How to prevent building where it isn’t safe

The landslides in Muriwai in Tāmaki Makaurau have had devastating and cruel consequences and draw attention to very site-specific idiosyncrasies and vulnerabilities of the land beneath our feet.

Muriwai is an exceptionally beautiful and striking beach on the west coast of Auckland, and also susceptible to rainfall-induced landslides. There were two landslides in Muriwai in August 1965 after substantial rain (220mm of rain recorded at Whenuapai). Landslides occurred on consecutive days as “debris avalanches”, which means they were highly mobile; one eyewitness described a “a fantastic wave of mud” travelling at 90kph. The landslides formed above Domain Crescent, and flowed 250m downslope, destroying two houses. Those landslides killed two of four people holidaying in two baches.

According to an article in the New Zealand Geographer in 1966, heavy rain played a critical role in the landslides, but the tragic event was dependent on other factors, too: the slope angle, near surface geology (weak, ‘young’ sands overlying more competent, less permeable older bedrock), groundwater flow, and possibly interference with surface drainage. Of course, the location of the buildings on the slope was important in exposure to the landslide hazard. There are now several houses on a site that failed in 1965, suggesting that planning and zoning in Muriwai is not quite what it should be.

Rainfall-triggered landslides are a feature of New Zealand’s landscape erosion, and Auckland’s in particular, where the number of landslides that have been triggered by the recent rainfall probably numbers in the 1000s. However, the interplay between rainstorm events, the preceding (‘antecedent') rainfall in the days and weeks leading up to a storm event, as well as soil and bedrock properties, is complicated.

Auckland’s varied geology and geomorphology is fundamentally important to land instability. Auckland typically has weak soils formed by the weathering and erosion of underlying (often) weak rocks. It also has a lot of steep slopes. Even in their natural state, these slopes can be prone to failing if the soils become saturated enough. Some slopes are modified, becoming steepened for road cuttings.

A basic region-wide approach to determining landslide hazard is landslide susceptibility mapping. This is often based on LiDAR mapping (airborne laser scanning of the land surface) which provides a high-resolution digital elevation model (DEM), revealing the geomorphology, and whether landslides have occurred previously. A range of other factors are “layered” on top of this model, and statistically analysed, and relate to slope angle, soil and rock properties, and land cover parameters. The output, a landslide susceptibility map, can in theory then be used to inform any future planning decision-making.

Auckland’s clay-rich soils show high “shrink and swell” properties, meaning there is a natural annual cycle of wetting (swelling) and drying (shrinking), which can cause a progressive weakening of the soils over years and decades, called “strain-softening'.

Some areas also have layers of volcanic ash within the surface soils, from eruptions that happened a million years ago, that when wet is the texture of petroleum jelly. The ash layer creates an impermeable barrier, what geologists call a ‘perched water table’, which stops water from readily draining away. This makes the soil above the ash layer even more saturated, and more vulnerable to failure under persistent rain. Other west coast areas of Auckland are underlain by weak sandy soils, which are porous and highly erodible from surface water flow. So, permeability of soils and rocks (‘hydrogeology’) at a site is important too.

Further factors to consider relate to geomorphology. Although a slope may not have failed since European settlement, it may have failed in the recent prehistoric past. Such landslides may reactivate during either heavy rainfall, land disturbance activities and/or earthquakes. Because the surface soils and rocks have failed in the past, they may be at a lower residual strength, rather than at a (stronger) peak strength.

Irrespective of variations in the local geology and geomorphology, much of Auckland’s housing stock is built above, on, or below slopes, which, given our changing weather, are likely to be at risk. Auckland Council knows this and has recently commissioned an assessment of the areas of coastline that are susceptible to instability and erosion. This will provide the council with useful coastal information for future zoning in response to anticipated housing intensification. Moreover, the council’s proposed Plan Change 78 (publicly notified 22 August 2022) limits development within the coastal hazard areas to avoid increasing the risk of adverse effects on people and property. However, this does not include hazardous unstable slopes inland from the coastal strip.

An important way of mitigating rainfall-induced landslides in Auckland is the use of setback distances, the horizontal distance (on a plan) a dwelling is from the edge of a cliff or the top of a steep slope. The concept of setback distance is that it (in theory) lowers the risk of landslide hazard because it controls and limits where dwellings are located.

There are different methods for calculating setback distances, which are used widely overseas, and often specified in national building codes. Gravity, for instance, is fundamental to the driving force of a landslide so engineers can, using the slope height, estimate the area of land that might be at risk of landslide. Setback distances are used in New Zealand, although the exact method of calculation can vary, and I suspect may depend on the engineer doing the calculation. Many may use an off-the-shelf method of calculating such distances from a building code or standard, which doesn’t account for the myriad and often site-specific factors that put a property at risk of landslides. However, setback distance from the bottom of slopes is also important. This is because houses can be affected by ‘runout’ of saturated material mobilised from slopes above their houses, as we have seen in recent weeks.

Overseas, the insurance industry understands the importance of setback distances and acts on this. There are many examples of the insurance industry refusing to insure homes in the UK where it’s clear that in coming years or even a couple of decades, the house will be damaged by land sliding.

Without insurance, people are unable to get mortgages from banks to buy such houses. This means the houses are deemed uninsurable by the insurance industry, either purchased cheaply by cash buyers (who should realise the risk, but often don’t), or the houses are simply removed by local authorities after owners have vacated.

How do we plan to build to make houses more resilient to extreme weather events or indeed, earthquakes? Many may be becoming more familiar with the concept of “managed retreat”, which in New Zealand has typically applied to houses in low-lying areas on floodplains. The largest managed retreat programme in New Zealand was from the “Red Zone” after the Christchurch earthquakes. Through voluntary buyouts, the Crown acquired and demolished or removed more than 8,000 properties around the Avon River.

Internationally, managed retreat has also included elevated, landslide-prone zones, although usually limited to what we might call affordable housing. How the concept of managed retreat could be applied to hundreds or thousands of multi-million-dollar homes at risk of landslides in Auckland is problematic, as determining risk is an inexact science because of the many site-specific factors outlined above.

But if there is any good to come out of the storms this year it is that it will encourage local and central government planners to develop processes that ensure building consent means well-informed consent, that is based on site-specific calculations, not generalised basic slope geometry. This should take account of the ongoing effects of climate change, the geotechnical properties of the near surface materials and geomorphic history of the site.

That would lead to the development of more watertight policies and processes that prevent people from building where it isn’t safe to, that is compatible with the 100-year planning timeframe of, for example, the New Zealand Coastal Policy Statement. Even if that means getting in the way of some people’s views.