Knysna Fires : Five factors that produced the Perfect Inferno

June 2017 will be remembered by South Africans for decades to come. A historical moment when Mother Nature showed her true power and the only option was to get out of her path and watch in awe.

For a week preceding the fire, extreme weather warnings had been issued with predictions of flooding in drought stricken Cape Town and surrounds and rain and strong winds in the Garden Route. Waking up on the 7th June, little did anyone know that within 72 hours 10,000 hectares and in excess of 500 structures in the Garden Route would be burnt, some houses simply reduced to a heap of rubble and vast swathes of pine plantations burnt.

Watch Video Playlist of Knysna Fires : Watch Here

While fires aren’t uncommon in the Garden Route, this fire had all the conditions to make the ‘Perfect Fire’, something that thankfully occurs only every 100 years. With hindsight being a perfect science, understanding the fire has produced insight into the elements that created this inferno.

There were five core conditions that made this fire so unique, namely :

  1. The regional drought conditions,
  2. The fuel load in the environment and suburbs,
  3. Topography of the area,
  4. Hot ambient air conditions,
  5. The speed of the wind.

Each of these conditions would contribute to a fire, indeed the combination of two or three conditions would generate a formidable fire, but the combination of all five factors produced a historical fire.

Looking at each factor and how it contributed to the perfect conditions will assist in understanding the mechanisms and how to plan to mitigate escalated damage in the future.
Drought Conditions :

The Garden Route, along with the rest of the Western Cape has been in the grips of a severe drought for 12 months. While Cape Town exhibits an established winter rainfall, the Garden Route between Mossel Bay and Storms River don’t. Contrary to popular belief, the Garden Route doesn’t have a rainfall season.

The impact of the current drought on vegetation and the resultant increase of fuel for a fire has been substantial. One drought survival mechanism of plants is to reduce the surface area of trans- evaporation, or simply put, to defoliate and drop leaves. The defoliation can represent up to 40% of the trees leaf mass.

With the accumulation of extra leaf mass, the usual systems of decomposition by both chemical (fungal) and mechanical (earthworms, crickets, Pill Millipedes etc) means is retarded and thicker layers build up. This build up in areas can result in the formation of natural compost heaps. Normal composting is an exothermic process reaching internal temperatures between 45 and 77°C. Under certain conditions a compost heap can spontaneously combust.

Alone, the additional leaf litter and potential compost heaps has a potential for starting a small fire, or series of fires.
Accumulation of Fuel :

It is important to understand the different vegetation types in the Garden Route to appreciate the contribution to the build-up of flammable material for fires.

Everyone speaks of the Knysna Forest and the Fynbos in the Garden Route. However, we also have pine plantations, alien stands, coastal thicket and Milkwood Forest. In addition we have agricultural practices which comprise crop production, orchards and pastures for dairy and livestock production.

The importance to distinguish each of these vegetation types is that each has a different contribution to the progression of a fire, some retarding fire and others fuelling fire.

The Afro-montane forest, as a natural stand is fire retardant with the border species preventing the spread of fire to the interior of the forest. This has evolved as a means of protection against the fire climax vegetation of Fynbos.

Milkwood forest is also fire retardant which can be clearly evidenced on the eastern end of Lake Pleasant where the fire was stopped dead in a straight line by this vegetation type.

Coastal Thicket is in some part fire retardant, but the leaf litter, when dry, and dead branches burn and smoulder. While not completely stopping a fire, is can slow the progress of the fire down. The biggest danger of this vegetation type is the potential of flare up after the main fire has stopped.

Then there is Fynbos. Every South African knows the fires of Fynbos. Fast, furious, extensive and most times unstoppable. Fynbos, is a fire climax vegetation and needs fire. There is no exact frequency period of a burn, but it does need to burn. Not burning it has two results. Firstly the build-up of flammable material and secondly the intrusion of either coastal thicket or forest species.

The Goukamma Nature Reserve east of Sedgefield hasn’t had a burn in 30 years except for a small portion near the Goukamma River which was burnt in 2006. It was primed with fuel for a fire and was completely burnt during this recent fire.

Pine and Eucalypt plantations are also prone to burning. Depending on the age and maintenance of the plantations the leaf litter layer can build up and add to the fuel base for a fire. Representing the largest surface area burnt in this fire, the contribution as fuel of the mosaic of plantations has to be addressed.

Finally, the gradual intrusion of alien vegetation, which burns readily, in the form of extensive stands of Black Wattle has also contributed vastly to the provision of fuel to the recent fire.

