4th October - Discussion & Atmospheric Analysis

(1) It is hard to judge from the "Geelong Info" picture what the funnel is actually attached to. There seems to be a low rainfree base in the background, but there's also a lighter-coloured lowering in the foreground. Clyve, was that lowering rotating with the tornado, i.e. could that be indicative of a low-level mesocyclone?

(2) The Laverton radar volume scans reveal a rather nondescript storm:
1800 LT: 43 dBZ core extends from ground to ~1 km AGL
1810 LT: 61 dBZ core developed centred around 2 km AGL;
50+ dBZ reflectivity extending from 1 to 3 km AGL

Formation of an echo overhang on the N side of the storm
1820 LT: 62.5 dBZ core (c.f. "red" starts at 55 dBZ) centred at 1 km AGL
50+ dBZ extend from ground -> 2.5 km AGL
40+ dBZs remain below 4 km AGL at all times !!
1830 LT: max dBZ ~ 51 dBZ at 1.5 km AGL
1840 LT: a new 60 dBZ core appears near the ground from the northernmost of 3 reflectivity cores ( the S core was associated with Clyve's TOR)

The amazing upshot of the reflectivity pattern is:
(a) the storm does not look like a supercell in the reflectivity pattern IF this was the final word on the storm type, the TOR would be a "landspout" or "non-supercell tornado"
(b) the 40 dBZ echo top never exceeded 4 km AGL - that's extremely low-topped for a tornadic storm (I am not sure whether any lower-topped tornadic storm has ever been documented)

(3) Surface mesoscale data (which I haven't seen yet) showed a surface trough/wind shift line in the region. Behind (W of) that line dewpoints were higher, and the wind direction changed from SW to W across the boundary. The storm probably crossed / rode along that boundary around its tornadic phase.

(4) The MML 23/2 UTC (0900 LT) sounding showed about 5 g/kg LL moisture quickly dropping off with height, a dry adiabatic lapse rate from the ground up to ~750 hPa (!) and a lacklustre wind field with near calm at ground level, and 20 knot westerlies around 500 hPa. The MML 11/3 UTC (2100 LT) sounding had 6-6.5 g/kg of moisture up to ~920 hPa, and had a dry adiabat running into the surface parcel LCL at ~920 hPa! The wind field has tightened up considerably with 10-20 knot W/SW near the surface turning to 40 knot SW around 500 hPa.

(5) The 23 UTC MesoLAPS run showed a cold pool (initialised form the 23 UTC MML sounding) that slowly drifted E. Leopold would have been along the back edge of the coldest mid-level air. MesoLAPS also showed a 25kt mid-level flow at 2200 LT, which is a little weaker than the mid-level winds in the 11 UTC MML sounding.

(6) The lightning network showed only a couple of strikes W and SW of Leopold between 1800-1810 LT. Maybe the tornadic cell was simply too shallow to produce all the microphysiscs needed for a lightning show. This also shows that lightning and tornadogenesis can happen apart from each other.

(7) The 0532 UTC VIS GMS-5 image shows that the cell was embedded in a disconnected line of cells arching counter-clockwise from King Island to Geelong to Ballarat to Horsham. A mesolow between King Island and Colac seems to have advected "useful" air across Cape Otway into VIC behind the surface boundary mentioned above.

My current hypothesis is that we had deep, dry adiabatic lapse rates from the ground up, combined with an increase in LL moisture on the NW periphery of a parent surface low somewhere E of TAS. This combination led to a short fat CAPE profile in the LL, and hence could be accomodated in a short fat storm. The surface boundary supplied extra lift and vorticity to the updraft, which took the vorticty, tilted and stretched it.

Forecasting an event like this with an acceptable FAR (False Alarm Ratio) is beyond anyone's capability ATM. I wish nothing more than being proven wrong on this statement! I suppose that the "discriminating " factors that made this overgrown Cumulus :) (remember: 40 dBZ cloud top at 4 km AGL) tornadic are tucked away into small time and spatial scales such as LL lapse rates at 1800 LT E of Geelong (not 11 UTC, not 23 UTC, not Melbourne Airport).

I would like to know more about what boundaries were in the area around 1810 LT, and what their vertical structure was.

Clyve Herbert (ASWA Victoria) - 5th October

Harald.
I am extremely interested in your reply, and I will endeavour to answer (from a visual perspective) your request for further information. At the moment I am still suffering from only what I can describe as retina retention of what I saw (I cant get it out of my mind),I am also continually analysing and trying to remember everything I observed. In respect to the local scale low you mentioned I observed this just south of Cape Otway on the Vic visual satpic for 1230hrs Mel time, it seems this local/meso scale low had moved northeast towards the Geelong area by 1530hrs local time. I feel the presence of this meso feature is a significant aspect of the study I will carry out on the occurrence of this tornado. In regards to the lightning you are correct I detected only two discharges prior to the storm moving over the Leopold area and this was when the system was just southwest of my location (I photographed what I would describe as a small multicell of rather low tops). Another area worthy of investigation is the possible presence of convergence points to the lee of the Otway range in a southwest flow this may have also aided in updraft potential. The tornado became very large between 1815 and 1820 and at times almost seemed wedge like, at this time the tornado had passed onto Corio Bay and effectively became a water spout. I would be keen to receive any ideas and suggestions.

Clyve Herbert (ASWA Victoria) - 6th October - Preliminary Damage Report

Most of the area affected by the tornado is grassland with a few trees and also private property with guard dogs!?.When this funnel first touched down it was rather weak, however I did find minor track damage near to the Sands Caravan Park with branches broken and on the ground also leaf litter etc, this appeared over a narrow track, also grass in the nearby paddock was showing a chaotic appearance. The locals I talked to indicated that they heard some noise about 1800 but could not be sure. A little further east about 500m I found some tea trees snapped off but nothing significant. From this point it seems the tornado skirted the coast and then moved onto Corio Bay. I will continue further investigation today. In respect to the description as a 'land spout' type this tornado did not exhibit this appearance, at its maximum (ex Video) it had an appearance of a large tornado almost wedge like complete with large rotating wall cloud.

Bill Hark - USA - 30th October

Great video still of the Leopold tornado from early October. The structure is interesting. From the surrounding clouds, I would expect a tornado more like a landspout yet this is a nice trunk. Where is the wall cloud and dry slot? Any speculation on it's formation and whether the parent cell was a supercell?

http://www.stormchasers.au.com/Leopoldtornado0018.htm

http://www.stormchasers.au.com/oct01.htm

Bill Hark
http://www.harkphoto.com