Movement at the horse paddock

Recently, while driving along Rainworth Road past the south-east corner of the governor’s property, I noticed something new. Running through this grassy area, which I understand to be the governor’s old horse paddock, were two rows of newly planted trees. Interspersed among the new trees, which still were housed in their protective plastic tents, were what looked to be some native grasses or lomandras.

The pond at the edge of the governor’s paddock. The two rows of newly planted trees can be seen in the background.

One of the two plantings follows a course that might be similar to that of the original stream of Western Creek, which once flowed through this land.

The two new plantings in the governor’s paddock at the south-east corner of the Fernberg property.

I’ve suggested previously that this piece of land is something of a missed opportunity. It is visually uninspiring, ecologically bland, and inaccessible to the public. Its one redeeming feature is the rehabilitated pond along the fence, the beauty of which only serves to illustrate what could be done with the rest of the site if the will was there.

These new plantings, while far from a full-scale restoration of the site, are a positive sign. I look forward to seeing how they develop, and to learning if they might be part of a larger rehabilitation effort.

May Day mayday! A tale of extraordinary weather and extra-ordinary lasagne.

It’s been a long time between posts. As per the forecast, my PhD has reduced this blog from a bubbling stream to a lifeless dry creek bed. But any stream can be revived with a little rain. And with a lot of rain, it all comes flooding back.

'UQ Falls' - posted to the Facebook group UQ StalkerSpace by Panashé Machinguara just after midday on 1 May 2015.

‘UQ Falls’ – posted to the Facebook group ‘UQ StalkerSpace’ by Panashé Machinguara just after midday on 1 May 2015.

‘A lot of rain’ is one way to describe what Brisbane experienced on Friday the 1st of May, 2015. ‘Extraordinary weather event’ would be another. The gauge in the CBD recorded 182mm of rain in the 24-hour period from 9am that morning. To put that in perspective, the historical average rainfall for the entire month of May is 74mm. The record for the largest daily fall at this time of year stands at 190mm on 26 April 1989, but the recent event set a new daily record for May, with the previous record being 149mm on 9 May 1980.1

Personally, I’d had enough of the rain long before any records were broken. My trip into the University of Queensland at St Lucia that morning took nearly two hours because of a track fault somewhere near Yeronga Station. ‘Local flooding’ was said to be the cause. I enjoyed the rain a lot more from the other side of my office window, but was forced to venture out again at around 4pm. By that time, every road and pathway on the campus was a raging torrent. Continue reading


  1. All of this rainfall data is available from the Climate Data Online section of the Bureau of Meteorology website. The historical data comes from the the Brisbane Regional Office station (040214), which operated from 1840 until 1994, while the more recent data comes from the Brisbane station (040913), which commenced operating in 1999.

The Western Creek floodgate: a reality after 125 years

Way back in January, I reported that the City Council had begun constructing a floodgate at the mouth of Western Creek. This is the spot where the Milton Drain meets the river, right next to where the dilapidated shell of the old floating restaurant still stands. If you have passed by this spot in the last few months, you will have seen that the construction is complete: Western creek finally has its floodgate.

Hugh Bell's letter to the editor of the Brisbane Courier in July 1889, arguing for the use of tidal valves to prevent flooding in Milton and Rosalie.

Hugh Bell’s letter to the editor of the Brisbane Courier in July 1889, arguing for the use of tidal valves to prevent flooding in Milton and Rosalie.

