Columbia University Center for Hazards and Risk Research

1
Columbia UniversityCenter for Hazards and Risk
Research

• Overview
• CENDIM/CHRR URM Workshop
• October 25-26, 2001

2
Center for Hazards and Risk Research

• Organized April 2001
• Part of Columbia Earth Institute
• Based at Lamont-Doherty Earth Observatory
• Partners include The Wharton School, Bogazici
  University, London School of Economics, others.

3
Organizing Themes for Hazard and Risk Studies

• Science in service of Sustainable International
  Development
• Reducing Vulnerability / Building Resiliency
• Methods and Mechanisms of Risk Assessment and
  Risk Management
• Technological and social measures of risk
• Multiple Hazards, Aggregate Risk
• Interactions, triggers, and amplification
• Human activities that generate risk

4
Organizing Themes for Hazard and Risk Studies
(cont.)

• Community Knowledge System
• Assessments of Predictive Skill in ST
• Assessment and communication of risk
• Assessment and communication of uncertainty
• Understanding decision strategies and mechanisms,
  and their inputs and outcomes
• Feedbacks for integrated studies
• Appropriate use of technology

5
What Is the relationship between Hazard and
Risk?

• Hazard is a process which has potential human
  impacts.
• Risk is a function of both hazard and
  accumulated human assets.
• Concentration of assets contributes to
  quantitative measures of risk.

Source USGS, CIESIN
6
Risk is a Cultural Quantity

• Different societies have different asset
  exposures
• System effects can amplify the evaluation of risk
• The study of risk mixes hazards (physical
  science) and vulnerability (engineering, social
  science)

7
Technical Risk vs. Social Risk

• Asset census and Asset fragility
• Integration over many scales
• Complexities due to interdependencies not
  included
• Not normative
• Measured in replacement and repair costs (or loss
  of use)
• Useful for cost-benefit analysis
• Risk transference as an economic cost

• Measures social and cultural disruptions
• Measured in risk tolerance in relation to a
  self-defined state of well being
• Highly normative
• Social and cultural interdependencies critical
• Useful for choice analysis
• Issues of risk transference harder to quantify,
  and hinge on moral arguments

8
Prediction and Predictability

• New science is available to improve
  predictability of events and their impacts
• New science is both empirical and model-based

9
Stress climate is a manifestation of plate
tectonics, but simple models of deformation are
surprisingly good at predicting seismicity.
(source J. Deng, Ph.D. Thesis 1996, Columbia)
10
Research Model
Basic Research
Physical Science/Engineering
Social Science
Integrated Research
Risk Assessment
Risk Management
Extension/ Applications
11
Source K. Boyer
12
Illustration K. Boyer
13
Center Programmatic Components

• Basic Disciplinary Research
• Integrated Research
• Research Extension
• Training Extension
• Graduate and Other Formal Education

14
Improving Resilience

• Emergency Response
• Hardening Built Environment
• Hardening Social Constructs
• Political
• Economic
• Cultural
• Alternative Development Paths
• