Received from Neil Johnstone
Wed, 3 May 2017 at 7:19 p.m.
Message: Last Thursday (27 April) I presented the remainder of my submission on Natural Hazards. Notes attached in case they might help anybody’s further efforts.
{The notes from Mr Johnstone are public domain by virtue of the consultative 2GP hearing process. -Eds}
****
2GP PRESENTATION NOTES: LANDSLIDES
Neil Johnstone
I have no property interest in any landslide hazard area (although I did previously), nor in the Water of Leith catchment, nor in South Dunedin. My main purpose in appearing at this stage is to bring to the panel’s attention that the expert (so-called) opinions received from Otago Regional Council’s (ORC) natural hazard analysts are often deficient to the detriment of the 2GP process and the city’s residents.
I am a long-term resident of Dunedin and am highly experienced in flood control issues and solutions. I am appearing here on my own behalf, therefore not strictly as an Expert Witness in this instance, although I have done so in past years both in both the High Court and the Environment Court. I also acted as lead technical advisor to the NZ Govt investigation into the massive 1999 Clutha flood. My detailed investigations have ranged from simple issues such as the Water of Leith (as Investigations Engineer at Otago Catchment Board and ORC) to the entire Clutha catchment (in varying roles). These investigations have often incorporated the construction and operation of accurate, properly verified models.
I am now semi-retired MIPENZ, but still running my own consultancy on a reduced basis. I am a highly experienced expert in flood issues, I am much less so wrt landslide identification and mitigation (but I know a nonsensical report when I read one). ORC hazard analysts responsible for the landslide buffer zones originally imposed across my former property (and many others) need to accept that their approach was seriously flawed, and far from expert. Paul Freeland has mentioned to me in a recent phone conversation that Dunedin City Council (DCC) should be able to have confidence that ORC hazard analysts are expert. I have no strong criticism of Mr Freeland, but those days have passed – in this region at least – when expertise was based on proven performance, and not on a position’s title. A property previously owned by my wife and me in Porterfield Street, Macandrew Bay was quite ridiculously misrepresented in ORC’s landslide report of September 2015. The landslide hazard zone on that property has apparently now been removed, but uncaring damage has been done to us, and no doubt to many others. The Hazard 2 zone was reportedly imposed without site inspection, or without anybody properly reviewing output or checking accuracy of references.
[Reason for submitting: Natural Hazards section of 2GP dominated (undermined) by ORC hazards staff input and DCC failure to verify/review; DCC presumption that ORC “experts” do/should have appropriate expertise. We appear to be witnessing a proliferation of Hazard Analysts in NZ Local Government with little relevant experience or skill.]
****
2GP PRESENTATION NOTES: SOUTH DUNEDIN
Neil Johnstone
The comments re South Dunedin flood hazard contained in my original written submission were written prior to DCC’s producing its inaccurate flood reports in respect of the South Dunedin flooding of early June 2015 in which high groundwater levels were held to blame. These DCC reports were eventually released in late November 2015 and April 2016 respectively. My analyses (well after my original submission) demonstrated that the prime cause of widespread flooding in South Dunedin was DCC’s failure (in order of probable significance) to utilise the bypass facility at Tahuna Wastewater Treatment Plant, to fully utilise its stormwater pumping capacity at Portobello Road, and to maintain its stormwater infrastructure (mudtanks etc). Inflow of “foreign” water from the St Clair catchment added to the depth of inundation in some areas. All these can be remedied by a diligent Council. Some have already been remedied, as positively demonstrated in the admittedly rather over-hyped rain event of the subtropical cyclone remnant around this past Easter.
ORC natural hazard analysts were probably responsible for the origin of the groundwater myth as a cause of the South Dunedin flooding in their Coastal Otago Flood Event 3 June 2015 report. Reference was made there to “elevated” ground water levels. They followed up with a contentious report (The Natural Hazards of South Dunedin, July 2016). This opens by stating that the June 2015 flooding was caused by heavy rainfall and high groundwater levels, with no mention of mudtanks, or pumping failures (plural). Such reporting cannot be treated as balanced, nor its authors credible. Elsewhere, ORC essentially conceded the groundwater myth in Rebecca Macfie’s excellent NZ Listener article entitled Flood Fiasco (June 11, 2016).
Shortly after, however, ORC produced the aforementioned South Dunedin Hazards report (backed up by an embarrassingly inaccurate video presentation) that seems to reflect a desire to preach doom rather than convey a balanced defendable scientific analysis of South Dunedin realities and solutions where needed.
