L A N D S L I D E !
7-8th November 2005
back to home page

One person's view - a report and images by Mark Barritt

Landslide and flash flood in Black Hill Conservation Park
At some time on or about the 7-8th November 2005, apparently witnessed by no-one, there was a landslide on the Black Hill Massif in Black Hill Conservation Park, which metamorphosed into a mysterious flash flood down Ghost Tree Gully.

Triggering rainfall event
During the 24 hours to 9 am on November 8th there was heavy rainfall in the central Mount Lofty Ranges, causing local flooding in several of the rivers including the River Torrens, and in many streams including Fifth Creek.

There were some convective showers on the afternoon of the 7th with local thunder, but in the late afternoon and during the night there was a prolonged period of mainly orographic rain - caused as very moist air coming from approximately the west rose up the west-facing scarp and over the Range.

As far as I know there is no rain guage within one kilometre of the point where the landslide started so we cannot know just how much rain fell. The landslide site is very favourably situated for orographic rain coming from the westerly quarter - it's near the top of the western scarp of the Range, at the head of a valley, facing the direction the rainclouds were coming from.

We can be confident that somewhere in the vicinity of 125 mm fell at the landslide site. Cherryville about 5 km to the south-east of the landslide received 29.4 mm for the 24 hour period to 9am on the 7th, and then the State's top official guaging for the 24 hour period to 9am on the 8th (along with Uraidla) - 118.2mm. Down near the bottom of the hills face on Addison Avenue in Athelstone close to the western edge of the Park the Bureau of Meteorology's automatic rain guage recorded about 30 mm to 9am on the 7th and about 67 mm to 9am on the 8th. There is a table of the hour by hour rainfall totals for this automatic guage here.

My thanks to Erik Dahl Senior Ranger Northern Lofty District and to the Bureau of Meteorology for providing the rainfall figures for the Addison Avenue gauge. I obtained the rainfall figures for Cherryville from Laurier William's Australian Weather News website www.australianweathernews.com on this webpage (there are rainfall figures for other nearby recording stations listed there too)

You can find a detailed report of this unusual weather event on Tim Thorpe's weather website for the Mount Lofty Ranges www.hillsrain.com. The direct link is http://hillsrain.com/Weather_Station/Events/2005/07-nov/index.htm.

Distant view
This image shows the top approx 75 metres of the landslide scar as seen from 1.5 kilometres west-north-west. Below that the slide is hidden by trees.
You can click on any of the thumbnail images for larger (approx 60 kb) versions and use your back button to return.


Start of landslide
The landslide began high up on the north-west facing scarp of the Black Hill Massif - an arc of high hills including Black Hill and the Sugar Loaves, which dominates much of Black Hill Conservation Park. The two images below show the top of the landslide and the view down the landslide scar from the top. The scar begins on the outer edge of a vehicle track which runs around the side of the steep slope in that area, although the exact point where the slide started is unclear. It may have started some metres below the track and eroded back up to the edge of the track.

The left image shows the top of the landslide - it's about fifteen metres wide, and a depth of about one metre of soil has slid away. You can see the kind of fairly dense vegetation that was carried down with the slide.  The right image is looking straight down the slide path, with the Sugar Loaves in the background.

I am not a geologist but I could see no evidence that the track contributed to the start of the slide. The soil in that location naturally becomes wet and waterlogged in winter and there is some wet-season seepage down the slope. It's also near the top of a wide shallow hollow in the steep slope  which would tend to concentrate water flow and may have led to supersaturation of the soil during the downpour.

The rock strata in that area appear to tilt slightly downwards towards approximately the north-west or west, which means soil would have considerably less purchase than it would on horizontal rock strata.

Another landslide, very small - perhaps only a ton or two of soil and rock, started a few metres above the track near the main landslide. Most of the material came to rest on the track and although some carried on down the slope it did not trigger a major slide.

It's easy to imagine that over hundreds of years soil slowly increased in depth on those upper slopes, and eventually, in the prolonged downpour, the soil became supersaturated with an immense weight of water, and a section gave way.

