Tom Wilson
Lecturer in Hazard and Disaster Management
Dept of Geological Sciences
University of Canterbury
Acknowledgements :
David Bell (UC), Jarg Pettinga (UC), Mark Quigley (UC),
Graham Leonard (GNS Science), William Power (GNS Science)
Overview of Seminar
 New Zealand: our world class natural hazards laboratory
 Active geological landscape
 EQ, Volcanic, Tsunami,
 What information is GeoScience aiming to produce
 Lessons from the Canterbury Earthquake Sequence
 Success and limitations of geoscience
 Geoscience and Emergency Management
 What can be expected...
 What we expect from you...
1: New Zealand’s active geological landscape
Plate tectonics…
What happens at plate
boundaries?
New Zealand
tectonics and
Geology
New Zealand is the ‘ideal’
natural hazards laboratory!
“It does us a power of good to remind ourselves that we live on two volcanic
rocks where two tectonic plates meet, in a somewhat lonely stretch of
windswept ocean just above the Roaring Forties. If you want drama - you've
come to the right place“
Former Prime Minister Sir Geoffrey Palmer
Define each of these:
 What is a Hazard?
 What is Vulnerability?
 What is Risk?
What are Geoscientists worried about?
Geohazards:
•
Magnitude – how big is the event(s) going to be in space and time?
•
Frequency – when is the next event(s) going to occur? (tied to
probability)
•
Ensuring the scientific data is
used appropriately and
effectively
- Market base approach
(value for money)
- Return on investment
•
Geological disasters appear
to be increasing...!
•
Who is exposed and how
vulnerable are they
Some of challenges...
•
Geological systems are complex and operate in 4-dimensions
• 4D system but often can only see 2D
• Geological timescales are very different to human timescales
• Preservation potential
•
Geological disasters are commonly “surprising” – “didn’t fit the model”
• Severe, complex, cascading implications...
•
In the context of a society looking to manage risk, society is demanding
that geological risk be better managed
•
So...looking at ourselves – what can we do better?
• Risk Assessment: Increased societal demand, understanding and
access to technology has driven an increasing quantification of
geological risk
• Use of Geological Risk information: our job doesn’t stop once we’ve
produced the data
EARTHQUAKES
BODY WAVES (P & S)
P = primary (compressional)
S = secondary (transverse)
In rock P>6km/sec & S
≈50% of P
SURFACE WAVES
Include Rayleigh & Love waves
Travel at surface & not through
rock and thus slower
Can cause major damage to
buildings & infrastructure
Effects of Earthquakes
 TERRAIN
 Ground warping &/or rupture offset (≤10m)
 Triggering of landslides, rockfalls & liquefaction
 WATER
 Well level changes & seiche effects in lakes
 Generation of tsunami waves (eg Boxing Day
2005)
 STRUCTURES
 Structural damage (depends on design & age
etc)
 Failure of “lifelines” due to shaking or offset
 PEOPLE
 Loss of life or injury from building collapse etc
 Social disruption from loss of services
 Psychosocial effects
Earthquakes in NZ
Source: Barnes et al. (2011)
Earthquake
Probability of Occurrence
Using seismology and geology
5 Years
15 Years
Probabilistic
Seismic
Hazard Analysis
(PSHA)
Peak ground acceleration (-g)
expected at 10% probability
in 50 years
(g)
100 Years
From: Stirling, M.; Gerstenberger, M.; Litchfield, N.; McVerry, G.; Smith, W.; Pettinga, J.;
Barnes, P. (2008):
Updated probabilistic seismic hazard assessment for the Canterbury Region.
Volcanic Hazards
 Examples of volcanic hazards:
 lava flows
 lava domes
 pyroclastic falls (ash fall) and
pyroclastic flows
 lahars and debris avalanches
 volcanic gases
The volcano problem...
 Range of hazards
 Range of volcano types
 Range of different eruption magnitudes from one
volcano
 Range of different hazards produced within an eruption
Taupo 22.5 kyr
Eruption Sizes
( a range)
Probabilistic ash
fall map for
central North
Island for multiple
volcanic sources
(excludes Auckland)
Hurst and Smith (2004)
Tsunami
For effective management we need to know:
•
What sources affect NZ?
Geology, seismology
•
How big and how often?
Seismology
•
How high will the waves be?
Math. Modelling
•
What are the run-up consequences? Engineering
There is a level of understanding of all of these, but uncertainties
are still large.
Geological investigations and historical studies
provide checks on all of these.
Then we can evaluate how often for any given wave height
Distant tsunami sources for Napier:
Location and magnitude - the principal factors that determine impact in NZ
Estimated Maximum water level at Napier for Mw 9.0, by source location
Modelling by William Power (GNS Science) using NOAA (PMEL)
MOST programs. G Leonard slide (GNS Science) .
Local sources – Hikurangi plate interface
Many unknowns, many possible
scenarios with variations in
• fault length
• distributed movement along and up
the fault
•
•
•
•
fault rupture terminates at shallow depth
fault ruptures to surface along one or more splay
faults
fault ruptures to surface along the plate interface
hybrid of these
Effect of various rupture
scenarios on vertical seafloor displacement (and
hence the tsunami
generation capacity)
Geist (2005)
Local earthquake sources –
Hikurangi Margin plate interface
Slip distribution
(per 100 years of released
slip deficit)
Press to play
Tsunami modelling by William Power (GNS Science) in
2006 based on Wallace et al (2004 ) geodetic model of
plate locking beneath Wellington/Wairarapa.
Power et al. EQC funded Project 2006/521
This is just one scenario:
rupture up splay faults
releasing 400 years of plate
interface locking in a very
large earthquake (over
magnitude 8).
Global
Disaster
Trends
2010 + 2011???
Physical
Systems
HAZARDS
+
DISASTERS
Human
Systems
The Canterbury Earthquake Sequence...
It’s rattled more than just the pot-plants
 Isn’t New Zealand highly motivated at promoting
earthquake resilience?
 Why did so many people build on liquefiable land?
 Aren’t our buildings meant to stay standing during
earthquakes?
 I thought EQC had HEAPS of money for disasters?
 And now what happens now with Wellington?
Consequences – the numbers
 10,000’s homes damaged and >1,100 buildings demolished
 Reduced habitability and strained city’s ability to accommodate it’s
residents
 7,500 residential properties retired: ~5% of total housing stock
 Central Business District closed for months
 New Zealand’s largest short-term migration
