Category Archives: Weather

Taunting the Dragon

This post is intended more for pilots than backpackers, presented in layman’s terms for both to read, but it should drive home the lesson that the Natal Drakensberg is Southern Africa’s highest place and must be treated with reverence. In all four of the aircraft accidents mentioned here, the pilots failed to account for the foreboding height of this mountain range.

I was a licensed private pilot many years ago. Several of the aircraft I piloted were later written off by other pilots, and I saw the wreckage of some lying forlornly and accusingly on a hangar floor. Since then, I have encountered two other wrecks in my wanderings in the Natal Drakensberg. The experience has a profound impact, trying to comprehend the sudden finality, the waste of life and a good aeroplane, and the anguish that must have followed.

On the first occasion, I summitted Gray’s Pass and settled into Nkosazana Cave with a small school group in heavy mist. It was summer and cattle were lowing across the valley from us, but every now and then we heard the puzzling clatter of metal through the mist. The morning broke bright and clear, and from the mouth of the cave we could clearly make out the wreckage of a light aircraft. It lay about halfway up the back of Champagne Castle, almost directly opposite us. I was unaware of an aircraft accident in this area, so we were obliged to investigate. I later learned that a father and his 15-year old son had lost their lives there a few months earlier, headed toward Virginia Airport at the coast from Gauteng.

There was very little left of the aircraft, which I quickly determined had been a twin-engine, six-seat Piper Seneca. It must have been at its cruise speed of 370 km/h when is left wing tip struck the slope. We could make out this impact point clearly, from the shattered red cover of the port wingtip light. The nose must have slewed to port, but the aircraft hit a rock embankment, disintegrating and exploding at the same time. Its two Continental flat-six engines were the largest remaining parts, but they were badly broken. I found the engine synchronisation tube and the shattered instrument panel, which included a 3-axis autopilot, radio altimeter and what looked to be a complete IFR suite. If the Seneca had been another 200 metres higher, father and son would have reached their destination none the wiser. It appeared that some of the wreckage had already been carried off – by the recovery team, the wind and inquisitive Basothos. One of the wing flaps made its way down to Keith Bush camp at the bottom of Gray’s Pass and remained there for several more years. This was what air crash investigators call a Controlled Flight Into Terrain (CFIT).

A year or two after this accident, another Piper Seneca went missing in the Natal Drakensberg. An extensive search found nothing, but months later, backpackers came across its wreckage near the bottom of Ifidi Pass. This was another CFIT accident. The pilot had flown his aircraft into a 1 km high, vertical escarpment face at 370 km/h. Needless to say, there was not much left to recognise. I heard that coins in the passengers’ clothing had been bent. If you lie on your belly on the edge of the escarpment above the crash site and carefully look down below you, you can still make out the wreckage to this day.

More recently, on 5 December 2012, a South African Airforce AMI Turbo C-47TP (a turboprop-powered Douglas C-47 Skytrain / DC-3 Dakota) carrying a team from Waterkloof to Nelson Mandela’s retirement home outside Mthatha in the Transkei, flew belly-first onto the top of the ridge behind Giant’s Castle at an elevation of 3260 metres (10695 feet), killing all eleven people on board. It was 09h45, Instrument Meteorological Conditions (IMC), and another instance of CFIT where the crew had failed to account for the height of the Natal Drakensberg escarpment. In this case, another 5 metres higher and they would probably have made it. The crash site was found as soon as the weather lifted. It would have been very easy to locate – the point on the map where the aircraft’s stated course and altitude intersected with the ground.

I now want to dwell on the second occasion that I saw a wrecked aircraft in the Natal Drakensberg. This time I was aware that it was there and had seen it from a distance on previous occasions – a tiny white speck below the escarpment near No Man’s Peak in the Garden Castle Wilderness Area. This time, though, I had planned to stay at nearby Fun Cave for two nights so we would have plenty of time to visit the crash site. My goal was to see for myself what had happened. We took only photographs and left everything where we found it out of respect for the families of those who died there, but I did report back to Ezemvelo KZN Wildlife that the site needed cleaning up, which they later did with the help of the SAAF.

In this case, the aircraft was a single-engine Beechcraft Bonanza V35, where the “V” denotes its unusual V-shaped tail. It flew into a steep mountainside at 13h16 on 10 October 2006, less than 200 metres below the lip of the escarpment between No Man’s Peak and Sandleni Buttress. Three adults and an unborn child were killed in the accident.

An intact Beechcraft Bonanza V35, displaying its V-tail
Wreckage as seen from Speir’s Cave

Without backpacks, we climbed the steep spur to the right of the crash site, then headed left across a small tributary of the Umzimkhulu River, where we found the aircraft’s flat-six Continental engine, similar to the ones used in the Piper Seneca, lying almost in the water. The engine was surprisingly intact, but the propeller was completely missing and the metal skin that was still attached had been severely twisted. This suggested that the V35 had not impacted the ground nose-first, that the engine had been running at high power, that the propeller had struck the ground and the torque had ripped the engine from its mountings and sheared off the propeller, freeing the engine and allowing it to roll down the hillside until it came to rest in the stream bed about 50 metres below the main wreckage.