I say ‘finally’ not because more than three and a half years have passed since the 2011 flood (some people might even call that response time quick), but because more than 125 years have passed since a floodgate at the mouth of Western Creek was first proposed. In July 1889, while the mud was still drying from a flood that peaked at 3.75m on the Port Office gauge (about 2 ft lower than the 2011 flood), a Milton resident named Hugh Bell wrote to the Brisbane Courier, asking “Why are not tidal valves or flood gates being used” to prevent flooding in Milton, Rosalie and other low-lying suburbs? This question was no doubt raised again in March of the following year, when the river peaked at 5.33m — just under the mark of the 1974 flood. In response, the Toowong Shire Council in October 1890 unveiled an ambitious whole-of-shire1 drainage scheme which featured floodgates at the end of Western Creek and Langsville Creek. But the council could not afford to build the whole scheme, and so built it piecemeal instead. Some components, such as the drain through Red Jacket Swamp, took years to get finished, while the floodgates never materialised at all.

But the idea didn’t die: it resurfaced after every flood, only to recede again with the memory of the floodwaters. It probably sunk into near-oblivion after the completion in 1985 of Wivenhoe Dam, which — if you believed the real estate industry — was supposed to save Brisbane from ever flooding again. But when the flood in January 2011 proved — unless you believe the real estate industry — that Wivenhoe Dam is not a bottomless pit after all, attention turned once again to more localised methods of flood mitigation. The Brisbane City Council commissioned a study into backflow prevention, and since 2012 has been rolling out the installation of floodgates — more properly known as backflow prevention devices — at prioritised spots along the river. Continue reading


  1. The Toowong Shire fronted the river all the way along the Milton Reach, from Boundary Creek (where Boomerang Street is today) to Toowong Creek (near Gailey Road). The boundary between Toowong and Ithaca shires began where the railway line crossed Boundary Creek, and continued along the railway until it reached Baroona Road, then known as the Boundary-road because it divided the two shires all the way up to the top of what is still called Boundary Road today. (Back then, Baroona and Boundary roads were joined at the upper end of Norman Buchan Park.)

oncewasacreek: a words-eye view

Note – the pictures in this post contain two images, which flip back forth when you hover your mouse over them. On a mobile device these images will switch when you touch them, and then switch back when you scroll away … if you are lucky. For best results, please view this article on a desktop or laptop.

Time now for something completely different.

This site has reached a point where I am starting to look back, usually with some degree of bemusement, at all that I have written and done over the last two years. For a while now I’ve been meaning to overhaul the site’s structure to better serve the unexpected array of topics that I have ended up covering. But invariably, something more interesting captures my attention. This post is a case in point.

Recently I have been playing with network graphs as a way of visualising concepts and data. Networks are used as an analytical tool across many branches of science, because just about any phenomenon can be understood at some level as a system of discrete but inter-related entities (or nodes, in network-speak). Perhaps the most obvious examples are social groups and computer networks, but a myriad of other things, from gene sequences to ecosystems to economies, have been modelled and explored with the help of network theory. (Check out this page for a some examples of network visualisations.)

Networks are also useful for analysing text, because they can show relationships among words, such as which words occur together most frequently. And the best bit is that we can get computers to do all the counting. We still need to tell the computer which words to look out for, but thankfully there are dictionaries and word-recognition tools that do much of this work as well.

I wanted to have a go at analysing some text, so I found some software1 and fed it a collection of eleven of my own essays from this website.2 The result, shown below, is a kind of word-map of the topics covered by the eleven articles. In this case, the topics are limited to the names of people and locations, simply because these are much easier to identify than other kinds of topics. In the future, I hope to bring other topics (floods, maps, roads, etc.) into the mix.

The network of all people and places mentioned in the eleven articles. Node size reflects term frequency, and line thickness reflects the strength of co-occurrence. Hover over the image to see the network with the labels removed.

Continue reading


  1. There are various packages out there, some free and some commercial, but the one that I used is called KNIME. It has an intuitive graphical interface, which means you don’t need to know any code.
  2. The essays were: A history of Gregory Park, John Olxey and the chain of ponds, Milton: Brisbane’s western frontier, The waters of Milton, Bottles and cans (an adventure in suburban archaeology), The broken lands of Toowong, Topography shock: postcards from the south, The history of the Coronation Drive Office Park, and the three installments of Uncovering Langsville Creek.

