CHAPTER FIVE

WESTERLY

‘And then the wind changed.’ This sentence recurs in almost every account by those in the valley that evening. I began my research with a simple narrative of the fire, and for almost two years I understood this to mean ‘and then the wind changed from north-westerly to south-westerly’. For months I read the transcripts of the interviews I had conducted in the light of what I thought I knew, only to become more and more confused. The Bureau of Meteorology had forecast the wind change — it had faxed prediction maps to CFA headquarters — but no one knew exactly when the change might arrive. Sometime late in the afternoon of 7 February, a CFA staff officer at the Kangaroo Ground Incident Control Centre had written on a scribble pad — on which he or she jotted reports as they arrived, tracking the fire front — ‘Wind change when?’ and beside it added, ‘2030–2100’ (8.30 to 9.00 p.m.). No one really knew, and when it arrived it was not necessarily from the direction expected.

Every historical narrative that is based in part upon the accounts of participants is the result of a three-cornered connection: between the participants, the events through which they lived, and the historian seeking to make sense of the story. From the very outset of research for this book, I acted on the conviction that those who experienced the fire had the best understanding of it. I realised that the CFA records which I had consulted at its regional headquarters — although testifying to immense resources and heroic efforts directed toward fighting the fire — did not help much to establish what happened in the valley of Steels Creek on the evening of 7 February 2009. This meant that, even if they were able to see only what had occurred on their properties, those who survived the fire in the valley could at least describe what they had seen and felt. Stitching enough of these accounts together would surely bring us closer to establishing what had happened, where and when. Early in the project — on the first day I spent in Steels Creek, in fact — I heard and accepted a straightforward account of the way the fire developed that I found satisfactory. The simple narrative went like this: the fire travelled along the rearward, western slopes of Yarra Ridge, driven south-eastwards by a powerful north-westerly wind. Then the wind changed and the fire was driven north-eastwards up the valley, broadening into a wide front as it went. Those two forces seemed to account for the destruction that the fire brought to Steels Creek and its people.

For about the first year of my research this simple narrative seemed to suffice. As I met and listened to the people I interviewed, their accounts seemed to fit into it, to corroborate and confirm it. The fire swept to the south-east and consumed Yarra Ridge; then it turned as the wind changed, roaring up the valley and away towards Toolangi. Increasingly, though, I began to realise that the simple narrative was unsatisfactory.

The simple narrative did not explain how the fire behaved in relation to particular places. Individual accounts had fire fronts arriving from all directions. Part of this problem was one of time: hardly anyone knew exactly when anything had happened, and as I learned more about fire behaviour I found that the very topography — allied with vagaries of wind, fuel, and slope — made the fire behave in erratic and utterly unpredictable ways. In the second year of my research, I began to incorporate these vagaries into my account. By this time, I began to reach some dispiriting conclusions. However many people I spoke to, I felt that I could never comprehend or describe what had happened. The more I went over survivors’ accounts, the more pieces of evidence did not fit what I had believed to be a general pattern. Re-reading my notes, I realised that this scenario did not reflect the evidence I had collected. Instead of finding that I had done justice to what people had told me, I saw inconsistencies and anomalies.

Again and again in interviewing those who saw and often fought the fire, I would ask questions predicated on the understanding that the fire basically reached Steels Creek properties from the north-west (before the change) or the south-west (after it). Again and again, though, survivors’ recollections did not square with that assumption. At first, I assumed that such anomalies were the result of simple misunderstandings, hazy recollections, and careless gestures. Sometimes the transcript reveals that confusion existed. Here are Joe and Pat Maurovic trying to get me to understand where the fire that struck their place at the head of the valley came from:

PS: Now, can I ask you, the fire that you faced … originally it came from over there? [points to the south-west]

Pat: No, it didn’t.

PS: Well, from down the ridge …

Joe: No, no, it came up to us.

PS: Yeah, but …

Pat: But before it comes up to us — [points to the west]

PS: Before it gets from here to there, it came down through the national park?

Joe: See, it’s from here. [points at an aerial photograph] From the west.

‘From the west’. As long as I understood the fire as a simple north-west wind/south-west wind pattern, this memory remained awkward to explain. Perhaps it was a mistake, but I began to see so many ‘mistakes’ that I began to doubt my ability ever to work out what actually occurred. I had bought into this project on the assumption that, though chaotic, a fire was as explicable as a battle. Perhaps I had been wrong about that. Could such a whirlwind of an event, such a complex phenomenon, ever be presented as a coherent narrative? For a time, such doubts all but paralysed me, though I smiled and spoke optimistically anyway.

Plotting the general course of the fire on a map, taking in more than just Steels Creek, deepened rather than resolved the mystery for me. The fire had consumed Strathewen, twelve kilometres to the west, at around 4.40 p.m. How had it reached Steels Creek only twenty or so minutes later? What drove it eastwards — almost perpendicular to its course — when a general change did not arrive for a further hour at the earliest? The fire certainly burned out Strathewen. But no place in Steels Creek was directly in the path of a fire following a south-eastward course from the places that burned around four. Something happened to the wind on the late afternoon of Black Saturday, something that no survivors’ evidence showed explicitly, but which many survivors’ accounts support.

