A PERSISTENT complaint from victims of the Black Saturday bushfires in Victoria was that they had “received no warning”. Over and again we heard statements like this: “There was no fire anywhere, but the next thing, we had fire all around us. There was no word of warning, and we never stood a chance”.
This issue has since been highlighted by the Royal Commission in its Interim Report, and is being taken to heart by fire authorities all over Australia. In Western Australia, for example, the Fire and Emergency Service (FESA) has rolled out a new warning arrangement based on mobile phones, and has carried out a substantial and well-publicised test in a Perth Hills suburb. It was said (by FESA) to have been a great success.
This is a delicate subject, because I don’t want to sound disrespectful to people who lost their lives or suffered in the Victorian fires. I realise that many people are perplexed by the way they were engulfed by fire and caught by surprise. I understand the desire of authorities to get warning systems in place. Officials realise that a failure to deal with this issue in future fires will come back to haunt them if complaints are made to Royal Commissions, Coronial inquiries and the media.
However, the downsides, weaknesses and dangers in bushfire warning systems must be properly understood.
The first problem is that while the behaviour of bushfires burning at low intensity in light fuels is well understood, high intensity fires in heavy fuels can behave erratically. Intense fires generate their own wind and throw spot fires kilometres ahead. This is the main reason people are caught unawares. One minute the fire may well be “miles away”. But the next minute a high wind brings a rain of burning embers. If these fall into heavy dry fuels, people rapidly find themselves enveloped by fire. High intensity fires will leapfrog across one ridge to another, and then swirl back, sucked into the intervening valleys and seemingly coming from the “wrong direction”. A bushfire can move from a mild ground fire with 1-2 metre flames, to an intense crown fire (throwing spotfires) within a matter of minutes…. it is simply a matter of a wind change turning a long flank into a headfire, or of a fire moving from an area of light to an area of heavy fuels.
Very rapid changes in fire behaviour, and mass spotfire generation present a nightmare for people with the job of activating a warning system. Decisions can only be made with very accurate and up-to-date information from the fire. Since the situation at the fire is often confused, and firefighters generally do not have any idea of the big picture, it makes decision-making about whether or when to activate a warning (and to whom) doubly difficult.
A further problem is that rarely do you get one fire at the one time, especially on a bad day. When there is a dry lightning storm, of where an arsonist is at his dirty work, it is not uncommon for several fires to start at about the same time and run parallel with each other. This can confuse efforts at fire detection, mapping, and spread prediction. When many separate fires start to coalesce and interact, fire prediction moves into the realms of the unknown, making it virtually impossible to know who to warn and when, other than in the broadest geographic sense.
Finally, any warning system based on communications technology is likely to break down in a serious bushfire situation. This is especially true of technologies that require mains electricity, which is generally the first to go when there is a fire, or static relay stations like phone towers that can be destroyed by fire or cyclonic winds. To this must be added the well-known problem of communications overload in a crisis situation.
There are two serious dangers with the whole concept of targeted warning systems. The first is that a mass warning will quite possibly lead to a mass evacution. People leaving the area will choke the roads, and these may well be the same roads on which there are incoming fire appliances. It is not clear to me that the authorities have sufficiently thought this issue through.
The second danger is that the authorities are raising expectations that they may not be able to fulfil. If people are expecting to get, and are waiting for a warning, and the warning does not arrive (for one reason or another), they are going to be set-up for calamity. I hate the idea of community and individual self-reliance being undermined.
To be effective and reliable, a bushfire warning system must meet a number of criteria. It must have access to accurate data on fire location, fuels and weather, together with the fire behaviour algorithms that can predict fire frontal development. It must be able to anticipate wind changes and instantly take on board new information from a fire where long-distance spotting is occurring. It must be flexible in responding to rapidly changing human as well as bushfire situations. There must be back-up in the event of a technological failure. Above all it must have a large and well-trained human resource to make everything work under extreme pressure, including very experienced and accountable decision-makers. A system meeting these requirements will be expensive to set up and maintain. It will also suffer steady degrade if a few years go by with no major fires.
It is will be the height of over-confidence to create an expectation in fire-prone communities that they will always receive timely warnings of imminent bushfires. The system will probably work under relatively mild weather and low fuel conditions. But the opposite will always be more likely when a killer bushfire is running. Then people will receive no warning, or warnings will be too late to enable appropriate actions.
There is another very real problem. This is when warnings are issued but are not followed by a fire. In the coming fire season or two we can expect that there will be a (wholly understandable) temptation to overdo the warnings. Fire officers with trigger fingers will not want to face a Coronial Court for failing to push the button. But if fires do not follow warnings, the result will be the “crying wolf syndrome” where people become blase, and then do not react when there really is a fire.
In my view the first priority for fire authorities should be to optimise the bushfire resilience of towns and communities – in particular reducing areas of heavy fuel within and adjoining residential areas, making houses and road verges safer, setting up local community refuge areas and maintaining a program of regular mild burning in hinterland forests. Secondly, they should be telling people that it is quite likely they will NOT be warned and that they must themselves take responsibility for finding out what is going on and having a sensible plan of action, including evacuation to a safe place well before a situation becomes remotely dangerous. In my view both of these actions will have greater value than spending millions of dollars on “technological-fix” bushfire warning systems.
The fundamental message that our governments should be putting out is this: if you live in, or close to the Australian bush, you should expect to get a bushfire on a hot windy day in summer….. and be prepared for it. To rely on a government warning system is to rely on something that is inherently unreliable.
There is a final factor. As that wise anthropologist George Silberbauer has pointed out, we already have a system in which the Bureau of Meteorology puts out twice-daily fire danger forecasts and these are published on the net and broadcast on the news. Most country roads have Fire Danger warning signs. The problem is, few people understand fire danger, and the system is unduly complicated with six, and soon to be seven, categories. It is possible that if we had a more simple way of expressing the fire danger index, which is a warning in itself, and we ensured it was more effectively transmitted and better understood by the whole community, the new technological gizmos would not be needed.
Other articles by Roger Underwood: http://jennifermarohasy.com/blog/author/roger-underwood/