 We're looking at some rainfall up in board areas over districts, right down to specific rainwater quantities in catchments and in valleys. Not only that, we're doing something what we call ensemble modelling as well. Running the same model over the same area, maybe 50 or 100 times, slightly different inputs just to see how much variability there is in the atmosphere or the inputs that result in different rainfall. Also, that gives us a really useful information when we talk about projects. Most of the watch art can be seen here through the media. Underlying that, there is usually a whole range of uncertainty and sometimes the forecast can vary out of certain. It sounds highly certain. So being able to communicate that using the information coming out of this recent computer will be another game changer. Just a little bit about it. You can see on the board here, if you're into numbers and things, teraflops, we've moved on from kilobytes and megabytes and gigabytes. We're now talking about teraflops now. But in broad terms, the Bureau runs a supercomputer at the moment and the new one will be 16 times faster. When that gets upgraded again in a couple of years' time, it will be as much as 50 times faster. So it has got the capability to be upgraded. You can see on the right-hand side there a number of the benefits of this will be obviously more accurate forecasts. We've been moving down this path for some years now in terms of the Bureau's forecast. Obviously the weather's a chaotic beast. So it's always going to throw us sort of spin balls here and there. But overall, the forecasting capability of the Bureau is improving largely because of supercomputer power. Also our understanding of the weather is improving through research, but it really hinges on this supercomputing power to be able to get more accurate and more dynamic and more localised weather forecasting. Okay, I want to talk briefly. This is the only slide I really have on fire. But there's been a huge change in the way we understand fire. This has taken place. It's too pronged really, not only seeing the sort of fire behaviour we're getting but being able to collect that data and the research is being able to analyse it and using supercomputers that model the atmosphere at high resolution. We're able to understand some of the physical and dynamical properties that give rise to sometimes what we call unexpected fire behaviour. But we're now moving from what we previously have called unexpected fire behaviour to becoming more predictable fire behaviour. This is an excellent example that took place in the Grampians this year. We're leading up to this event that was fire in the landscape. The day before the fire really took off and what we call became a plume dominated fire. Fire behaviour analysts were ahead of the game. They knew that this was a high probability of occurring so every strategic and tactical decision was based on the fact that the fire behaviour ultimately was going to look like what you see on the screen there which means very different tactical operations on the ground than any more normal behaviour that might be more wind driven or terrain driven. So there's been a lot of research done and we've really switched on now to the situations where this can occur and the incident control centres have got some excellent resources, some excellent information at their fingertips now to highlight the possibility of this occurring rather than being surprised which was probably what used to be the case in the past. So just a bit of information there of what we call pyrocumulus if you're not familiar with that term. Most fires don't develop their own clouds but when they do they become highly dynamic and that manifests itself on the ground as quite intense fire behaviour and in the most extreme cases these clouds can reach right up to the top of our tropper pours and it seems ironic but a fire which is very intense at the surface can ultimately end up producing its own ice crystals at the top of the atmosphere in the most extreme instances. And we use something now called the continuous Haines Index a bit like the fire danger index, it's meant to complement it but gives us an idea of whether a fire can end up generating this type of fire behaviour. That's on the fire with a forecast and available through a few other mechanisms as well. Talking a little bit about heavy rain leading to flood events as a meteorologist we often look at or use pattern recognition in combination with information from our dynamic weather models but there are three patterns typically it boils down to that we're looking for that generate our most heaviest rain events and we've come a long way since events like Benella which really surprised us in terms of the quantity of rainfall it's pretty rare now for us to sort of be surprised that there will be a heavy rainfall at least we might not get the quantities always right but predicting heavy rainfall events now are becoming much easier thanks to numerical weather modelling but the one you see here is what we call the east coast low if you were to look at a synoptic weather pattern 24 hours prior to that you would have no idea that a low could deepen so quickly on the eastern seaboard and now our numerical models can do that whereas previously they weren't able to capture that depth of dynamic rapid deepening of these low pressure systems so obviously this type of system because it deepens so rapidly and because it is often influenced by enhanced moisture off the Tasman Sea is responsible for heavy rainfall particularly in Gippsland but also can be responsible for heavy rainfall in Melbourne and also in the Otways interestingly it's this type of event that has been responsible for Victoria's heaviest ever rainfall recorded and that was in the Otways a number of decades ago and it was due to an east coast low like this the second heaviest rainfall event occurred in Victoria occurred in Wilson's Promontory just a few years ago not quite this type of event but very similar and both those rainfall, daily rainfall totals were in excess of sort of 350 to 400 millimetres now the other one which some of you will be familiar with this occurred a few years ago back in 2011 the flooding across the northern plains but also