 70,000 people evacuated: 19% of population
 3,000 long term migration: <1% of population
 8 millions tonnes of disaster waste: 40 years of Chch waste
 Total damage: estimated at >$20 billion (>10% GDP)
 EQC reserved exhausted and central government very tight (GFC)
 One of the greatest geotechnical disasters of the modern age
Lessons from the Canterbury
Earthquake Sequence
 Success and limitations of geoscience...
 What can we unpack from this?
 Very difficult situation
 Sustained earthquake sequence in a previously
seismically ‘quiet’ region
 Unmapped faults
 Sequence has migrated across an major urban area
 Most damaging EQ since the 1931 Napier earthquake
Seismicity for August 2010
To this…
No systematic increase in earthquake
rate or magnitude prior to Sept 4 2010
Aftershocks
Seismicity……..How many so far?
Numbers of Earthquakes in the Canterbury Region since 4 September
2010
Magnitude range
Number
7.0 and above
1
6.0 - 6.9
3
5.0 - 5.9
55
4.0 - 4.9
423 (+1 to June 24)
3.0 - 3.9
3143 (3024 on 31 May 2012)
This table was last updated on 19 June 2012
Source: www.Geonet.org.nz
Total now well over 10,000
recorded events
B Duffy using Geonet data
Aftershock Decay Sequence for Canterbury
(X10)
- Magnitude 3 and greater
Dec 23/11
M6.0
June 13/11
M6.3
Feb 22/11
M6.3
Source: www.geonet.org.nz
Sept 4/10
M7.1
Graph depicts the significant effect of the February, June and December 2011 earthquakes reenergising the Darfield aftershock sequence (Source: GNS Science).
Successes
 High quality seismological dataset
 Informed detailed forecasts and crustal strain models
 Significant input to geotechnical hazard and risk analysis &
planning
 liquefaction studies and rockfall/slope stability)
 Helped explain the phenomena to emergency management
and a “highly unsettled” public
 Hazard assessment worked. Correctly identified:
 Chch was exposed to seismic hazards
 Major potential for liquefaction
Surprises
 Earthquakes on unmapped faults
 Long duration of aftershock sequence, migrating
across the city
 High ground shaking experienced in Chch city
 Cliff collapse – Port Hills
 Transparency of analysis & decision making processes
This is not a blame game – it is about learning lessons for the future
Limitations
 Forecasting locating, timing and magnitude of earthquakes
– even after the sequence began
 We can’t predict earthquakes
 We can’t communicate our forecasts
 Hazard assessment was successful...but...
 Liquefaction and “blind” faults in Canterbury
 Why wasn’t this information better used?
 Our probabilistic analysis needed slight revision
 “Floating” statistical earthquake
This is not a blame game – it is about learning lessons for the future
Prediction vs. Forecasting
Prediction Requires:
 A time interval during which the earthquake is predicted (short-term =
hours to months, medium term = years to decades, long-term = decades to
centuries)
 An area within which the earthquake is predicted,
 A depth range for the earthquake,
 A range of magnitudes within which the earthquake is predicted to occur,
And, very importantly,
 The probability that the earthquake will fall within all these intervals, as
measured by past successes and failures of the predictive method.
 Unless all these requirements are met, the “prediction” cannot be tested
against the occurrence or non-occurrence of the earthquake, and therefore
the “prediction” is valueless
Mark Quigley & Jarg Pettinga - 2012
Probabilistic Forecasting Earthquakes:
• Short-term forecasting makes use of aftershock models
• Medium and long-term forecasting from geology,
GPS and seismicity modeling
Without Aftershocks:
Earthquake Sources in 3D
Background (Point)
Sources
• Constructing a seismic
hazard model
• Requires geological investigations
Seismological
vs
Geological
Fault
Sources
Earthquake
Hypocentres
Magnitude - Frequency
Distributions
Background
n/yr
M
+
Fault
=
Combined
Forecasting Earthquake Activity in Canterbury
• Short Term (hours to months)
• Medium Term (years)
• Long Term (decades)
Source: “Geonet”
Aftershock
Forecasting
Canterbury region extended probabilities:
Short-Term
Medium-Term
One month: 28 May – 27 June ‘12
One year: 28 May ‘12 – 27 May ‘13
Magnitude
range
Expected
range
Expected
average
Probability
Expected
range
Expected
average
Probability
5.0 - 5.4
0-1
0.19
17%
0-4
1.4
75%
5.5 - 5.9
0-1
0.05
5%
0-2
0.4
33%
6.0 - 6.4
0-1
0.014
1%
0-1
0.1
10%
6.5 - 6.9
0-1
0.004
<1%
0-1
0.03
3%
7.0 - 7.9
0-1
0.0015
<1%
0-1
0.01
1%
This table was last updated on 28 May 2012
Source: www.geonet.org.nz
Chch EQ Hazard
Assessment
 Fault sources which might
affect Chch had been well
identified
 Difficult to deal with
blind (unmapped)
sources...but:
 They had been theorised
 Some even identified
 Hindsight is a...wonderful
thing
We’ve known about
Christchurch’s liquefaction
hazard for a long time
(map compilation - 2005)
http://ecan.govt.nz/publications/General/solidfacts-christchurch-liquefaction.pdf
Feb 2011 Liquefaction
From Tonkin + Taylor with contributions from many
Unprecedented level of liquefaction across an urban
3: Geoscience & Emergency Management
Some reflections from the past 18 months:
 Most of us (geoscientists) have no idea how you
operate
 What is your role? What do you do?
 What language do you speak?
 CIMS... Isn’t that a computer game...?
 Research timescales vs operational demands...
 Geoscience research paper: 2-5 years: start to finish
Communication...
Communicating the right information in the
right form to the right people
•
Two ways to look at it:
• Politicians, emergency managers and the general public do not understand
uncertainty...
• Scientists do not communicate uncertainty well...
•
Successes (apparent or real) create the expectation that geoscientists
will get it correct, every time.
•
Reality is that geologic hazard and risk assessment is fraught with
uncertainty...
•
This MUST be adequately communicated:
UNDERSTAND your audience
Minimise JARGON
The audience is SMART, it’s just that they aren’t an expert like you
TRUST is easy to lose and hard to win back
Integration of researchers with emergency
management
“Science” (in broad terms) produces a product which is fundamental for
emergency managers
 WHAT HAPPENED??!! - Geotechnical + seismological
 Emerging risks (first-order geological and second-order sociological)
Challenges for researchers
•
Who did the research serve?
•
•
•
Operational emergency management demands?
Or more lofty research goals?
Who decides?