One of the worst features of the report and subsequent video was the depiction of projected permanently inundated areas of South Dunedin based on ORC modelling of rising sea level effects. These depictions made front page news in the Otago Daily Times with flow-on reporting nationally. The mapped areas of inundation are actually taken from an earlier ORC report entitled The South Dunedin Coastal Aquifer and Effect of Sea Level Fluctuations (October 2012). The modelling was based on limited information, and the findings would therefore be expected to be of limited reliability. The 2012 report essentially confirms this, noting that modelling of existing conditions overestimates actual groundwater levels (by the order of half a metre in places). Figure 2 (Scenario 0) of that report shows significant permanent ponding for current conditions. None exists in reality. Almost lost (in Section 3.8) are the following (abbreviated, and amongst other) concessions:
• Uncertainty of input data
• Potential inaccuracy of model predictions
• High level of uncertainty
• Groundwater system is poorly to moderately well characterised
• Aquifer properties are poorly understood or quantified
• Each of these uncertainties could have the effect of overestimating the groundwater ponding in the current setting.
The reader is advised to read the full Section 3.8 to ensure contextual accuracy. In my view (as an experienced modeller), a study that cannot even replicate known existing relationships is imperfectly calibrated and unverified. It cannot therefore be relied on. Strictly speaking, it does not qualify as a model. The relationship between possible sea level rise and consequent groundwater impact remains highly uncertain.
Unfortunately, the 2016 ORC South Dunedin Hazards report (and video) chose to reproduce the 2012 ponding predictions using more recent data (but without any better appreciation of aquifer characteristics), but the predictions are similar. It is noted that no Scenario 0 mapping is included in the latter report, nor are the model’s inherent weaknesses described. No admission of the potential modelling inaccuracies is presented other than the following note in Section 4.1: “Further discussion of the original model parameters, model calibration and potential pitfalls is included in the ORC (2012a) report, which can be accessed on the ORC website”. I believe that all parties were entitled to know unequivocally that the modelling was unreliable and unverified.
The 2016 report also makes reference to the fact that dry-weather ground water levels at the Culling Park recorder are at or below mean sea level. This is attributed by the authors to leakage of ground water into the stormwater and wastewater sewers. If that is correct (I would reserve judgement as to whether there may be other factors), then we are witnessing just one example of how an engineered solution could be utilised to dissipate increasing depth of groundwater. Such solutions are canvassed in the BECA report commissioned by DCC several years back.
To summarise, South Dunedin’s exposure to flood (current or future) is poorly described by ORC hazard analysts. The 2GP process seems to have seen these analysts “adopted” by DCC planners as their experts. I consider that to be an inappropriate approach to the detriment of our citizens.
The proposal to require relocatable housing in South Dunedin seems premature, and based on highly questionable information. The proposal for relocatable housing in South Dunedin also rather pre-empts the currently-planned DCC study of overseas approaches to sea level rise solutions.
Requiring relocatable houses will likely simply mean that aged houses that should in time be replaced will be repaired instead. Who is going to build a new relocatable house if they have nowhere to relocate to and probably insufficient money to acquire the requisite land? The proposal to require relocatable housing is ill-considered and premature in my opinion.
With respect to ground water issues across South Dunedin, the 2016 Hazard Report presents –
The reason for my pointing out these facts is to encourage Commissioners to take a step back from the current hysteria surrounding South Dunedin. Had the 2015 flooding extent been restricted (as it should have been) to that which occurred in a slightly larger rainfall event in March 1968, the event would have already been forgotten. Seemingly, at least partly as a result of that hysteria, the proposal to require relocatable housing in South Dunedin seems premature, and based on highly questionable information. Just as ORC floodplain mapping contradicts its in-place flood protection philosophy, so does the proposal for relocatable housing in South Dunedin also rather pre-empt the currently planned DCC study of overseas approaches to sea level rise.
Requiring relocatable houses will simply mean that aged houses that should in time be replaced will be repaired instead. Who is going to build a new relocatable house if they have nowhere to relocate to and probably no money to acquire the requisite land? The proposal for relocatable housing is ill-considered and premature in my opinion.
****
2GP PRESENTATION: URBAN STREAM HAZARDS
Neil Johnstone
Urban Stream Comment re Leith and Lindsay Streams:
ORC’s mapping is said to be of residual flooding (post-flood protection works of the past 80-plus years), but actually represents what might have been envisaged many decades back in something considerably greater than the record 1929 flood with none of the very significant channel works of the 1930s, 1940s and 1960s; or even those lesser improvement of the 2010s in place. The ORC 2GP mapping includes areas that didn’t get flooded in 1923 or 1929. I agree with some potential dangers of stream blockage (especially in Lindsay Creek, and to a lesser extent at Clyde Street and Rockside Road), but one can only consider locations of feasible blockage in today’s conditions. Furthermore, accepted professional practice for flood plain mapping requires detailed hydrology, probability analyses, climate change allowance, hydrograph routing, in-channel modelling (allowing for stream capacity variability), and overland flow modelling. ORC’s flood mapping incorporates none of these fundamentals; instead, it reads as little more than a colouring-in exercise, when a professionally researched technical document is required. In short, ORC’s hazard analysts have carried out no fit-for-purpose analysis for a District Plan process.
Interestingly, the concerns expressed by ORC hazard analysts re channel blockage are entirely inconsistent with ORC’s own design philosophy and consent application evidence for the recent Flood protection scheme (so called). Design Philosophy minimises the issue.
