地震可能对建立在盆地的遥远城市造成破坏——现在我们了解了原因

Sedimentary basins – depressions in Earth’s crust caused by tectonic activity – tend to be flat and are favoured places to build cities. But during earthquakes, they can become natural resonance chambers.

Just like sound waves echoing around an empty hall, this means seismic waves can get trapped in these basins and bounce from side to side to create “seismic echoes”.

Depending on the shape and depth of the basin, these echoes can become amplified and be highly destructive.

New Zealand’s capital city Wellington is built on a sedimentary basin and is an example of this phenomenon.

In the 2016 magnitude 7.8 Kaikōura earthquake, Wellington’s central business district experienced shaking that exceeded design predictions. Even though the quake was located 80 kilometres from the city, it caused severe damage to infrastructure, with many multi-storey buildings damaged or destroyed.

Archival records also show that during the 1942 magnitude 7.3 Wairarapa quake, which was located about 80 kilometres north of Wellington, some 10,000 chimneys were destroyed in the city.

Our new research provides an updated model for the central Wellington basin.

We found it is almost twice as deep (about 500 metres) than previously thought and that its shape is significantly different from the previous model. These differences go some way toward explaining why the shaking was stronger than expected.

The deadliest example from history

Historically, the most devastating example of seismic echoes, or trapped waves, was the 1985 Mexico City earthquake, which killed 8,000 people and destroyed high-rise buildings.

The quake’s epicentre was 350 kilometres west of the city, but when waves of moderate amplitude arrived in the city, they became trapped in the low-wave-speed sediments of the basin on which it is built and became amplified.

As the waves bounced from side to side, they created a standing wave, similar to water waves in a bath.

As a result, the city experienced specific narrow zones of extreme destruction, underscoring the risk from even very distant earthquakes for cities built on sedimentary basins.

Seismic echo chambers

Seismic waves become trapped and amplified for two main reasons.

First, as the waves move from a fast wave-speed medium (solid basement rocks) to the low wave-speed of sedimentary rocks, the amplitude of the waves will increase to compensate for the drop in wave speed.

This is similar to a tsunami wave that travels across the deep ocean with a small amplitude but high speed, and then, as it comes closer to shore in shallow water, slows down but dramatically increases in amplitude.

This interchange between wave amplitude and speed is to do with the energy of the waves, which must remain constant.

The second reason for the amplification of seismic waves in a basin is resonance – when the wavelengths of the incoming seismic waves are similar to either the vertical and horizontal dimensions of the basin.

If the basin has steep sides, an edge effect is also generated where strong amplification can occur close to the edge of the basin due to a buildup of different wave types.

Perhaps the most surprising of our findings is the shape of the basin under Wellington. Its effective western edge is not the Wellington Fault, as previously assumed.

Instead, the edge cuts across the basin at a high angle to the Wellington fault and follows the line of two previously identified, low-activity faults – the Terrace and Lambton faults.

These differences between the new and old basin models have significant impacts on the predicted shaking Wellington might expect. In particular, there will be an effect linked to the newly described edge, and the predicted amplification will be higher for a deeper basin.

We used a 3D model of the basin in a computer simulation for the shaking at frequencies of 0.7 Hertz (the dominant shaking frequency recorded during past quakes) . We found that the amplifications of horizontal ground motion could be 2.5-3 times the background level adjacent to the western edge of the basin.

When we compared this predicted pattern of amplified shaking to where the actual damaged buildings were located during the Kaikōura earthquake, we observed some correlation with the western edge of the basin.

However, we need to be cautious when making this comparison as this pattern could be linked to other factors, such as the distribution of reclaimed land and clustering of inadequately designed buildings.

Our study nevertheless highlights two key points.

First, simple geophysical methods can now be used in urban areas to map out the depth and shape of basins that cities are built on. From these models, we can then generate computer simulations to predict the location of amplified shaking.

This will lead to more granular zoning for what parts of cities may be more vulnerable.

The second key point is higher awareness of the risk to cities built on sedimentary basins from not only local, but also distant earthquakes.

Timothy Stern has received funding from the NZ Earthquake Commission (EQC) for the initial gravity survey; QuakeCore for some of the seismic work; and from an internal Victoria University grant for student stipend support.