In the absence of a plan, it was often up to the individual researcher
4 Sept 2010 was a practise run for 22 Feb 2011 event. Did we learn our
lessons...???
Challenges for the researchers
•
Who did the research serve?
•
Information needs were/are complex and usually
transcend disciplines
Requires excellent coordination and leadership
between researchers and research teams, and with
stakeholders
•
•
•
•
•
•
Dynamic, tiring, conflicting, confusing conditions
High profile...which is a double edged sword...
Simple, binary answers are required by stakeholders
and the public...
But this is not how scientists are trained to
communicate...
Physical scientists and engineers have had to learn
about the ethics committee
Integration of researchers with
emergency management
 Inclusion of researchers with “recovery groups” seemed to be mostly positive
 Yet it didn’t happen in many cases (no pre-existing relationship)
 Research benefits

Increased situational awareness = could begin to anticipate information
needs...

Will inform the current and future research (this disaster, and the
next...)
•
Operational benefits

Had their own personal scientist

Often researchers had greater access to information and resources

Researchers have significant capability to offer
A potential solution?
Sarah Beaven (UC)
The desired state:
 improve trust and collaboration between researchers and emergency
management
 familiarize researchers with emergency management processes and
requirements
 demystify (and validate the value of) research and science for emergency
managers  and vice-versa
Earthquake response indicates value of:
 Continuing to develop the Natural Hazards Platform into a more networked,
transparent open structure, grounded in public domain
• Better integration with the CDEM model, using detailed strategies and
guidelines for research collaboration at all levels, providing for:
o Regular and specific training requirements with emergency services
o Protocols and procedures during emergencies
o Relevant local/regional organizations to initiate (& coordinate in
collaboration with GNS & others) research responses to events occurring in
their region
- reliance on local research expertise in first few days
• Potential to increase national research and emergency management
networking, and so adaptive capacity
Recap
 Geohazards driven by planetary scale forces (plate
tectonics)
 Hazards occur at a variety of scales in time and space
 New Zealand is located at an active plate margin
 We will have more disasters
 Lessons from Canterbury... difficult and limiations on
geoscience
 Expectations of Geoscience and from Geoscience need
to be known, established, dyanmic, transparent and
well-practised!