The collection of fuel biomass from pine plantations, alien stand and a fynbos without partitioning corridors of fire retarding forest was a major contributing factor in the rapid spread of the fires.
Regional Topography :

Ask any old farmer or fire fighter where to run to when a fire gets out of hand and they will all direct you to the kloofs. This isn’t random advice, but the wisdom of experience. Fire likes to race up slopes and along ridge lines, bypassing gorges.

The Elandskraal fire did exactly that, twice splitting along ridge lines and then joining up again. A good indication of the traditional fire paths, because the vegetation has been controlled by fire for millennia, is to check the vegetation that prevails. Again, if Afro-montane Forest occurs naturally in an area, then the chances are that fire hasn’t traveled that way in the past and is unlikely to do so in the future.
Hot ambient Air Conditions : Berg Wind

Dendrochronological studies from trees in the Afro-montane forest lack any clear seasonal growth patterns in their growth rings which indicates a lack of a clear and defined rainfall in the region over a time frame that extends back at least 650 years.

The relevance of this distinction in rainfall patterns is important when weather predictions indicate heavy rain and flooding in Cape Town. When a winter storm is predicted for Cape Town, the anticyclonic weather mechanisms of the southern hemisphere will result in the pressure system veering north east from Cape Town and passing slightly north of the Garden Route.

This deflection north of the Garden Route creates a north westerly wind into the region which results in hot dry air known as a Berg Wind. Depending on the state and strength of El Nino and La Nina conditions, the degree of deflection of the pressure systems varies and can create a period of winter Berg Winds in the Garden Route. A previously notable period of prevailing Berg Winds in the Garden Route was from May 1995 for six weeks.

The mechanics of a Berg Wind are simple. As air descends from altitude, in this case over the Outeniqua Mountain Range, it heats up to approximately 32°C, but can be as high as 38°C. In addition to being hot, the air is extremely dry.

These hot dry conditions played a major role as a predisposition for the fires of 7th June. In the preceding week there were two days of Berg Winds which dried and wilted vegetation in the area. This, added to the extra layer of defoliated material as a result of the drought, prepared ample fuel that required a simple spark to ignite it.
Wind Speed :

Something beyond all human control is the speed of the wind. When a barometric chart indicates a large pressure differential, then expect strong wind. On the 6th June, the barometer started dropping from 1024mb at 00h00 to approximately 997mb by 15h00.

Accounts of how fast the wind was traveling on the 7th June vary, but it was recorded at between 90km/h and 100km/h with gusts exceeding 110km/h, strong enough to divert one aircraft from landing at George Airport and to close the airport till the late afternoon.

Like a bellows, winds of this speed can fan a fire and superheat it in excess of 2000°C which is exactly what occurred on the 7th June 2017.
Thermal Wave :

Add all the above conditions in with the strong wind blowing from the north west and you have the makings of the perfect fire and the creation of a phenomena known as a Thermal Wave. Referenced in literature and rarely seen, a thermal wave is a sine wave flow of super-heated air associated with a fire.

Heat from the fire rises, while the wind blows it horizontally before it touches down and ignites a new fire and then again bounces off downwind. The wave length of this thermal wave can vary between 300m and 1000m allowing it to jump over valleys and rivers and resulting in the seemingly random effect of single houses exploding into flames while those around them are left unscathed.

The mechanics of the thermal wave are interesting. The superheated air rises from the flames and moves laterally driven by the wind. As the air descends into the trough of the wave (of the sine wave form) the high temperature heats everything before it, be it trees or a structure, which then erupts into flame spontaneously before any flame reaches the area. When this wave descends on a structure like a house, it forces the roof down with immense pressure while the extreme heat melts glass and disintegrates bricks. The result is a collapsed pile of rubble.

Eye witness accounts of this leading edge of the thermal wave describe it as a rolling ‘tumbleweed’ flying through the air at between 100km/h and 110km/h. One account even related how the fire overtook their car at 110km/h. The area beneath the peak of the thermal wave has been described by Knysna Fire Chief Clinton Manual as being beneath the ‘dome’, a smokeless zone of earie silence and no wind.

In the Garden Route, during a few days starting on the 7th June, we lived through a historical event, another which has only ever been recorded in1869. Nothing could have prepared us for this fire and nothing could have combatted it. It was the perfect fire, a combination of factors which fuelled the inferno.