the sort of pattern that was responsible to some extent for the vanilla floods back in 93 where we see a low pressure system it doesn't look all that low it's only about a thousand hectopascals compared to the lows that we can get off the Southern Ocean and Tasman Sea but the issue is it stays there for a long time and it's helped by that high pressure system that you can see cradling that low to the south if a low stays in one position for a long time over multiple days and it's linked to tropical moisture it's just a very unique and high probable outcome of getting heavy rainfall over a wide area and that's what happened back in 2003 so it's that slow moving low that doesn't move anywhere due to that quite strong high moving southwards and it's linked to tropical moisture so it usually only occurs during the January sort of February timeframe sometimes into March you can see there we've got a tropical cyclone off the northwest coast and the last system again this is fairly recent only last year where we had a quite intense low near 980 hectopascals coming out of the Southern Ocean not linked to tropical moisture but a very intense low low pressure systems are responsible for air rising in the atmosphere and you need air to rise to generate rainfall pretty much and the lower the pressure is the greater that lifting mechanism is and although this was moving through quite quickly it's the rapid deepening of the low, the intensity of the low with moisture in the atmosphere that results in the heavy rain so there are the three systems and as a meteorologist when I communicate why do we often get heavy rainfall out of the atmosphere and it's these lows that act as the trigger and if you think of the atmosphere like a sponge sponge can hold a lot of moisture and our atmosphere can actually hold a lot of moisture you might not get any rain out of it but there's a lot of moisture out there you need something to trigger it you need what we call a something to squeeze that sponge squeeze the moisture out of the atmosphere and the low pressure systems are one very strong mechanism that results in that squeezing process so that the rapid deepening of the lows the intensities of the lows of what ultimately responds it ultimately results in that squeezing of the moisture and fortunately now with Unumerical Models they give us a really good idea of how much moisture can be squeezed out of the atmosphere whereas up until probably a couple of decades ago a lot of that process of working out how much rainfall was based on experience and knowledge of past events just very briefly just got a few slides here about severe thunderstorms because they are a very different beast compared to trying to forecast heavy rainfall events or wind events when we're talking about severe thunderstorms we're talking about a few different phenomena associated with them obviously large hail not like the thunderstorms we get in winter which obviously have usually quite small hail it's the extra moisture in summer that generates the larger hail the wind gusts of 90 kilometres an hour or more often very quick very rapid and short lived heavy rainfall usually quite localised and of an intensity over a short period of time that results in flash flooding and the last one that defines a severe thunderstorm is a tornado which is very difficult to forecast for usually because they're quite short lived and require very specific conditions under the thunderstorm for them to develop so in terms of the life cycle of a thunderstorm whereas a low pressure system you can often see tracking across the southern ocean or coming down the east coast a thunderstorm usually is quite different it's not like you can see the storms in western Australia they're travelling across south Australia and they make their way to Victoria sometimes that occurs but usually the life cycle of a thunderstorm is only one hour so we've got to wait for them to develop to see where they're developing and then issue the warning and if we look at where the best practice of warning services severe thunderstorms occur and that's in the United States they usually aim for about a 30 minute warning if they can get that they're pretty happy anything longer than that is usually a bonus and they've got highly specialised infrastructure in place to be able to analyse a lot of the storms that develop in their atmospheres and we're not too far behind but we just don't have the extensive nature of the infrastructure but I write our network nevertheless pretty good particularly in Victoria so in terms of thunderstorm we get this what we call the development stage and you'll have seen this out in the land where we see rapid developing of a cloud and then it gets to a tipping point where the what we call the updraft collapses and all that rain that was the spill precipitation that was suspended aloft ends up falling down to the ground and usually that is only for about 20 minutes sometimes 30 minutes it's all over in a very short timeframe and ultimately it's that outflow that you can see there on the right that ends up killing the updraft of a storm and then dissipates in fact what you see here that outflow from the storm can actually then be the next trigger to develop another thunderstorm so when you see a line of storms moving across the state along that line usually the storms will only last an hour but each storm may then trigger another one from developing so it might actually look like one or more storms lasting for a long time but in case what's happening is this redevelopment of thunderstorm activity along the line you can get severe weather from these types of storms but usually 90% of these types of storms won't generate severe weather in summer some of them will be severe but the ones we perhaps more we can competently warn for is this next type which we call the supercell thunderstorm which has a different from the ones you saw before these ones have what we call a rotating cloud system you may have seen this in the environment where the middle of the storm is actually rotating it might be masked by rain so you might not actually see it and it's that rotating process just like you see the water going down the plug hole and the bath or the tub once you get it set up it can often last