The Puysegur Trench Quakes of 30-9-2007 – Part 2

How They Were Managed

Two large earthquakes in the Puysegur Trench on the evening of Sunday 30th September 2007 required a number of New Zealand government agencies to carry out their responsibilities for assessing, reporting and managing the risk to life and property.

As the events of the evening unfolded, the New Zealand public was left to rely on government agencies in Australia and the United States of America for up-to-date information on the earthquakes and the associated tsunami risk. New Zealand has been in this situation before, when a strong earthquake near Tonga on the 4th of May 2006 highlighted serious tsunami information mis-management issues at the Ministry of Civil Defence and Emergency Management.

This is how the events of the evening unfolded.

The largest of the two earthquakes struck without warning in the Puysegur Trench at 6:24 p.m. on Sunday 30th September 2007. The magnitude 7.3 quake was located undersea 475 km south-west of Invercargill. Less than a minute later residents of Southland, Otago, Fiordland and South Canterbury experienced a moderate level of shaking and suspended objects were seen to sway.

The shaking was not severe but five of the 296 reports since filed by the public with GeoNet were categorised at level 5 on the Modified Mercalli scale which, in New Zealand, represents the following conditions: “Generally felt outside and by almost everyone indoors. Most sleepers are awakened and a few people alarmed. Small objects are shifted or overturned, and pictures knock against the wall. Some glassware and crockery may break, and loosely secured doors may swing open and shut.”

By 6:30 p.m. visitors to the GeoNet website were able to see the first traces of a very large earthquake which registered on all of the 15 seismograph traces which are updated every 5 minutes on the seismograph drums page.

As the 6:40 p.m. update of the drums page became available it would have been obvious to an experienced website visitor that a large earthquake had occurred somewhere in the south of New Zealand, with the traces for Denniston (erroneously), Kahutara, McQueen’s Valley, Wanaka, Otahua Downs and Wether Hill Road still showing varying degrees of ground tremor. No report on the details of the quake was available on the GeoNet website.

Within 15 minutes of the earthquake, the US Geological Survey website had posted an initial assessment of the earthquake, which was remarkably close to the final values determined for location and depth, and accurate as to final magnitude.

An assessment of the tsunami risk was prepared by the U.S. National Oceanic & Atmospheric Administration’s Pacific Tsunami Warning Centre (PTWC) at 6:38 p.m. and was posted on their website within a minute. The warning centre’s evaluation was that no destructive widespread tsunami threat existed, based on historical data. However, the bulletin pointed out that, “earthquakes of this size can generate local tsunamis that can be destructive along coasts located within a hundred kilometres of the earthquake epicentre. Authorities in the region of the epicentre should be aware of this possibility and take appropriate action.”

PTWC bulletins are usually compiled using standard stock phrases, so the bulletin issued gave no indication of the risk to coastlines 400-600 km from the epicentre which would include Stewart Island and the Southland and Otago coasts.

During the 35-odd minutes following the earthquake Radio New Zealand National, a party to the memorandum of understanding between the Ministry of Civil Defence and broadcasters for the provision of timely authoritative messages in an emergency, was broadcasting pre-recorded programme material.

At 6:57 p.m. a wave gauge at Dart, Australia recorded a 4 cm change in sea level. This was reported a few hours later by the PTWC.

At 7 p.m. Radio New Zealand National’s hourly news and weather segment did not mention the earthquake or any activity associated with it. The broadcaster’s continuity announcer returned to presenting another two pre-recorded programmes which ran until the 8 p.m. news bulletin.

Had a significant tsunami been generated, the first wave could have been close to the Southland coast at about this time. Tsunami waves travel at about 700 km/h in deep ocean, but their speed drops as they encounter shallower coastal waters, so it is not a simple calculation of 475/700ths of an hour for an arrival time at Invercargill, 475 km away from the earthquake’s epicentre, but an approximate arrival time of more than 41 minutes after the earthquake would apply under typical oceanic conditions.

A check of the GeoNet website showed there was still no report of the Puysegur earthquake, and the MCDEM website had no information on the event at all.

At 7:35 p.m. the Ministry of Civil Defence and Emergency Management (MCDEM) appears to have prepared a media release in which it stated “there is nothing yet to suggest that a tsunami of any size has been generated.” Bearing in mind that the first wave of a tsunami would already have arrived at southern coasts of the South Island by this time, this was a valid observation.

But the next sentence showed that the ministry was dithering again. “The GNS Science convened tsunami expert panel and NIWA scientists are monitoring and assessing information about the earthquake and sea level movements.” At more than 70 minutes after the quake, the panel should already have met, assessed and acted.