Fortunately we have learnt from this event and can formulate plans to never again allow Mother Nature to play a Royal Flush of all five contributing factors to produce a thermal wave through the Garden Route. We can’t prevent droughts or stop the Berg Wind or retard the wind speed, but we can manage the fuel load of the region and establish corridors of fire retardant vegetation and plan a mosaic of safe zones.

Written By Mark Dixon

33 thoughts on “Knysna Fires : Five factors that produced the Perfect Inferno

  1. Check: hectors for hectares.
    Check: the second mention of ‘between 100km/h.” without range delimiter.

    Excellent article!

      1. Thanks…want to share on FB…good to make reference to Author.

  2. Thanks for a well considered article. The ‘simple spark to ignite it’ could have been a lightning strike. Tineke Kraaij reports in an article in SAEON ( that “natural fires (ignited by lightning) dominated the fire regime, accounting for 60% of the area burnt overall, and for more than 83% of the area burnt in the Tsitsikamma region”.

  3. This article misses a CRITICAL issue. It is not the fuel load of the well managed pine plantations that are the issue. But rather the unmanaged out-of-control spread of alien pines, gums etc on state land not under plantation. Richard Cowling and Brian van Wilgen wrote of this looming catastrophe back in 2009. how no-mans-land is now everyone’s problem. This is essentially a governance problem with the commercial imperative to profit from plantations leaving vast areas of unprofitable land uncared for.

    Click to access 8.%20VandFSeptember2009B.pdf

  4. Thanks you for this, Mark Dixon, Sir….good article…but you complicate matters…you know it requires some intellectual effort, like reading it and applying reasoning and science and…stuff!.
    Much simpler to just blame angry vengeful god & those homosexual people you know? Why can’t we just go with that? 😉
    (Sorry, I couldn’t resist that).

    Seriously, thank you for helping me understand this. Yes, even I could understand this and I’m so not a science kinda mind. Appreciate this.

  5. Is that the same Mark Dixon who worked for the Fire Protection Association and helped set up Fire Management Units on the Karatara Road perhaps ten years ago?

  6. Great article. I will certainly be sharing it.
    One thing I found remarkable about the wind during those days: I could hear it approaching for quite a distance as if a large invisible boulder was tumbling over the terrain. A gigantic “ball” of wind, slamming into the house. Is this a specific phenomenon? I’m not the only person who remarked on it.

    1. Apparently the thermal wave phenomenon is difficult to study due to its rarity. However, a combination of wind speed and intense heat would have an influence on the air in front of it, heating it up with a by-product being sound. In addition, the effect on vegetation and structures in front of the wave would be immense, again with sound being one of the energetic by-products. An analogy would be the sound produced when a kettle boils which is an energetic by-product of boiling.

  7. Seeing as a man induced factor has been listed as one of the primary causes (fuel loads), then another crucial, if not THE most important factor was the failure to bring in addition resources earlier to control the spreading of the multiple fires.

  8. Thanks for a very informative article. Do you know if there will be serious efforts to start large scale alien clearing programs in the Garden Route? Experts warned that they were getting out of control and that they were fueling abnormally intense fires but government never prioritised clearing the catchments of aliens. Not only do they create mammoth fires but they also deplete our water supplies and ruin our very special biodiversity. At last year’s Fynbos Forum Brian van Wilgen explained how government will need to more than double their expenditure on alien clearing in the Garden Route just to keep up with the runaway invasions. Is this message actually getting through? I feel more awareness about this critical issue needs to be created. And I was also wondering how it works with the pine companies – are they not obliged to clear the invasions from their plantations into the surrounding veld, or is that purely the land owner’s responsibility who owns the land that’s being invaded? I suppose it’s difficult to prove exactly where individual invasions come from, but surely the forestry companies have a role to play?

  9. Hi Mark, Very interesting article.
    But I’m confused as I also received the identical article via email, but with the following byline: “Dr Guy Preston, PhD (Environmental Science) Deputy Director-General: Environmental Programmes, Department of Environmental Affairs.”?
    Peter Hers

      1. Just for clarity, the article was forwarded to me. I thought the article is insightful, and circulated it to several people. The authorship was not clear and it was then erroneously attributed to me. I did call Mark to apologize for the error. I am the head of the Working for Water, Working on Fire and other programmes, including Biosecurity, that will be working with the Local Authorities to ensure that we prevent, or at least reduce the impacts of, such fires. Guy Preston

  10. Thank you for this extremely interesting and insightful read, I can’t wait to share this on fb for all ‘those’ people who still believe this was the result of arson. My husband and I felt like in the minority, being convinced that this was a natural disaster.