The first hundred metres
For the first approximately one hundred metres the slide went straight down an approximately 45 degree slope (guestimate),  widening to about twenty to thirty metres.  It does not appear to have slid over the soil but rather to have torn out the top one to two metres of the soil as it went, so that by the time it reached the bottom it had greatly increased in mass.

There was a considerable amount of erosion down the bed of the scar - in parts a gully about two to three metres below the original surface has been eroded into the lower part of the steep slope. What caused this erosion I'm not sure - it was either mainly water, or perhaps it was supersurated soil moving as a slurry or fluid.

In the top left image below we are looking up to the top of the landslide - the white dot at the top of the scar is a person. In the top right image you can see that a 2 - 3 metre deep gully has been dug or eroded down approx the centre of the lower part of the first hundred metres or so of the slide scar. The bottom left image is another shot looking to the top of the landslide with a figure to give a good idea of scale. The bottom right image shows mangled vegetation and rocks lining the edges of this part of the slide. The white post is one metre tall.

What happened at the bottom of the steep slope? Your typical Adelaide Hills landslide would have come to a halt at this point, leaving a large heap of soil, rocks and mangled vegetation. But this landslide behaved quite differently. It kept going down the valley floor, not as a typical landslide but as a flash flood.

By the time it reached the bottom of the initial steep slope it was not a sliding mass of solid debris but a fluid of some kind. Perhaps akin to the mudflows that we sometimes hear about in other countries. There is no big heap of debris at all at the bottom of the steep slope.

It puzzles me considerably as to how it could have contained so much water that the whole mass flowed as a fluid rather than as a solid, but perhaps the answer lies in the saturated nature of the soil, and the kind of soil it was.

Perhaps the sandy, gravelly and stony nature of the soil on that slope can hold enough water so that when saturated and sliding, it effectively turns into a flow of water in which the soil particles small and large are suspended.  More likely I think it turned into a slurry of some kind. But I will leave that to the experts to judge, if any should study the slide, and I hope they will.

One explanation I can rule out is that there was already a flood
coming down the northern tributary of Ghost Tree Gully and the slide debris
entered the stream as solid rubble and was simply carried downstream by the
floodwaters. There was no such preexisting flood coming down the tributary.
There's no visual evidence of such a preexisting flood, and I doubt if even a
cloudburst of several inches in half an hour could produce a flood of
anywhere near the required magnitude in a catchment of that size.

Entering the northern Ghost Tree Gully tributary streambed
At the bottom of the steep slope the slide weaved to right and left a bit, and then crossed the vehicle track that comes up Ghost Tree Gully and entered the streambed of the northern branch of Ghost Tree Gully. Fortunately it did not erode the track significantly - it mostly flowed over the top.

Descent through Red Gum woodland
For the next approx 200 metres the flow descended along a broad gully floor perhaps thirty metres wide on average, and a downward slope of perhaps 20 to 25 degrees (guess only - not measured). This is quite steep for a stream, and the flood eroded channels up to about 2 metres deep in places, and deposited rocks, stones and tangled vegetation in other places particularly along the edges of the flow and against some of the gum trees. Below are three images taken along this stretch of the flow.

The Upper Ghost Tree Gully track crossing
Here this northern tributary of Ghost Tree Gully joins the southern tributary, and the united stream continues downwards as the main section of Ghost Tree Gully, with a considerably reduced slope. The flood took a sharp turn to the right at this point, and met some large logs lying in the stream bed. These jammed against a large gum tree and acted as a sieve and collected more logs from the flow and created the biggest log jam along the flow.

Long run down Ghost Tree Gully
This main section of Ghost Tree Gully is about seven hundred metres long, the stream bed is narrow and one of the adjacent slopes is quite steep in places. The downslope of the streambed is about 15 degrees (guess), and the gully floor is wooded with Red Gums and locally Sheoaks and wattles, with a mostly grassy understorey. As a rough approximation, a bit more than half this stretch of streambed has no sudden bends and is without many obstructions,and there the flood stripped soil, stones and ground cover vegetation from the stream bed and banks but deposited very little material. Below is an image taken along this straighter less obstructed stretch


A bit under half that seven hundred metre stretch of Ghost Tree Gully proper is bendy with some flatter areas and some large gums standing in the stream bed. There was considerable deposition of logs and other bits and pieces of vegetation, and stones, gravel and mud, and some incursions onto the nearby vehicle track and adjacent gentle slopes.

