Thunderstorm through Woomera, SA tracking across the Eyre Peninsula toward Adelaide in 2016. Photo: Bureau of Meteorology SA
Background
On September 28, 2016 one of the most significant thunderstorm outbreaks recorded in South Australia impacted central and eastern parts of the state. Multiple supercell thunderstorms were embedded in a Quasi-Linear Convective System aligned with a strong cold front that was associated with an intense low-pressure system. The storms produced at least seven tornadoes, destructive wind gusts, large hail and intense rainfall. Transmission lines were brought down in four different locations, which contributed to a state-wide power outage.
A simulated study
Accurate prediction and understanding of tornadoes and other hazards associated with severe thunderstorms is very important, for timely preparation and announcement of warnings. By conducting high-resolution simulations, this study by the Bureau of Meteorology’s Dragana Zovko-Rajak, Kevin Tory and Jeff Kepert, aimed to offer a better understanding of the meteorology of the South Australian thunderstorm and tornados.
It also contributes to improving knowledge of how to best predict similar severe weather events, which in turn enables better risk management and preparedness for such events.
Updraft helicity, a severe storm surrogate that indicates the potential for updraft rotation in simulated storms, was used to investigate the ability of the model to predict the likelihood of a supercell and a tornado.
Summary
The ensemble simulations highlighted the uncertainty associated with timing, location and intensity of the convective systems that spawned the tornadoes. While each indicated some potential for tornado formation, this potential varied in magnitude, timing and location.
It may not be appropriate to assign a numerical probability to tornado formation in this case, since the ensemble has not been calibrated for this purpose. Nevertheless, such an ensemble would have strongly supported a forecast of a high risk of tornado formation.
This ensemble simulation presents a good example of an important advantage of ensembles over deterministic simulations. The control member of the ensemble provided a more modest indication of tornadoes, especially to the north of the region of interest, as did two of the others. If any of these had been the sole deterministic member available, the forecasters would have not received as strong an indication of the potential for tornadoes. A single deterministic forecast can be regarded as a random choice from the set of all possible ensemble members. Although in this case the deterministic forecast happened to strongly indicate tornado risk, it could well have been weaker.
The most extreme forms of severe weather are, almost by definition, rare events. Thus, the use of an ensemble reduces the chances the numerical guidance will unluckily miss the true magnitude of the event.
Malcolm Jackman, South Australian Fire and Emergency Services Commission
Jillian Edwards, National Resilience Taskforce, Department of Home Affairs
Prof Alan March, University of Melbourne and Bushfire and Natural Hazards CRC Project Leader
Frank Crisci, SA Power Networks
Peta O’Donohue, Country Fire Service, South Australia
Using this September 2016 extreme weather event as a real-world scenario, the panelists explored the real cost of disasters, efforts to build disaster resilient communities, and plans to reduce disaster risk.
The forum was run in partnership with the South Australian Fire and Emergency Services Commission and the Australian Institute for Disaster Resilience, with support from the Department of Home Affairs and SA Water. Read more here.