for a long time sometimes to the order of several hours which is why they can be self-sustaining and continue to sort of generate the severe weather as they cross the landscape these are the ones that our radar system can pick up quite well particularly if you're looking at the Doppler radar and they're the storms that will propagate for quite a period of time so often when you see these storms on the radar and they will last for some time the Bureau's got a greater capacity to then warn for them just getting back to communicating this type of these severe weather events whether we're talking about fire, floods or storms I thought I'd show an example this is a blast from the past this was the forecast issue on Ash Wednesday have a little read of that gotta love the text this is as good as it got this was the weather forecast for the day along with this there were probably 20 town forecasts as well and that was it this is 5am in the morning on the day of Ash Wednesday let me note the comments light winds and yeah the extreme fire danger just in the melly I think we all know what happened that day so that's back in 1983 and that was the that's the high profile weather forecast this went to the media through the radio there was no internet then remember so you got your weather via the radio paper and in fact the forecast the day before was pretty similar it hadn't changed between the 6pm or 5pm forecast the day before and 5am it was updated at 11am to produce to highlight extreme fire danger across the state but it was 11am let's contrast that to Black Saturday ok so the conditions for Black Saturday were acknowledged several days in advance the weather was going to be bad this is 4 days look at the terminology used absolute extreme fire weather spike day this is the sort of information now being presented at the state control centre to the emergency managers you sort of get the feeling that we've come a long way in 20 or so years not only being able to forecast these bad days but also use the terminology that galvanises the emergency services and the community in recognising the threat or preparing for the threat obviously when the threat emerges it's often hang on tight but at least from a preparedness point of view we've come a long way and similar to things like the flooding that occurred in Benella where there was only mention of some rain overnight not the intensities we're moving into the landscape now of talking about probabilities of different outcomes and just to highlight that you may have seen this particularly if you've been in the emergency services this is a briefing product now that we work with the SES I'm trying to highlight these days in advance as best we can and also when we get to the day itself or hopefully within 24 hours trying to identify time frames locations and amounts from a preparedness point of view obviously situational awareness will gazump this on the day because that's highly important but at least from a preparedness point of view we've come a long way to get to the full wind as well just sort of highlighting those really top end wind events how long they will last for and the types of gusts that we could potentially see and just lastly if you haven't been onto the Bureau's website for a while there's a link to the right of the satellite image to a new platform called Metai where you can get some very specific weather forecast information it's used quite a lot by the fire agencies now doing their plan burns and managing fires but obviously has a great planning tool for the public and for the emergency services as well where you can get a forecast anywhere in the state so pretty much that's all I wanted to touch on as part of this presentation the last image there is from the new Himawari satellite that the Japanese have launched it's going to be a huge bonus for Australia the Bureau of Meteorology and Emergency Services providing 10-minute high resolution satellite data from a number of different channels that will really keep our situational awareness up to date so that's pretty much all I have so thank you very much and I might open it up to some questions if you have any very difficult the question was how do you distinguish between just a normal severe thunderstorm warning and one that is called a super severe thunderstorm the best way is when you're seeing those severe warnings and if there's a tag on it that says this is a dangerous thunderstorm which is a tag that's occasionally used on the product then you know it's the latter which is a super-cell severe thunderstorm that could last for quite a few hours and produce damaging weather over a long distance the question was is there much moisture in a pyro-accumulus excellent question but there's one thing that the researchers have they haven't been able to nail exactly how much but we know the atmosphere has moisture and the process of combustion of fire releases moisture obviously the more fuel that's consumed by the fire, the heavier the fuel the more moisture that will be released so that in turn gives it a greater possibility of generating pyro-accumulus so crown fires are the ones that tend to generate the pyro-accumulus so the more fuel consumed means usually the more intense the fire behaviour will end up being sort of ironic you're thinking it's a lucky moisture but it ends up meaning you end up with a more intense fire behaviour so the question was about tropical moisture usually in our monsoon which is the summer months we see that dip into northern Australia and some weather patterns allow that moisture in the upper atmosphere to basically advect, advect just means move from one location to another, advect from that part of the world to Victoria meteorologists spend a lot of time analysing the upper atmosphere looking at where that moisture is going and sometimes it's quite dynamic, it can happen very quickly the key is you can have all that moisture coming towards Victoria but you still need what we call a trigger to squeeze it that's it I go in 5 minute warning but you just wave back at me no I saw the 5 minute warning and the wave was the acknowledgement so we're finished there now I'll hang around for 10 minutes or so if anyone's got any further questions but yeah have a great end of the day