The Duckbill Dimension

I wrote last week about the backflow prevention device — let’s call it a floodgate — that the council is installing at the end of the Milton Drain to protect the Western Creek drainage system from backflow flooding (that is, flooding from the river rather than from local stormwater). The Western Creek floodgate is the most recent — and probably the most ambitious — in a series of such works being done across Brisbane as part of the City Council’s Backflow Program.

Another place where a floodgate is presently being constructed is the outlet of the Leybourne Street drainage system in Chelmer. This system meets the river as a large open channel at Faulkner Park. There isn’t much of a creek left, and I don’t know what it used to be called. Evidently though, it flooded quite seriously in January 2011.

The end of the Leybourne Street drainage system in Chelmer, soon after the peak of the 2011 flood (note that a larger area was flooded — including the road — at the peak). Hover over the image (or tap it if you are using a tablet or smartphone) to see the normal landscape. Flood imagery from Queensland Globe.

Two large duckbill valves waiting to be installed near Leybourne Street, Chelmer.

Two large duckbill valves waiting to be installed near Leybourne Street, Chelmer.

I see this place several times a week because I ride past it on my way to work. For some weeks now there has been excavation work going on in the channel on both sides of the road. There have been tractors, trucks, fences, signs, portaloos, the whole works. In this last week, the stars of the show have arrived: two enormous duckbill valves.

Duckbill valves seem to be the City Council’s favoured backflow prevention device, as they are popping up everywhere. There are two already at Milton just downstream from the Western Creek system, and just last weekend I noticed one at Merthyr Park at Newfarm. But the specimens I have seen so far are all modestly sized — certainly less than a metre across. The two at Leybourne Street are of another scale altogether. Each of them must be nearly two metres wide, and in their current position on the grass they stand more than two metres tall (even at full stretch I could not reach their tops). They are made of some sort of very hard rubber-like material — so hard that I have trouble imagining it yielding even to a torrent of stormwater. But this is exactly what these devices have been designed to to, or else they would cause local flooding every time there is a downpour.

A temporary dam at the outlet of the Leybourne Street drainage system, where the duckbill valves will be installed.

A temporary dam at the outlet of the Leybourne Street drainage system, where the duckbill valves will be installed.

Looking over the embankment towards the river, you can see the new concrete structure into which these duckbills will be fixed. Curiously, the area in front of it has been made into a small temporary dam. I’m not sure why.

These devices will bring some comfort to residents in this area who were flooded in 2011. But it is worth being aware of their limitations. If you look closely at the Queensland Globe flood imagery (best done via Google Earth rather than the image above), you will notice that there is mud all over the road at Leybourne Street. This is a telltale sign that the river broke its banks here, flooding the area directly rather than just through the drains. Indeed, the river spilled over the banks for much of the length of the Chelmer Reach. This means that the two duckbill valves, as big as they are, will not protect this drainage system from a flood as high as 2011 unless the bank itself is raised. Even so, they will hold off the flood until the riverbank is breached, and should stop most floods that are below the 2011 mark.

These are not the sort of devices that I expect to see installed at Western Creek, where to have any real impact a floodgate will need to fill up the large space around the footbridge under Coronation Drive. But as the Backflow Program proceeds, I will be curious to see if any duckbills appear that are bigger than these two.

A sticker showing the manufacturer of the two big duckbill valves.

A sticker showing the manufacturer of the two big duckbill valves.

The tip of the duckbill valve.

The tip of the duckbill valve. The rubber is thick and hard, but will open up when there is enough water pressure behind it. We hope.

UPDATE – 11 August 2014

The duckbills have of course been in place for some time now, but I only recently got around to checking them out. When I visited, there was a small but constant stream of water trickling through one of the valves, as you can see in this short video.

The installed duckbill valves. If you look closely, you can see some water trickling through the one on the right.

The installed duckbill valves. If you look closely, you can see some water trickling through the one on the right.