Just when I was feeling that I could not give to the people of Steels Creek the clear narrative they craved, a piece of serendipity occurred (as it so often has in research as I have known it). It took the form of Kevin Tolhurst’s fire maps. As we have seen, on the afternoon of 7 February Tolhurst was in a corridor at the Integrated Emergency Coordination Centre, collating and mapping reports of fires, wind directions and speeds, and other variables. He used these to produce a series of predictions that appear in the CFA records but which, as the 2009 Victorian Bushfires Royal Commission disclosed, did not reach those in charge of responding to the unfolding disaster.

Though disregarded on the day by the CFA’s Chief Officer, Kevin Tolhurst’s maps have assumed a huge importance as historical documents. In addition to fires proper, the maps show a plume of tiny red spots flung out ahead of a fire. This is ‘ember fall’ — fragments of burning leaves, branches, and forest litter, plucked by the wind and carried far ahead of the main fire. Those who endured it tend to call it ‘ember attack’, because it feels as if the rain of burning litter has been directed. Journalist Gary Hughes, who, with his family, narrowly escaped from his burning house in nearby St Andrews, described it in The Australian as feeling like ‘a fiery hailstorm from hell driving relentlessly at you. The wind and driving embers explore, like claws of a predator …’ The royal commission heard evidence that spotting occurred at distances of up to twenty or thirty kilometres. While many embers cooled as they rose and fell, others remained able to start fresh blazes, fanned by a fierce wind and falling into highly combustible country. To those facing the phenomenon, ember fall seems malevolent and deliberate, like an air raid. In reality, it is driven by the force of the wind, and it is wind that provides the hitherto hidden explanation of how fire reached Steels Creek when it did.

For months after the fire, Dr Tolhurst continued to gather, collate, and analyse the physical evidence from what the CFA calls the ‘fireground’, eventually creating a detailed animated map of the fire’s growth and progress, which enabled him to view its course almost minute by minute. The interactive map shows the plume of the fire’s ember fall and the course of the fire front moving across the landscape in blocks that are only a few hundred metres across. Replaying it, paying particular attention to the area around Steels Creek between about 5.00 and 6.00 p.m., suggests several revealing conclusions.

The first is that the fire did not advance at a uniform speed across the ranges from Wandong. This insight will surprise neither bushfire researchers nor those who faced the fire on the ground. For its first hour or two, its progress was relatively slow. But this relatively slow pace raises the tantalising question: could those in its path have been warned in time? The royal commission, especially in its 2009 interim report, asked the same question.

The second conclusion suggested by Kevin Tolhurst’s fire plotting is new, and of particular relevance to Steels Creek. It shows that to the north-west of Steels Creek a heavy ember fall stretched out ahead of the relatively narrow fire front in a vivid red plume twenty kilometres long. The embers were by this time falling around Yarra Glen and the Yarra Valley. This plume caused the spot fires afflicting the area between Yarra Glen and Healesville — the fires seen in the distance to the south and south-east by the anxious residents of Steels Creek and Yarra Ridge. But move the cursor on a further ten minutes and the plotting shows ember falls reaching the valley of Steels Creek — all of which is north and east of the area that had been affected a few minutes earlier. Red dots, each representing the actual or predicted extent of a spot fire (many traced or reported on the ground by investigation teams in the aftermath of the fire), appear on the screen to the east of the main plume. Played at a slow speed, the sudden appearance of spot fires in Steels Creek shows dramatically that fire first arrived in Steels Creek not from the north-west, where it had traditionally been expected, but from the west.

What seems to have happened around 5.00 p.m. is what we can only describe as a miniature change of wind direction. This explains why, again and again, witnesses reported seeing fire moving eastwards. Because I accepted the broad explanation that the fire came from the north-west, I had not noticed these mentions; then I discounted those I found. As Kevin Tolhurst’s model explained what had probably occurred, I realised that, in fact, witnesses in Steels Creek were corroborating and confirming what Tolhurst had shown using quite independent observations.

Complicated by local shifts in direction due to geography and the colossal forces unleashed within the conflagration — we can never forget how terrifying it would have been to face such a fire — Tolhurst’s model explains the broader pattern of the winds governing the movements of fire in Steels Creek. It explains why the eastern face of Yarra Ridge was alight by around 5.30 p.m. It explains why the fire seemed to observers on the valley floor to move down the ridges from the west; and why those on the crest of Yarra Ridge experienced fire first from the south and west — before the wind shifted again to the north-west, until finally (probably between 5.15 and 5.45 p.m.) it shifted around decisively to blow from the south-west.

The realisation that the fire reached Steels Creek in embers driven and deposited by a westerly wind explains many comments in my interviews with its people. Ivan Filsell described how ‘our fire’ came from a spot fire ahead of the main front. Brian Spears said, ‘It came from a totally different direction to that we [had] prepared [for] and assumed it would come from.’

Margaret McLoughlin remembered that ‘the fire started behind us, over there somewhere’ — pointing to a place where, according to the map (and the accounts I had of it), it should not have. But then she began to describe where she saw embers falling (from a kitchen with a panoramic view south-east over the Yarra Valley). She described how ‘they were coming, they were basically coming …’ and I supplied ‘south-east’ — so imbued was I with the explanation that the fire came from the north-west. I now understand that Margaret was trying to say that they came from somewhere to the south-west. She described the embers flying overhead and then the flames approaching. Her account makes sense in the light of Tolhurst’s map. And looked at in this light, Ivan Filsell’s explanation is clear: ‘the fire’, he wrote, ‘was burning in different places at different times’.