The severed Continental flat-6 engine

We then headed up the next spur to the main wreckage. It was clear the aircraft had caught fire, but it had not completely burned out even though the fuselage had remained in one piece. Both wings had been sheared off. The integral fuel tanks in the wings had completely burnt out, taking the wing skin away, but curiously, close to the edge of where the tanks had been, there was unscorched white paint. These observations suggested that the aircraft might have come down in rain, so although there was a fire, the rain extinguished it before everything was completely burnt out. The fact that the fuselage had remained in one piece after the impact with the ground suggested that this contact had taken place at a relatively low speed – perhaps the stalling speed, which would have been around 120 km/h “clean”, at this altitude and load. This presented the horrifying spectre that the pilot could see the ground rising beneath him and had tried to take evasive action by opening the throttle fully and pulling the nose up into a steep climb.

Most of the fuselage was burnt out
The V-tail was mostly intact
One of the wings with its burnt out fuel tank

We came across some haunting personal effects: a pair of neatly folded blue jeans with fire damage; a dented deodorant can; a Smith’s Pilot Guide lying open in the grass with edges burnt and a page flapping on the gentle breeze. These had been real people whose lives should still have lain ahead of them. Further to the left on the spur, under a small, low rock overhang, I found what I presumed to be the aircraft’s locator beacon, whose battery had been removed. It had just been left there, as if to say that the hunt for the crash site was now over.

The tattered remains of a pilot’s manual

Ahead of the fuselage, the most significant items were the destroyed propeller, including its variable pitch hub, and the nose-wheel leg complete with its very heavy lead ballast. One propeller blade had been torn off, while one was almost straight, again suggesting that it had come to an almost instantaneous stop. The tip of the propeller spinner was obviously damaged, but its condition confirmed my earlier notion that the aircraft had pancaked onto the ground. The aircraft may even have been “hanging” on its propeller – i.e. pointing almost vertically and about to fall backwards. If so, it would have hit the ground at much less than 120 km/h, suggesting a possibly survivable impact except for the ensuing fire.

The damaged propeller hub
The missing propeller blade seems to have melted

Above the wreckage was perhaps the most significant tell-tale sign of what happened during the aircraft’s last moments. Above an embankment, and just below a cliff face, a heavy impact scar was clearly visible. I surmised that the pilot had seen the rising ground, pulled the V35 into a steep climb at full power in the hope of getting over the escarpment, then saw the cliff face coming up, pulled back further into a climbing left turn, hanging on the propeller or perhaps already entering a full stall, at which point the propeller struck the ground, breaking the engine free of its mountings, ripping off the nose gear, and releasing the fuselage, which then slid backwards downhill before coming to rest. The wings would have broken off during this slide.

The impact scar above the main wreckage
The wreckage as seen from the impact scar

The Civil Aviation Authority’s official accident report, which I only came across recently, lists the root cause of the accident as the pilot’s failure to revise his altitude after selecting a more direct route to Bloemfontein. A passenger or air traffic controller who knew the Natal Drakensberg could have warned him of his error. No Man’s Peak is 3100 metres or 10170 feet above sea level. The pilot had chosen to fly at 8000 feet and impacted the ground at 8957 feet (2730 metres) while climbing to avoid the rising terrain. Even if he had been able to clear the edge of the escarpment, he would still have had to clear the ridge behind No Man’s Peak, which is not much lower than the peak itself. To safely clear the Drakensberg mountains on this heading to Bloemfontein, the pilot would have had to fly at an altitude of at least 12700 feet. He was flying almost 4000 feet too low – another case of CFIT. To put this into perspective, the highest peak in the United Kingdom, Ben Nevis, is 4413 feet!

Not mentioned in the report is the influence the presence of a pregnant passenger may have had on the pilot’s decisions. She was apparently due any moment, hence, possibly, the rush to get back to Bloemfontein. The Beechcraft Bonanza V35 is unpressurised and this one was apparently not carrying oxygen. The pilot may have wanted to avoid subjecting her and her unborn child to high altitude flight, where the risk of her getting hypoxia would have been much greater. In any case, air law requires that, for flights exceeding 30 minutes at cabin pressure-altitudes between 10000 feet and 13000 feet above sea level, pilots of unpressurised aircraft must use supplemental oxygen. Failure to do so can result in hypoxia, whose effects are very similar to being drunk. The difference for backpackers is that they usually have time to acclimatise before reaching this altitude and can therefore manage without supplemental oxygen. Even so, some hikers still suffer from altitude sickness at the top of the Natal Drakensberg escarpment.

