See : 128km Radar Loop for Mildura, 05:00 12/07/2012 to 16:00 12/07/2012 UTC
And the 256km loop which shows my target area clearly:
See : 256km Radar Loop for Mildura, 05:00 12/07/2012 to 16:00 12/07/2012 UTC
So its pretty obvious when we look at it that the atmosphere didn't pan out quite like I thought it might. Or did it? You might notice the cell that forms near Robinvale NSW at around 6:30GMT (4:30 pm), which initially moves south but then turns hard east around 7:30pm local before anvil debris from a cell to the north interrupts its lifecycle (note that Mildura is an under-reading C band radar, and is notorious for showing far weaker echoes than the storms actually occuring, so red/orange is quite impressive). The deviation to the mean flow (some 45 degrees) and longevity (at least 2 hours as a discrete cell) would suggest that this was a likely left moving supercell. In fact a number of the cells show the tendency to turn to the left of the flow in the northern sector, until a pesky right split from one of the three initial cells kills the cell heading for Mildura and causes a multicellular collapse, and a second right split from the northeast cell kills the middle of the three. So looking back on it actually the forecast was pretty good - we were looking for organised, rotating storms and found them, but unfortunately without enough capping we ended up with too many storms which had some interaction issues.
So why were the storms forming in these two areas, here is the 00Z forecast from GFS for 5pm.
Hang on you might say, that looks very different to the forecast this morning - You are right, the system has retrograded further west into South Australia. A quick look at the CAPE at 00Z:
So the Instability is still in Victoria, despite the shift in the upper system, but shifting the upper system might have changed the shear - so what if we have a look at 06Z: a 4pm close to analysis:
The instability is smaller than forecast and shifted northwest, so less energy is available for the updrafts, perhaps explaining the tendency to collapse as the cells moved away from the favourable CAPE towards the southeast.
Ahh, we have moderate unidirectional shearing of the atmosphere, with relatively higher bases (taller LCL) and more instability. So what was this going to do to the storms? Cause splitting as both the left and right rotational couplet of the supercells are equally favoured by this sort of wind regime. When you end up with cell interaction you never know what you will get (a monster or a collapse) so in this case the interactions killed the cells, and with balance not quite right any attempts at being a supercell were pretty transient. So in conclusion, the forecast based on the data available was a good one, (20/13 was seen in Mildura) but the evolution in the atmosphere wasn't quite right to produce the strong storms we were looking for, so I'm a little glad I couldn't chase but at the same time disappointed as its always interesting to see what evolves in different environments. What is interesting is the storms in this area weren't the most impressive of the day. In fact check out this sounding east of Adelaide:
Boring and capped you might say? Well, it has reasonable turning and ok deep layer shear of around 45 knots, but why am I mentioning it as it has little CAPE and looks like it has no chance of producing storms. If you look at this radar loop though:
Whoah, suddenly we have strong storms, and supercells forming amongst them. But how? The environment wasn't really primed for that, or was it? I draw your attention to the area between Maitland and Port Lincoln at 12pm, and subsequent scans - which shows a gustfront boundary with preferential lifting along its axis, and though not visible on the radar scans (probably due to scan height), it extends to the area east of Adelaide. So what has happened because of that boundary? Clearly cells have been initiated by the additional lift, but additionally the Storm Relative winds have been increased - in effect the forecast shear looks nothing like what the storms saw on the day, and the resulting development was favourable for organised rotating storms and where things were too clustered, multicells. The lesson here is: forecasting for a chase will only get you so far, what the atmosphere realises can be worlds apart from that forecast that you thought was perfect. It is really important though to analyse why your forecast didn't (or did) pan out as pattern recognition improves your forecasting markedly (though not every day will pan out the same with the same conditions) Adapting to mesoscale features (such as boundaries) on what might appear to be marginal days will definitely make you a better chaser, so keep your eye on the observations and the skies.
Oh and was my forecast a bust? Not sure which way to call this one, while storms formed and organized as I expected it would have been a long drive for just that Robinvale cell - but it could have been really spectacular and worth it. You be the judge, tell us what you think on facebook or in the comments below.