    1. Glad that you enjoyed it. Yes, I too had some heated debates with conspiracy theorists. When the damage from these fires is compared to the damage from the fires between the 3rd and 16th December 2016, it was exponentially bigger. Reports from the December fires were arson but due to the lack of preceding berg winds and weaker southern winds, they didn’t spread so dramatically. In fact, due to the wind being southerly it’s moisture content was high which is markedly different from the dry hot conditions of the berg wind.

  11. It is not true to forests and milkwood retard fires. They burn readily, especially under drought and Berg Wind conditions. What is true is that they occur in fire-safe habitats: areas that fire cannot get into. Under Berg Wind conditions these are areas to the south of mountains (where the wind descends, just east of deep kloofs, which stop the fires getting in.

    Also, the real cause of the fuel accumulation is the current paradigm of putting out all the fires. What happens when you put out small fires. The veld gets older and the fuel accumulates. And the fuel will continue accumulating and the fires will continue being put out, until the day arrives which is hotter and drier and windier than normal, and then all hell will break out.
    Working on Fire and the legal people behind the fire Act do not understand this. Fynbos cannot burn when it is young: there is no fuel. After 15 years it will burn and under most conditions this can be managed, but not under extreme conditions. And after 30 years there is so much fuel it is one Berg Wind away from a conflagatory disaster.

    Until the authorities learn these basic facts, all they are doing is preparing the next disaster. Put out the small fires until the big one overwhelms resources. Next time throw in more resources and surprize, surprize an even bigger disaster!
    We need to mimic the natural fire cycle. “Fuel reduction” burns are needed so that we can have fires that are controllable. These burns should be “prescribed” so that they can be done under conditions that are manageable, and not under extreme conditions. The term “control burns” should be avoided: no fire is ever under total control.
    What we have to watch out is that the authorities burn in the correct season. That is the fire season: the hot and windy conditions. Fires at the wrong time burn too cool and dont sterilize the area, promoting disease and inhibiting recovery from those species that need hot fires for the germination cues. They also kill birds that are breeding and amphibians that are active: instead of aestivating safe below ground in the dry, hot perfect fire conditions.

    Fire is natural and part of our ecosystem. Unless we accept that we are part of the ecosystem, and accept that fires are part of Fynbos, we are settign ourselves up for disaster. Like the Spanish, and Portuguese and Californians: the more money you throw at trying to control the ecosystem, the harder it hits back.

    1. Hi Tonyh, all valid points on accumulation of fuel. The human component of the the fires is one of perception, and if we are honest, over population. Sadly neither are going to chance anytime soon. With regards to the susceptibility of burning of Afro-montane and Milkwood, I do beg to differ. Admittedly the former grows predominantly on southern slopes, however some are on northern slopes. With regards to Milkwood Forest, a blatant example of the fire stopping ability of this vegetation is on the northern slope of the dune overlooking Lake Pleasant.The burn line is vivid on both the eastern and northern border of Milkwood when viewed from the N2. Still, under the correct conditions every source of fuel will burn. But I would prefer to be surrounded by Afro-montane on a southern slope in a big fire. Let’s hope that the natural role fire is appreciated sooner than later, before it is too late.

  12. Hi Mark, Please can I publish this article in the next LIVING in Knysna magazine. It is a free magazine we distribute to home-owners in Knsyna.

  13. Hi Mark, this is a well written article and I hope it will inform the general public and assist with the awareness and education so needed in this area. There are a couple of points I wanted to add.

    The first is on the thermal wave, it is important to note and make clear that while this was very likely occurring during a conflagration on this scale, the majority of home ignitions that occur during a wildfire incident are the result of embers from ember storms alighting on and igniting vegetation situated close to structures and entering the vulnerable areas of structures, such as eaves, gutters, air vents, etc.

    The second is that it is very very important to clarify that when you speak of kloofs being safe zones, it is made very clear that a kloof can only be considered a safe zone if filled with afromontane forest. A kloof is like a magnet to wildfire and literally acts as a chimney sucking the heated air and flames up and into it. This is such a phenomenon that it is stated in the SA wildland firefighters Fireline Safety Pocket Guide Book that one should “Stay away from chimneys, sloots or the top of kloofs”. Firefighters have been killed in exactly these geographic locations. Fire moves through these topographic features the same way that water would flow down them; at an increased rate of spread.

    Thank you again though for such an informative and well put article.
    Patrick Ryan, Vulcan Wildfire Management.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s