What Kevin Tolhurst’s model (and the evidence on which it is based) gives us is a new understanding of what happened at Steels Creek in the late afternoon of 7 February 2009. As we know, strong north-westerly winds pushed a fire (originating from a fault in an electrical transmission line at Kilmore) south-eastwards. It reached the base of Mount Disappointment at about 3.30 p.m. By about 4.00 p.m., accelerated by ember falls far ahead of it, it had reached around Smiths Gully, seven kilometres west of Valley Road. Had the fire front continued in a south-eastward direction, it would have struck Yarra Glen, as spot fires blown ahead of it did. A westerly wind — possibly generated by the forces within the firestorm developing over the ranges — blew embers eastward around 4.45 p.m., bringing the first ember falls to the valley. By 5.00 p.m., spot fires were burning across the valley floor and fire was advancing over the ranges to the west of the valley. The valley was alight in many places when, certainly by 6.00 p.m., the south-westerly change drove the fire in a north-eastward direction. The hitherto unexplained additional change explains why the valley burned so swiftly: in effect, the fire did a three-point turn over Steels Creek, spreading across the valley and its surrounding hills more or less simultaneously.

The ember fall provides a vital clue to the puzzle, of how fire moved into the valley, that many Steels Creek people have sought to solve since the day of the fire. What Kevin Tolhurst’s expert analysis and modelling cannot do, however, is explain what it was like to experience Black Saturday at Steels Creek from within the blaze. To understand that, we need to turn again to the accounts of those who survived.

Around four in the afternoon of 7 February, Brian Spears took two B&B guests to a wedding at the Immerse winery in Dixons Creek. Jo Spears, who had been watching the plumes of smoke to the north-west, kept consulting various websites. Her sister called from Alexandra, asking whether she knew what was happening. Jo — like so many others that day — reassured her, saying, ‘Don’t worry, it’s in Kilmore.’ When Brian returned, he and Jo walked to the end of their verandah.

Looking at the smoke, Jo said, ‘Oh Brian, we’re in real trouble.’

‘Why’s that?’ Brian asked.

‘Because that’s a firestorm’, she replied.

‘It’s a jet’, Brian said.

‘No. I’ve been in one of those before. It’s a firestorm. We are stuffed.’

Everyone I spoke to remembers the noise. Many described it as indescribable — chilling, frightening. When groping for similes, people talked about jet engines and express trains, ‘only much louder’.

Sometime before five, at the southern end of Steels Creek Road, Michael Boffa heard a roar, coming from over the hills to the west. The Boffas’ house seems to have been struck by the very extremity of the fire front that roared out of the Kinglake National Park. It skipped along Yarra Ridge at a frightening speed, carried by flaming embers pushed far ahead of the main front, crowning in the wind and rocketing out in great tongues of flame. (Michael noticed that the CFA website recorded ‘fire in Steels Creek Road’ at 5.17 p.m. The exact location of this sighting is unknown.) By about six, the hillside north-west of the Boffas’ property was full of clouds of choking brown smoke. As the wind changed, flames reached the house and it filled with smoke. The Boffas sent their two boys into a bedroom, where they lay down with their cat, while Michael and Chris began to hose the flames.

Although John O’Neill had spent the afternoon readying his house in case fire arrived, he had a commitment with a friend from Dixons Creek to stock the bar at the Healesville racecourse for the next day’s race meeting. He and Peter left Old Kinglake Road at about four. As they passed Yarra Glen, they saw grass fires in paddocks beside the road near Train Trak. John called his former neighbours at Tarrawarra to warn them, but he and Peter pressed on to Healesville. As they stacked slabs in the refrigerator, they kept an ear cocked for the radio. It was not so much what the announcer said, but the tone of his voice that alarmed John. Already uneasy, he and Peter now felt ‘really edgy’, and they abandoned unloading the beer to race home. As they sped along, they saw more grass fires. Started by ember fall, these fires were spreading all over the valley. On Gulf Road, there was one moving slowly, with no one attending it. As Peter dropped John off (and drove off to face the fire at his own place in Dixons Creek), John picked up the phone to hear neighbour Dave Allan report that he had seen fire on the ridge line of Mount Everard, less than three kilometres away. It was just after 5.30 p.m.

John mobilised his family. He started the pumps and, with his sons, began soaking their weatherboard house and its wooden deck. Looking to the north-west, John noticed that the sky was darkening: the smoke was thickening, billowing purple behind an angry red glow. The wind rose and soon he heard a noise — growing louder until it seemed like a steam train, a ‘rumbling, roaring deep noise’ — heralding the arrival of the fire front.

The Filsells also took the appearance of smoke around Mount Everard as a sign that it was ‘time to get serious’. Ivan described the drill typical of well-prepared bush householders: ‘fill gutters of shed and cabin, damp down all around … start filling gutters, watering cans, buckets, and all available containers’.