Very briefly, the mapping is challenged for the following reasons (inter alia):
No descriptions of the effective flood protection initiatives (OHB -1920s and 1930s, DCC -1940s, OCB -1960s) are included. These works have ensured that overtopping is practically impossible in the George Street to Cumberland Street reach, the Clock Tower reach and Forth Street to Harbour reaches. Flood protection in these areas are all built to a much higher hydraulic standard than the so-called ORC scheme of the past decade, and to a far, far higher standard than existed pre-1929.
It is further noted that ORC’s own Design Philosophy Report (OPUS for ORC, 2005) for the proposed Leith/Lindsay flood protection scheme is adamant that debris traps recently (then) constructed at Malvern Street and Bethunes Gully would further mitigate any debris problems. Refer paras 7.7 and 10.6 of that document.
Ponding is mapped where water couldn’t even reach in 1929 (peak flood currently estimated at 220 cumecs, and predating flood protection measures) in the wider CBD area. Flows along George Street in the 1920s only occurred south as far as about Howe Street, then re-entered the river. Nowadays, the accelerating weir above George Street and the structural high velocity channel immediately downstream provide much more clearance than existed in 1929. [Most outflow then from the river occurred much further downstream.] In those downstream reaches, many of the bridges have been replaced or upgraded. Possible remaining points of interest are the hydraulically insignificant extension (circa 2015) of the St David Street footbridge, the historic Union Street arch footbridge, and the widened (circa 2012) Clyde Street road bridge. The flimsy St David Street bridge would not survive any hydraulic heading up so there would likely be of little flood consequence, and backing up upstream of Union St would be largely inconsequential because of the height of the Clock Tower reach banks immediately upstream. The Clyde Street bridge is acknowledged as being lower than optimum, but it has not created any issues in its half century existence. Any overtopping there could only impact on a limited area between the bridge and the railway line.
Overland lows beyond (east of) the rail line remain highly improbable because of the ongoing blocking effect of road and rail embankments. Flows as far as the railway station to the west of the rail line are also highly improbable nowadays as only the Clyde Street area could conceivably contribute.
The 1923 photograph showing ponding along Harrow Street is presented by ORC with an unfortunate caption stating that the water is sourced from the Leith. Some undoubtedly was, but the whole of the city was subject to “internal” stormwater flooding from Caversham tunnel, across South Dunedin to the CBD and beyond. To illustrate further, a NIWA April 1923 flood summary (accessible online) provides a summary of some of the information more fully described in technical reports and newspaper accounts, including:
• Portions of Caversham, South Dunedin, St Kilda, the lower portions of central and northern areas of the City and North East Valley were completely inundated.
• Water in South Dunedin was waist deep.
• The Water of Leith rose considerably and burst its banks in many places, causing extensive damage along its banks and flooding low-lying areas.
Today’s stormwater infrastructure is rather more extensive and effective (when maintained), and DCC has a continuing legal obligation to provide to maintain that service.
The levels plotted across Lindsay Creek seem highly pessimistic. Levels are shown to be of the order of 2 metres above North Road in some locations at least. I have [no] knowledge of any such levels ever having been approached. Care must be taken not to include unfloodable areas in the mapping. I don’t however discount localised channel blockage, and the channel capacity is substandard in many areas. The valley slope ensures that overland flow will achieve damaging velocities. Such velocities are noted in the NIWA summary.
Of greater concern to me, however, is that ORC’s mapping appears to have seriously underestimated the significance of potential Woodhaugh flood issues:-
The river channel through here is both steep and confined. The influences of Pine Hill Creek (immediately upstream) and Ross Creek (immediately downstream) add to turbulence and bank attack. The area was ravaged in 1923 and 1929, and there have been evacuations in some much lesser events in later decades. These areas are at considerable risk in a 50- to 100-year plus event. Hardin Street, Malvern Street had houses evacuated in the 1960s flood. High velocity, rock laden flows and mudslides can all be anticipated, and difficult to counter. Area below camping ground / Woodhaugh was overwhelmed in floods of the 1920s – a focus for flooding depth and velocity.
If the 2GP process is to include urban flood maps, these should be diligently derived, based on historical record and appropriate modelling. The mapping should reflect the real flood risks (including likelihood, velocity and depth). The decreasing flood risk from Woodhaugh (potentially high impact) through North East Valley (moderate impact) through to the main urban area south of the Leith waterway (localised and of little-to-zero impact) should be reflected in the mapping.
[ends]
—
█ 2GP Hearing Topic: Natural Hazards
https://2gp.dunedin.govt.nz/2gp/hearings-schedule/natural-hazards.html
█ For more, enter the terms *johnstone*, *flood* and *south dunedin* in the search box at right.
Related Posts and Comments
6.6.16 Listener June 11-17 2016 : Revisiting distress and mismanagement #SouthDunedinFlood
10.6.16 “Civic administration” reacts to hard hitting Listener article
—
[DCC Map differs from what was notified]
—
Posted by Elizabeth Kerr
This post is offered in the public interest.
What if? Dunedin... 