Beware, all you who enter into the dragon’s lair. Do not taunt it, for it will awaken to smite you down, then return to its rest. Step softly, step carefully, lest you wake it once more!

Wreckage Impact Scar Melted Metal Propeller Hub Smith's Pilot Guide Wing V-Tail The Main Wreckage Engine Beechcraft Bonanza V35 Beechcraft V35 Crash Site

White Squalls

If you have ever seen the 1996 movie White Squall starring Jeff Bridges, you will have some idea of what a white squall looks like at sea. According to Wikipedia, “A white squall is a sudden and violent windstorm at sea which is not accompanied by the black clouds generally characteristic of a squall. It manifests as a sudden increase in wind velocity in tropical and sub-tropical waters, and may be a microburst.

Microbursts also occur over land and have been identified as the main cause of some devastating aircraft landing accidents. Unsurprisingly, given the violent summer storms in the Natal Drakensberg, they happen there too.

My first experience of one happened many years ago when I was leading a group down from The Hub to Gxalingenwa Cave. It had been an oppressively hot afternoon, and at a narrow section of the Gxalingenwa Valley just above the cave, we stopped for a swim. With additional heat radiating off the rock faces there, it must have been close on 40 °C. A major storm had been building over the Giant’s Cup, and the first indication that something was about to happen was when a few large hailstones starting falling amongst us.

We made a beeline for the cave, entering it via the top approach and the waterfall to save time. We had no sooner arrived than a white wall passed down the valley in front of us. It was not mist – it was a vertical wall of white, marking a sharp boundary between the clear air in front of it and the hail and very heavy rain behind it. The precipitation was so intense that it did look a little like mist while it was over us. The worst was over in less than 5 minutes, but the storm continued unabated overhead for another hour.

One day, returning home after a hike, we ran into a microburst on the road just outside Bulwer, on the way down the Umzimkhulu Valley. I had the headlights on, and the wipers were making little difference to visibility even though they were on maximum speed. There was some hail too. The next moment we popped out of the wall of the microburst with the wipers flailing wildly, and us squinting in the sudden brightness. The road was bone dry. We passed a lady walking in the opposite direction. She could not have had any idea of what was coming her way.

On another hot afternoon, we were busy settling into Vaalribbokkop Cave in the Monk’s Cowl area, fetching water and washing off the sweat at the junction of the streams below the cave. I had just returned and was hanging stuff out to dry when I noticed a wall cloud coming up the valley. Some of the others were still making their way back and were about halfway to the cave. There is a big, smooth rock at the front of the cave which makes an excellent stage, so I jumped onto it, yelled at the stragglers, pointing to the wall cloud and miming for them to run!

They got the message and ran, making it into the cave as the wall cloud reached us. It had been approaching at probably around 20 km/h. Again, as it passed over the cave, it contained torrential rain and a lot of hail. Some of the hailstones were ricocheting off the rocks into the cave. And then it was gone!

Being by now fairly familiar with these wall clouds, or “white squalls” as I like to call them, put me at an advantage on the next occasion that I encountered one. We had stopped at Lakes Cave in the Cobham area for lunch, and a major storm was once again developing over Hodgson’s Peaks. We were headed for Glade Cave, so as soon as lunch was over, we left Lakes Cave and started making our way up to the tarns on the ridge above it. I was in the lead, and as I approached the summit of the ridge and could see over it, I looked straight into a wall cloud.

I yelled to everyone to get back to Lakes Cave as fast as they could, no explanation given other than that “all hell is about to break loose!”. I was even encouraging them to run if they could, but to be careful not to twist an ankle. We arrived back under the shelter of the cave only a few seconds apart, everyone looking at me in expectation of an explanation. All I said was “Watch!”. On cue, the wall cloud reached us and there was torrential rain and lots of large hailstones. This time, the wind was gusting so hard I thought we might even have a tornado on our hands, so we got the smallest kids to climb up onto the ledge at the back of the cave and wedge themselves in there. Again, it was all over is less than 5 minutes and we continued on our way.

I have never seen a fully developed tornado in the Natal Drakensberg, but I have little doubt that they do occur. The rural village of Impendle, which is not far from Underberg, has had two devastating tornadoes. I was on the plateau below Lammergeier Cave once, heading down to Surprise Cave, when we saw a funnel cloud drop out of a storm cloud over the Bushman’s Nek Valley. We watched fascinated, then with growing alarm, as the funnel slowly stretched downwards towards the valley floor. My mind was racing about what to do if it came our way. There was a long rock band just behind us with a low, shallow overhang, so I decided we would have to lie down on our sides using our shoulders to wedge ourselves under it and our backpacks for protection. When I looked back at the funnel though, it had slewed horizontally and then it disappeared.