Around this time, the fire front raced down Yarra Ridge and struck the Dahlstroms’ property on Steels Creek Road. Ray and Bronwyn and their adult daughter, Karen, reacted swiftly as they realised that the fire they thought was nowhere near was, in fact, bearing down on them. They changed into heavier clothing, wetted down the house again, and began patrolling the outside to douse embers. As the front approached, they acted as they had been trained and withdrew inside the house. But the fire came from the west or north-west, then swung around from the south, seemingly driving into the carport, from where the house caught fire. Inside the house, they looked up to see the recessed light fittings glowing — flames had lodged in the roof. With extraordinary discipline, they stayed in the burning building for as long as possible, realising that the fire was raging outside. While they could not find torches, Ray pocketed the money he had made at the Healesville market that morning. When they could wait no longer, Ray tentatively opened the front door, and they ran into the garden. For about half an hour, the three stood under a water hose near their pig’s wallow, watching their house burn. Their only possessions now were a car; a mobile phone running out of credit; the dirty, wet clothes they wore; and a pocketful of wet bank notes. Not having packed their car, they lost precious family photographs, all of Ray’s artworks, and Bronwyn’s unfinished book.

At 719 Steels Creek Road, the Filsells realised that the peak of the fire had passed and they decided to move outside:

… and the real work started, everything around us was burning, the garden beds, the retaining walls, the trees in all directions, my neighbours’ houses on both sides, and my farm shed with a lifetime’s collection of tools and equipment … all around was thick smoke and a red glow and heat. All the gravity hoses were still working, so we … worked our way out from the house …

At the far end of Scott Road, on the south-eastern slopes of Yarra Ridge, Barry Richards watched the ember fall in his paddocks. In what seemed like half a minute to him, thirty-odd embers fell on one of his paddocks in a huge circle, ‘like candles on a birthday cake’. Soon the fires joined together and edged closer to the Richardses’ house, while another front gathered momentum as it raced uphill towards their neighbours’ place. The wind changed from north-westerly to south-westerly and — while flames continued to whirl about houses on Scott Road — the wind drove the fire north-eastwards. Flames advanced rapidly toward the cluster of houses around Valley Road. ‘It just kept going’, Margaret McLoughlin said simply — a recollection offered by many Steels Creek people about that evening.

Spurred on by spotting and driven by the change, fires roared up the gullies to the south, heading toward Valley Road and taking everyone by surprise — its residents, too, had expected fire to come from the north. Ian Whitford described how ‘a huge black cloud’ moved over at about 5.15 p.m. ‘Almost immediately we were threatened by a fiery blast from the south’. For three hours, he and Peta struggled against flames on their southern side, using hoses, buckets, and mops. The heat was so fierce that it peeled the render on their house’s mudbricks. Fire destroyed three of the area’s twenty houses — some of whose occupants escaped only narrowly — but the residents’ preparations were mostly as effective as their communication. John Houston modestly claimed that he and his wife, Margaret, ‘did no “fighting” as such’, but relied on what he described as ‘preparation, prevention, and protection’. But he and his fellow residents spent hours suppressing spot fires, and they anxiously watched the thick bush around them burn throughout the night.

The fire soon claimed its first victims in Steels Creek. The deaths of Leigh and Charmian Ahern demonstrate, as no other shocking example could, the power of the Black Saturday fire to confound the most well-prepared residents. Police investigators were able to reconstruct Leigh and Charmian’s last hours with reasonable completeness, using the testimony of those who saw or spoke to them up to only minutes before the fire consumed their concrete-brick house, on a densely forested slope at 171 Old Kinglake Road. The investigators were able to reconstruct the sequence of events that led to the deaths because they not only called upon the testimony of the Aherns’ children (who were able to describe their parents’ preparations), but had an eyewitness in Dave Twentyman, who lived virtually next door.

The Aherns were both equipped and prepared for a bushfire. They had seen at least two before, although their house had never been threatened directly. The police confirmed that Leigh had installed and maintained ‘an elaborate firefighting system’, including two large tanks, holding forty thousand litres of water, both with electric- and petrol-powered pumps, as well as hand pumps. Leigh had also installed a system of mist sprays and jet sprays to wet down the exposed wood of the verandah, the carport, and a nearby bungalow. He regularly tested his pumps and sprinklers, and raked up leaves and litter. No one could have been better prepared, but the very position of the Aherns’ home negated all of this.

That morning, Leigh and Charmian prepared their house against fire. Charmian told her daughter, Chloe, that they knew about the possibility of fire (acting upon the general warnings issued the day before); even before the fire ignited, Leigh asked his brother Gane to monitor the CFA website and radio, and to advise him if anything untoward happened. The Aherns had internet access, but Leigh preferred to be out readying the house in case of fire rather than remaining indoors monitoring websites. Around lunchtime, other members of the Ahern family called, but Leigh assured them that he and Charmian were prepared for what remained a possible threat.

Around 4.00 p.m., Charmian Ahern telephoned her son, Dale. She said that they knew of fires, but that they were ‘a fair distance away’ as far as they knew. The situation changed rapidly later in the afternoon. Around 5.00 p.m., Dave Allan, the area’s fireguard coordinator, called the Aherns. He had seen smoke around Mount Everard. The Aherns could see this smoke, too. Leigh sent a text message to Gane at 5.26 p.m.: ‘Mount Everard not lookin good’. Gane replied, telling his brother to think about evacuating. Leigh said it was too late to leave, but that he was sure he and Charmian would be all right.

The Aherns must have watched anxiously as the fire approached, but they expected to survive — Chloe recalls her mother saying that she ‘had to go and get water ready and would call as soon as she could’. Soon after, not realising how close the fire was, Chloe telephoned to ask her mother about a recipe. Charmian said, ‘The fire’s coming over the ridge.’ A minute later, a friend from Dixons Creek phoned, concerned to see smoke over Yarra Ridge. He heard Charmian answer, obviously distressed. Moments later, Dave Twentyman heard the roar of the approaching fire and saw Charmian run out towards the Aherns’ carport; she shouted, ‘The fire’s coming’, then ran back into the house. Dave was the last person to see Charmian alive. He thought that the fire took perhaps two minutes to reach the house from when she first saw flames. Leigh must have been on the other side of the house, wetting the side facing the fire, but soon after both retreated indoors to the bathroom and laundry. The fire passed over their house and swiftly penetrated, consuming it and killing Leigh and Charmian.

The details must be conjectural, of course. It seems that the fire first approached from the north-west, but then swung about and struck from the south-west, smashing the big living-room windows — the house was totally destroyed on that side. Karen Ireland, a police forensic detective, concluded that the fire had approached the Aherns from the south-west. As Dave Twentyman saw orange flame filling the windows of their house, he went upstairs to investigate a smoke alarm. Through the window facing the Aherns’ house, he saw how the fire had peeled back their roof panels and had blown out the wall on the eastern side — a reminder of both the power and caprice of the forces that fire generates as it swirls about structures. Karen Ireland determined that ‘had the roof stayed attached, the sprinkler system may not have been compromised and the contents in the house would not have been open to direct ember and fire attack’.

Forensic detectives concluded that the steep road, probably blocked by burning or fallen timber, made escape impossible: ‘evacuation after the wind shift would probably have resulted in any vehicle being caught by the fire’. The next morning, concerned for his neighbours, John O’Neill walked up the road and looked around the ruins of the Aherns’ house. There he found their bodies. Charmian and Leigh were well prepared for a bushfire, but not for a fire of the scale they faced. Their daughter, Chloe, wrote that ‘They loved the bush, they loved our home and they loved each other. They knew there were risks but they believed that was worth fighting for.’

Around Hargreaves Road, the telephone tree prompted people to call their neighbours. The Maskiels called the Montells and the Montells called the Williamses, who in turn called their other neighbours. To the Montells, the links worked; they felt a part of a network of neighbourly support, even if that support extended to grim exchanges: ‘It doesn’t look good’, Lindy Montell and Amanda Williams agreed. The tree’s effectiveness depended upon the links in the chain being at home — and the Hargreaves Road tree included the annotations ‘weekends’, ‘occasional weekdays’, and ‘not resident’. But, even if everyone had been present and reachable, the suddenness of the fire’s arrival would have negated whatever careful preparations the best community fireguard groups had made.

Somewhere around 6.15 or 6.30 p.m., a CFA tanker nosed through the smoke into the Boffas’ driveway. It seemed to be using the driveway merely to reverse, but to Michael it was a godsend. The fire was licking into his eaves — the most vulnerable point in the house. With the house seemingly minutes away from being consumed by fire, the crew swiftly and skilfully doused the flames. Then they left, evidently turning right to go south toward Yarra Glen and the fires burning around the town. There are few other reports of firefighting tankers anywhere else in Steels Creek until the next day, although early in the evening another CFA crew checked on Judy Anderson, warned her of an open window, commended her on her firefighting skills, and left.

The Boffas fought the fire — in their sheds and wood stacks, in trees and scrub lining the Melbourne Water aqueduct behind the house — for the next ten to twelve hours. Like dozens of others around the valley, they saved their house. Michael Boffa described how, although he faced fires to the west and north, the south and east remained fairly free of all but spot fires from burning embers. He, too, recorded being helped by neighbours — Scotty from across the road, and friends Richard and Kelly, who brought a water tank over. But for the most part, the Boffas fought the fire with what they had — with hoses, and even water pipes, which melted in the intense radiant heat. They poured vast amounts of water onto the burning sheds, sleepers, and trees. Later, Melbourne Water presented them with a bill for the thirty thousand litres that they had used; something of an irony, Michael thinks, considering that Melbourne Water’s unkempt aqueduct reserve helped to fuel the fire at the rear of his place.

From the vantage point of her kitchen, Margaret McLoughlin saw embers falling after four o’clock, as the ‘Kilmore’ fire gathered strength and swept down from the north-west. The embers were bursting into flame around Yarra Glen and the Yarra Valley. Margaret saw spot fires igniting near Train Trak and TarraWarra — a particular worry to Margaret because she worked as a voluntary guide at the gallery. She called friends and repeatedly dialled triple-O, but could not get through. She eventually reached the CFA headquarters at Lilydale, who were even by then overstretched. Margaret and a few neighbours around Scott Road had a small, informal telephone tree, and she and they exchanged worried conversations as the afternoon wore on. Helen Richards, higher up the ridge and closer to the Valley Road telephone tree (of which she was a member), acted as an informal link to the McLoughlins in Scott Road.

Around 5.20 p.m., Richard McLoughlin — who had been keeping a wary eye out, especially on the nearby ridge — called to Margaret, ‘Come and look at the hill!’ They saw flames roaring down the flanks of Yarra Ridge, blowing pretty much straight toward their property. Hastily donning heavier clothes, Margaret and Richard saw the whole of Skyline Road aflame as the fire raced towards them, seemingly heading straight for Yarra Glen. They spent the following hours beating and hosing flames that were licking at their garden retaining walls.

Those familiar with the environment of Steels Creek had a hard-enough time coping with the malevolence of the bush. Those to whom the bush was unfamiliar had an even harder time. Jo Spears remembers one of her guests, a Scotswoman, speaking on the telephone to her family in Scotland, saying, ‘We’re on fire, we’re going to die’. Jo describes the woman as ‘unbelievably scared’ — and with justice. She felt responsible for her husband, who had been incapacitated by a stroke.

On Yarra Ridge, Allan and Rob Giffard had prepared the hoses and pump for a blaze that they were only now able to see, with ember-fuelled fires ‘popping up’ all over the hillside. ‘And then all of a sudden,’ Allan recalled, ‘an ember hit the paddock over there … and it exploded … like a cluster bomb’, scattering burning fragments all over the dry grass. He heard the ‘roaring noise’ that so many bushfire survivors remember. Allan said to Rob, ‘Mate, we’ve got to get out of here or we’re going to die.’ Rob replied, ‘Mate, if you get out of here now, we will die.’ Imbued with the habit of command, Rob directed Allan to protect one side of the house while he defended the other. ‘Right’, he said. ‘We’ve got to stay and defend your ground. Don’t chase it. Let it come to you.’ Soon more embers landed on the paddock before them — the fire came to them.

Those living on Old Kinglake Road and Brennan Avenue felt the most extreme force of the fire front as it roared out of the national park in the wake of the afternoon’s ember fall. Survivors told stories of miraculous escapes. Dave Twentyman described how he laid down inside his house with his two dogs, listening to the roar of the flames and the sound of his eight-millimetre windows shaking and splintering as the pressure of the fire battered them. As the ordeal continued, Dave thought he was ‘stuffed’; but as the noise diminished, he realised, ‘I’ve got through.’ He had survived, he told journalist Jonathan Dart, because he was ‘70 per cent prepared’; but ‘the other 30 per cent was just luck’. Waiting awhile, he emerged with a wetted cotton shirt around his face and immediately started the pump to hose down burning timber around the house, decking, and stairs. Then, after texting Dave Allan (who was sheltering at the De Bortoli winery), he raced uphill to the Aherns’, where the house was already a smouldering ruin. He called out, but got no reply. Dave attributed his survival to the layers of retardant rendering on the weatherboard on his house’s walls, and to the toughened glass that protected the windows from imploding.

Although understandably preoccupied with saving his own house, Hannah Sky’s neighbour John O’Neill was able to describe how the fire struck at Old Kinglake Road. He thought that three fires struck Jehosaphat Gully, one from the north, and then one coming down the creek from the direction of Mount Everard — the front that killed the Aherns. Later, another came from the south, propelled by the change. Within the gully, with its steep sides and curves, the wind created its own eddies and whorls, with terrifying power. John saw a big manna gum on the bank of the creek snapped off by a wind blowing back up the gully.

The O’Neills saved their house. They had a well-thought-out plan and — crucially — five able bodies to suppress the flames. John reported that ‘my family took on their designated posts’. As the noise grew louder, his sons, Matthew and John, held hoses, wetting down the house until the last possible moment. When the front approached, John shouted, ‘Boys, into the house’, and the family sheltered in the smoke-filled study, from where John, kneeling, peered out towards the west and north-west. He saw fire approach Hannah’s house like ‘a wave’ that ‘picked up the house’ — instantly turning it into what he described as ‘an incandescent ball of sparks and explosions’.

The sky was black. The power failed. The battery-powered smoke alarm screamed until John’s son smashed it. The wind and flying embers battered at their windows. Every so often, John would climb a stepladder and check the roof for flames — although he later conceded that, with only buckets of water, he had little chance of suppressing fire in such an awkward area. He ventured out once to douse a burning verandah post, causing him to be scorched by radiant heat, though he was outside for less than half a minute. For a further seven hours, the O’Neills beat and sprayed outbreaks around their two adjoining houses, saving both.

Whether or not anyone saved their house depended very much on its setting. Houses on hillsides, in thick bush, and especially ones susceptible to embers — during or after the fire — stood little chance regardless. As we have seen, there were exceptions, most of which could be explained by their situation. Keith and Lindy Montell fought flames licking around their house, which stands both on a hillside and in thick bush, although their alpaca enclosure acted as a firebreak. At the height of the onslaught, the Montells told a relative by mobile phone that they were ‘okay but surrounded by flames’. They survived and, extraordinarily, their eighteen alpacas survived as well.

On Lawrence Lane, Alby and Gwen Leckie had sensibly stayed indoors with the air conditioning on. Like many that day, they occasionally walked outside to look — over to the north-west, from where bushfires had always come. They knew from the radio news that a fire was approaching and that, ‘if it was going to come over, it was going to come over there’.

The Leckies had been invited to dinner, an arrangement they confirmed by phone at about 5.55 p.m. While Gwen went down to the paddock to feed the stock, Alby decided to shower. During this time — and, in the midst of a long drought, he did not shower for long — someone rang and left a message on the phone. As Alby dried his hair, he listened to the message. ‘Sorry, mate, if you’re not out now, you’re too late’, it said. Puzzled, Alby again walked outside. ‘Oh, gee,’ he later remembered, ‘you should have seen it. Had it changed!’ Billowing over the ridge just a few hundred metres from his front door was a huge pillar of smoke, with flames visible over the crest. Clad only in his underpants, Alby raced for the paddock and shouted to Gwen to drop everything and get out. Alby hastily pulled on some clothes and jumped into the first of the Leckies’ two cars, calling for Gwen to follow in the other. If smoke or flames blocked them, he called out, they should look for a dam somewhere. At the bottom of Lawrence Lane, Alby turned south onto Steels Creek Road.