The lower track crossing
At the bottom of this confined steep-sided stretch of Gully the vehicle track crosses the stream bed and the stream slope lessens considerably, and there was a mixture of erosion and deposition of material in this area. The first image below shows the vehicle track where it crosses the Ghost Tree Gully streambed, and the second shot is looking downstream from the crossing. Some quite large rocks were still being rolled along by the flood so it still had plenty of momentum, although I'm not sure of the original location of the large 0.7 approx metre tall rock in the middle ground.

The flood continues downstream
About a hundred metres below that crossing the stream bed leaves the Park. I haven't examined the downstream stretch outside the Park, but it appears from a distance that the flood kept going as a flash flood for about three hundred metres depositing considerable quantities of debris, and then may have begun petering out.


Erosion and deposition
There was abundant erosion and abundant deposition during the event. The flash flood was very abrasive and erosive. It stripped the ground cover vegetation and some of the tall shrubs and sapling gums from the stream floor (or knocked the shrubs and saplings over), and in some places deepened the pre-existing stream bed, and it dug a new stream bed up to 2 to 3 metres deep in places down the lower part of the initial 100 metre steep slope.

Stones and small boulders and logs were caught up in the flood or came down in the initial slide and are a feature of the debris patterns left by the flood.

Another feature is the lines of stones, rocks and vegetation left along the edges of the flow at its peak.

Effect on vegetation
All vegetation along the initial approx 150 metres of the slide has been carried downhill or strewn along the fringes of the slide, except for one tree which was larger than most and not near the centre of the slide. The original vegetation cover of that stretch was a pretty dense heath shrub cover with many eucalypts perhaps up to 6 or 7 metres tall  (I haven't looked at its actual composition).

Further downstream, wherever the flood was fast-flowing and deep it stripped all the vegetation ground cover and most of the shrubs and often the topsoil from the bed and lower banks of the original stream bed so now there is a newly-shaped stream bed mostly bare of vegetation except for the gums.

I saw a couple of small Red Gums in the upper part of the Valley with trunks about 30 cm in diameter with all the bark right round the trunks stripped off over about a two metre section of lower trunk - so I suppose they will die unless they sucker. I didn't notice any living trees larger than 30 cm trunk diameter knocked over or ringbarked, but many probably lost some of their bark on the side facing the oncoming flood.

Usual flow of Ghost Tree Gully stream
The stream in Ghost Tree Gully is annual, on average I think it would be dry for 9 - 10 months of each year, and usually very mild mannered. I don't recall any obviously erosive flood events in the past 6 years. A few thousand gallons an hour or less would probably be a typical flow for this stream in early spring after a wetter than average winter. In a particularly dry year it probably would not flow at all. A tangle of grass and a few shrubs typically grew right in the stream bed.

Abundant evidence of flash flood
The flash flood has left abundant and very obvious evidence of its passage and some of this evidence should be very long-lasting, still visible in fifty to a hundred years I would think on careful inspection. This evidence includes various log jams, and lines of stones and rocks and logs well above the stream bed.

I wonder if even a one in two hundred year thunderstorm deluge could have caused a flood of this magnitude. I don't think so. Even with this five inch downpour, the unaffected southern tributary of Ghost Tree Gully which has a considerably larger catchement area showed no evidence of a significant erosive flood.

Another perhaps more subtle sign of the flood is the damage to the bark of some of the gum trees in the path of the flood. Those scars might be visible for a long time, and might be a tell-tale sign to future generations of some dramatic flood event, and could be something we could look for in other stream beds as an indication of a past flood drama.

Unusual nature of this slide and flood
In my opinion this has probably been quite a rare natural event for South Australia, and one we can learn a lot from. We are accustomed to thinking of  slow erosion and deposition by streams over the years as being the main force shaping the details of the stream beds in the Mount Lofty Ranges, with an occasional minor flood to do a bit of rapid sculpting here and there.

But this slide and flash flood was on a bigger scale. In perhaps a minute it changed the stream bed and banks perhaps more than the past  hundred years of slow stream erosion, and it's added features that probably could never be formed by rainfall runoff alone.

I wonder if there are any other records of such slides and subsequent flash floods in the Mount Lofty Ranges and elsewhere in Australia. Can we find other placid streams with log jams, lines of rocks high up on banks where normal floods could never put them, gum trees on the valley floor with bark eroded away on the upstream side, and other evidence indicating that once there was a flash flood caused by a landslide?

Perhaps a few kinds of soils are capable of flowing as a fluid when they are saturated with water. Or maybe there's some other explanation for why this particular slide became a flash flood. Perhaps when the slide started it released a large volume of water from underground which came flooding down along with the soil, and that was the source of the water that caused the flash flood. That would make it an even more unusual event.

One of the features of the flood is the way it moved rocks and stones about with ease - they are strewn everywhere. This might (speculation) be evidence that the flow was a slurry of soil and water rather than a flood of water carrying soil particles. A slurry could be perhaps two or three times as heavy as water (guess) and exert much more force on stones and rocks.

I am particularly interested in the idea that floods of various kinds are more important in shaping our streams and alluvial fans and similar landform  features in South Australia than we tend to think. These kinds of flood events of biblical proportions (well they would seem that way to the plants and animals living in the streambeds) might be so infrequent that most of us never experience them.

There's a lot we can learn from this event, about erosion and deposition generally in our streams, and about the forces that shape our stream beds. And about signs to look for that might indicate that similar events occurred in the past.

The easiest way to reach the Ghost Tree Gully is to park your car at the top of Addison Avenue in Athelstone near the Wildflower Garden, and walk through the Park entrance right at the end of Addison Avenue. Proceed about fifty metres up the vehicle track where you will come to an intersection with tracks leading in three directions. Take the vehicle track leading to your left - the one heading approximately north (not the track going straight ahead). Follow that vehicle track for about 600 metres - while you are walking along that track you should have rising ground on your right and ground falling away to suburbia on your left. At about 600 metres you will come to a four-way track intersection. Take a sharp right turn at this point to follow the track that runs up the valley (Ghost Tree Gully). Along that stretch you should have rising ground to your right and ground falling away gently to the valley floor on your left. In about 300 metres you will come to the lower Ghost Tree Gully crossing where the effects of the flood will be abundantly visible.

The vehicle track then closely follows the stream bed of Ghost Tree Gully for a bit over a kilometre, right up to the point where the landslide first entered the Gully floor.

To get to the top of the landslide, keep following that vehicle track - it takes a u-turn to your left at the point where the slide crossed it and climbs steeply for about 200 metres where you will come to an intersection - a branch of the track continues on to the top of the ridge and another branch goes sharply right. Take that right hand branch and proceed up a moderate slope for perhaps three hundred metres when you will come to the point where the landslide commenced. (Along that stretch of track you should have steep rising ground on your left and steep falling ground on your right).

I would allow one and a half hours for a walk to, and inspection of, the lower part of Ghost Tree Gully and return, two and a half hours for a leisurely walk up to the point where the slide first enters the northern tributary and inspection and return (recommended if you have time), or three and a half hours if you want to include a walk right up to the start of the slide (quite steep slopes in parts and tiring if you are not a regular bushwalker).

I would strongly recommend not attempting to climb either up or down the initial approx 100 metres of the slide - it's very steep, slippery and there are abundant loose rocks and stones. There's a vehicle track which takes you right to the point where the slide started and it's a lovely walk with a great view over Ghost Tree Gully and the plains. I would also suggest caution at the top of the slide - I don't know how unstable the edge of the track is up there where the slide started - it might be quite unstable.

Throughout the length of the slide there are plenty of loose stones and rocks and logs that might move underfoot so it pays to be cautious all the time when wandering about on the debris.

Adding information, comments or corrections to this article
If you have any additional information or observations or you can refine my guestimates of slope angles, distances etc you are welcome to email them to me at mark@blackhillpark.com and I will add relevant info to this article.

Mark Barritt

top of page