Battling wildfires needs radical new thinking
What set the University of Cape Town alight and what should we do to limit future fire damage? I think the answers are far from straightforward and require some radical rethinking and new action. I have worked on fire ecology for decades, including on how fires spread. In savannas, we have shown that continuous grass fuel is critical. But not in the UCT fire. A section of the top floor of the Pearson building, housing the Plant Conservation Unit, was obliterated by fire (my office, in the wing next door, survived untouched). The building is deep within UCT’s upper campus, protected by a firebreak of over 100m of concrete, tar road and brick paving. It’s the sort of protection foresters dream of. The African Studies library is also deep within the campus, far from the burning veld. How did the fires get there? How did they spread so far with no fuel to feed them? What should be done to stop fires spreading from fynbos deep into the city?
The fynbos shrublands that border the suburbs of Cape Town are intrinsically flammable. Fires recur every 10-30 years or, rarely, as long as 50 years apart. But fynbos will, and must, burn. Similar flammable ecosystems occur on the urban edges in California, Australia, Portugal, Spain, Greece, France, Israel, Morocco, Chile and more. Catastrophic fires in these regions have led to terrible losses.
The frequency and severity of these extreme fires seem to have been increasing worldwide in the past few decades, suggesting a possible link to global warming, or uncontrolled urban sprawl, changes in firefighting, more arsonists. Cape Town is remarkable relative to other Mediterranean-climate regions for the very low levels of property damage from fires. This may be an accident of urban planning. Cape Town has sharp urban limits with buildings restricted to the lower slopes of the mountains — and fires burn slowly downslope.
Until recently, attempts to protect citizens and their properties from fire at the wildland urban interface (WUI) concentrated on vegetation management. That means making firebreaks, suppressing fire starts or prescribed burning under safe conditions to reduce fuel loads. In the Cape it has also included removing invasive alien trees such as pines, eucalypts and Australian wattles. The basic idea is that fire spreads as a front and by managing the fuels the severity of the fire is reduced and fewer buildings burn. Think of fire as a wave — reducing the fuel reduces wave height, turning a potential tsunami into a ripple.
But the fires at UCT did not spread as a wave. Nor did they spread through continuous fuels. They jumped roads, leapt over buildings, crossed rugby fields, shot across the M3 freeway and sparked fire far from the burning fynbos. The stone pines on the edge of campus did not create a raging inferno. On the contrary, many canopies did not burn and are still green or merely singed. Stone pines have long straight trunks with no side branches and therefore no ladders for flames to climb into the canopy. So what have we got wrong about how fires spread?
The fire damage to buildings deep in the UCT campus was not caused by flames from burning fynbos but by streams of embers (or radiant heat?). Driven by extreme, hot, dry, gusty berg winds, and perhaps winds generated by the fire itself, embers penetrated deep into the complex of buildings. As in a campfire, embers usually go out. But some fall on flammable ember traps. Traps include plants, litter and debris on roofs and gutters, air vents, open or broken windows, exposed rafters — it’s a long list.
A quick survey suggests the worst tree ember traps at UCT were palms, evergreen conifers, especially cypresses and young pine trees, strelitzias and the iconic Virginia creeper. The creeper burnt patchily but was seen to carry fire up walls, endangering roofs. Broad-leaved trees, especially deciduous species such as oak, plane and liquidambar, did not burn.
The phenomenon of embers or firebrands igniting fire in spots distant from the fire front is known as spotting. Eucalypts are notorious for spotting. Ribbon gums (Eucalyptus viminalis) produce bark that hangs in long streamers. These curl into little glowing balloons lofted up in the fire plume, where they can start fires 30km from the fire line. Imagine fighting fires that can leap from Table Mountain to Stellenbosch without touching the ground!
Not all eucalypts are that lethal, but very little is known, worldwide, about how ember production varies among tree species or vegetation types. Though spotting is not considered a major feature of fynbos- or grass-fuelled veld fires in SA, it is clearly central to understanding how fires spread from a veld fire across the urban edge.
New research on embers, spotting and fire spread is beginning to change fire management at the WUI in California and Australia. Though the usual fuel management in the wildland continues, there is new emphasis on ember-proofing gardens and buildings. Analyses of property damage in California and
Australia have shown that the biggest risk factor is not the fuel in the adjacent chaparral shrublands or eucalypt woodlands, but the flammable vegetation within a few metres of the house! Flammable garden plants, trees that overhang roofs, and buildings vulnerable to ember entry are the biggest risk. So the responsibility for fire safety is switching to the urban side of the WUI and away from fuel management in the wildlands.
This radical change in managing fire at the WUI has not yet reached SA. There is no research experience on ember production, ember distribution or on the garden plants that embers set alight far from the fire front. We don’t know the source of the embers in the UCT fire. Was it the fynbos burning, or the understorey beneath the pines, or hotspots where canopies of tall pines burnt?
Will removing more pines or broadening fynbos firebreaks really solve the problem? Or is the first priority to get rid of palms, cypresses and other flammable garden plants and to make buildings fire safe by checking and fixing potential ember traps? I think it’s still early days for attribution.
We can learn a great deal from our Australian and US colleagues, and it is certainly time for us to catch up. If episodes of high winds, high temperatures and low humidity increase with global warming, we will need to make sure that our urban properties are fire safe. Advice on how to do so is available, for example from Working on Fire. But we need to do a far better job of advising citizens of when they are at risk and what they need to do about it, underpinned by research in our local circumstances.