On the corner of Steels Creek Road and Hunts Lane, Liz and Grant Williamson were standing outside, watching ‘cars streaming past going south’. Then Liz ‘saw the same cars come back and, to my amazement, go up Hunts Lane’. Liz did not know it at the time, but a water tanker had jackknifed on the road south, blocking the road. As the road cleared, Liz thought it safe to back out onto Steels Creek Road to drive her car (loaded with photographs and documents) to a safer place a short way from the house that they planned to stay with and defend. As she was backing out, she saw headlights through the smoke. It was Alby Leckie, followed by Gwen. ‘Which way do we go?’ he asked. Liz told him about cars coming back to leave by Hunts Lane. He thanked her and the Leckies drove out by Hunts Lane — a decision that probably saved their lives.

As Alby drove up the lane, he saw flames devouring the house of a friend. The Leckies had nothing with which to fight such a fire and had to make a quick decision. ‘Honestly,’ Alby said, ‘there was nothing we could do — the flames were so high.’ ‘We couldn’t get in there’, Gwen remembered thinking. With fire and smoke now filling their rear-view mirrors, they drove on.

As the fire arrived, it seemed to drive out any idea of community, consuming altruism just as it consumed oxygen. The imminent threat of fire brought physiological responses — particularly the production of adrenaline — that enabled people to keep working and fighting to suppress or douse flames, but also focused their attention upon their own predicament. While the fire was among them, people had eyes and energy only to face the immediate threat. But, as we will see, the people of Steels Creek tell many stories of others’ selflessness.

At the top of Hargreaves Road, Keith and Lindy Montell had about ten minutes’ warning when they realised that a fire front was heading towards them. It arrived at about 5.40 p.m. They, too, had always expected the fire to come from the north-west, but this was not a fire that came in orderly waves on a single front. They fought blasts from the west and south-west, but, as they stood outside their house, directing fire hoses onto the burning garden, eddies of wind brought flames from the east — and, indeed, from every direction. Later, they faced ember falls from the east. As at other properties, the fire behaved capriciously: it halted on a line across a paddock and did not ignite a pile of wood heaped for burning although the surrounding area was burned.

At 271 Steels Creek Road, Bob and Dianne Curtis survived through a fortuitous combination of siting and preparation. Their winery building at Yileena Park is a solid masonry construction, a hundred metres or so off Steels Creek Road and set into the slope. Above them, the steeply rising ground is bush. Bob had spent the morning pumping semitrailer loads of water into Yileena Park’s dam; not because he feared an imminent fire, but because, in a drought, the wellbeing of his vines was his principal concern. In the extreme heat, Bob used the water to create a humid mist over his vines — a mist that saved both his harvest and his home. ‘The fire came over the hill from the south’, he told the Leader, but ‘when it got to our boundary fence it pretty much stopped’. The Curtises sheltered some fifty people as the fire struck; many had been guests attending a birthday party at the winery, while others were residents who had fled from the fire elsewhere in the valley.

The story of Alan Mitchell and his partner, Antoinette, was, as Jo Spears described, ‘unbelievably dramatic’. Alan dropped Antoinette at a roadblock, then went back to try to save their house on Steels Creek Road. Because he worked for VicRoads, he had tools in his vehicle — saws and axes — and was able to cut his way through the logs and debris blocking the road. (Jo thinks that Alan’s work helped several others to make their way up Steels Creek Road on the day.) By the time he arrived at his house, the fire was burning at its most intense, and he had to seek shelter. Alan later described how he consciously said to himself, ‘Think!’ He parked his car in a clearing, donned heavier clothes, and then crawled towards a dam near the house. Unable to reach the dam, he was able to find shelter in the lee of a small knoll. This tiny hillock provided enough unscorched ground to shelter him from the deadly heat.

Once the fire front had passed, Alan jumped up, raced back to his car, and drove up to his house. Already flames were licking around the eaves. Finding that his pump had failed, Alan grabbed a slab of tinnies from his fridge. Shaking them vigorously, he used one after another of the twenty-four cans to spray the flames — and succeeded in dousing the fire. Alan had saved his house with beer.

Waiting tensely on the other side of the police roadblock, Antoinette decided that, if Alan returned, she would marry him (as he had long been asking). Three months after the fire, they were indeed married.

On Steels Creek itself, off Hunts Lane, Luke Smith did something that no one else thought to do. As fire raced up the creek towards the north, Luke got out his video camera and began to film the valley burning. Luke’s parents, Athol and Heather, were away from their property that weekend, so Luke was at home alone (except for his dogs). Having worked in forests and on fire crews for the Department of Sustainability and Environment, he knew his way around the bush, burning or not. Clad in overalls and protective gear after readying the house against fire, Luke had little else to do but film the very centre of the firestorm that hit Steels Creek at around six o’clock that evening.