It is not known when this release was actually issued, as the two media releases from the ministry were not made available to the public until they magically appeared in the Media Releases/Speeches section of the MCDEM website late morning or early afternoon the following day, Monday.

Continuous monitoring of Radio New Zealand National showed that no change had been made to its scheduled programme between 7 and 8 p.m. and no announcements were made. Just before the 8 o’clock news, the GeoNet website was checked to see whether a report on the earthquake had been posted, but nothing had.

Radio New Zealand National’s news at 8 p.m. reported the earthquake, and a sound byte from Dr Ken Gledhill, the GeoNet Project Director indicated that he had stepped into the void left by the National Institute of Water and Atmospherics and the Ministry of Civil Defence (the organisations responsible for tsunami events) by stating that there was no indication of a tsunami. A media release from MCDEM was not mentioned.

Sometime between 8:40 p.m. and 9:14 p.m. GeoNet finally added information about the earthquake to the Recent Quakes list on its website.

At 9:00 p.m. Radio New Zealand National provided more detailed information on the earthquake, some information on the response to the event, and GeoNet advised that 70 people had reported feeling the tremor.

A check of the MCDEM and NIWA websites showed that no information on either the quake or tsunami risk had been posted.

At 9:25 p.m. a sea-level change of 10 cm was detected by a wave gauge at Spring Bay in Australia.

At 9:31 p.m. the wave gauge at Port Kembla, Australia reported a sea-level change of 11 cm.

A second media release from MCDEM prepared at 9:45 p.m. (about 2½ hours after the first tsunami wave would have reached Southland coasts) was more definite, being headed “Auckland Islands earthquake no tsunami risk to New Zealand.” However, the release also stated “The tsunami advisory for Southland and Otago has been withdrawn.” It is not clear how this advisory was originally issued and to whom, as it did not appear in the earlier media release.

Once again, this media release was not made available to the general public until it magically appeared on the MCDEM website late morning or early afternoon the following day, Monday.

At 9:40 p.m., according to The Domion Post of Tuesday 2nd of October 2007, a 250 mm rise in sea level was recorded at Dog Island near Bluff.

Radio New Zealand National’s news at 10 p.m. repeated the 9 p.m. content regarding the earthquake and related tsunami risk.

At 10:14 p.m. the Pacific Tsunami Warning Centre issued a second and final bulletin on the first of the Puysegur Trench earthquakes confirming that a tsunami wave had been generated, and included the readings from the Australian wave gauges.

At 10:48 p.m. a second strong, shallow earthquake struck the Puysegur Trench. This magnitude 6.8 quake was located 235 km west-north-west of Auckland Island (about 35 km away from the earlier event) at a depth of 10 km.

Within a minute, residents of Southland, Otago, Fiordland, South Canterbury and Christchurch city had felt the tremor from the quake located 505 km south-west of Invercargill. More than 700 reports would be lodged with the Geonet website over the next 12 hours, with more people being sensitive to earthquake activity following publicity about the earlier event. The second quake has also attracted slightly more damage reports, with 11 people from Queenstown, Dunedin and various locations in Southland reporting MM5 shaking.

When the US Geological Survey website was checked at 11:16 p.m. an accurate report of the quake’s details was available. The magnitude would be revised downward to 6.6 by about 11:32 p.m. but it was changed back to 6.8 after further analysis.

The Pacific Tsunami Warning Centre issued a bulletin at 11:01 p.m., reporting the magnitude 6.8 quake and concluding that a destructive widespread tsunami threat did not exist. The bulletin also contained the usual warning that local tsunami could be destructive within 100 km of the epicentre.

By 11:27 GeoNet had posted a report on the second event on its website. It is worth noting that the US Geological Survey assessment of the epicentre puts it further south than GeoNet’s which reports the epicentre to be 440 km south-west of Invercargill. This will be due to the different information available to the two organisations and will, no doubt, be corrected as they retrospectively analyse the events with data from more sources.

Had a tsunami been generated by this second event, the approximate arrival time of a wave at the Southland coast using GeoNet’s epicentral location would have been sometime after half-past eleven.

Radio New Zealand National’s midnight news bulletin (more than an hour after the event) made no mention at all of a second quake in the Puysegur Trench.

Neither NIWA nor MCDEM have issued public media releases on the second event.

[Compiled from data supplied by the US Geological Survey and its contributing agencies, the U.S. National Oceanic & Atmospheric Administration’s Pacific Tsunami Warning Centre, the Geonet project and its sponsors EQC, GNS Science and FRST, and Geoscience Australia.]

Next – Part 3: Problems and Solutions
A look at how the organisations involved failed to keep the public adequately informed and some suggestions for improving performance.

Leave a Reply