The fifteen minutes of footage are extraordinary. Luke puts his dogs in his ute and walks about the property, filming and commenting on the scene. The footage, in several segments, begins with a surprisingly still evening. The birds sing in the trees; the only sign of danger is the pall of smoke over the ridges. Then the wind gets up. Cattle run from advancing spot fires and a grass fire races over the paddocks. Luke is unconcerned — he knows he can practically step over such a fire. At first it seems bizarre, even irresponsible, to film, but Luke knows exactly what he is doing, and his laconic commentary demonstrates that confidence. If anyone exemplifies a characteristically Australian response to fire, it is Luke. ‘The horse sheds are gone’, he says. ‘The fences are going now … We’re lucky we cut the hay … Better get out from underneath these trees’, he says coolly. ‘Not going to touch the firewood’, he comments as he passes a burning woodpile, ‘That’s fucked.’ In the background, mature trees light up like candles. Luke walks about, suppressing spot fires, keeping a wary eye on the gas tank by the house and on the property’s buffaloes. Then he films from the house’s roof, taking in a terrifying panorama of burning trees. ‘The dogs are all right’, he says, as they loll in the ute’s cab — they are as unconcerned as he is. Luke’s film shows how rapidly and intensely the valley floor had caught fire, and how the trees lining the creek acted like a wick to carry the flames with the wind. It also shows how, even in the heart of the firestorm, areas remained unburned: the Smiths’ house did not catch fire. At a few minutes past six, in the background cars can be seen racing along Hunts Lane to the east. Two of them must have been Alby and Gwen Leckie’s.

On Old Kinglake Road, John and Jenny Barnett were long-term weekenders. They loved their steel-framed timber house, looking out from Yarra Ridge, with a view over the national park. Although they had never seen a fire come close during the thirty years that they had owned the property, the Barnetts had long realised the implications of living in a bushfire-prone area. They had installed roof-mounted sprinklers, had six connected water tanks and pumps, were active members of the local fireguard group, and, on the day of the fire, monitored the relevant websites. Their family and friends were divided about whether the Barnetts planned to stay and defend their property or leave. Around 1.30 p.m., John spoke to a friend and said that, if fire approached, they would probably leave. But they did not, and by the time they realised that a fire was close, it was too late.

At 6.05 p.m., the Barnetts attempted to contact the Aherns (who were dead by this time) — the last evidence that they were alive. Not until the morning of 11 February did a Disaster Victim Identification Team locate the Barnetts’ bodies, at the southern end of their house, of which all that remained was its steel frame. The location of their bodies suggested that they had been sheltering in the bathroom. Investigators concluded that the fire took hold of the house, possibly from the verandah. Establishing exactly what happened at the Barnetts is not as straightforward as at the Aherns, partly because of the absence of direct testimony. Investigators were unsure whether the main fire front or a spot fire destroyed the house. But it is certain that the lack of any worthwhile, specific warning had left people vulnerable. John Barnett’s stepfather believed that ‘as Jenny was checking the CFA website, if it had shown a fire nearby, they would have definitely left’. But Jenny’s brother, Richard Force, understood that ‘their preference was always to stay and fight the fire’. In that situation, even timely warnings might not have deterred the Barnetts from following a plan they had devised well in advance.

Police investigators noticed, though, that ‘there does not appear to have been any firefighting undertaken’ — the water taps were turned off, for example. This is understandable, given the terrifyingly swift approach of the fire. More puzzling, though, is that they did not find signs of ‘any specific preparations’. The police concluded that this was explained by John and Jenny’s plans to leave, although the car seemed not to have been packed. The royal commission accepted that ‘it seems likely that the couple were surprised by the fire and that their only exit route was cut by the fire before they were able to escape’. Again, the lack of a specific warning contributed to the deaths caused by the fire.

Further along the ridge line, on Skyline Road, Guy Williams was still alive and fighting, hosing down the concrete walls of the house he had built himself, within which he and his dog, Buzz, sheltered. The change of wind brought fresh challenges from new, unexpected directions. It drove embers that lodged in his carport and set all three of his cars alight — something he had not anticipated. Suffocated by the smoke and beaten by radiant heat, Guy ran indoors, retreating towards the back of the house as he realised that the door between the house and the carport was giving off toxic fumes. At last the firestorm abated and, after hours of this ordeal, Guy emerged alive, his house intact, standing amid a blackened wilderness.

On the flat of Steels Creek, biodynamic farmer Judy Anderson keeps a daily log, part of the documentation required for her certification as an organic farmer. It records the happenings on her property on 7 February since 1998, with entries such as ‘counted replants’ and ‘picked beans & zucchini’. The entry for 7 February 2009 is brief. It records the successive rises in temperature — twenty-nine degrees, thirty-six degrees, and then forty-five degrees — followed by ‘bushfire burnt us out’, with ‘bushfire’ crossed out and ‘wildfire’ added. This is a central fact of the fire for virtually all of those whose properties it consumed. Fire did not come from the direction most expected, nor did it come from one direction. Instead, many describe it as a ‘wildfire’, a ‘firestorm’, or ‘like a cyclone’. For Judy Anderson, ‘it came from every direction’, she told the Upper Yarra Mail, ‘like a hurricane’. Judy and her Cambodian friend and worker, Kim Tiang Siang, threw buckets of water on fires all over her place for sixteen hours without a break. (‘He’s such a magic bloke’, Judy recalled.) Together, they defended her place — saving Judy’s mudbrick house.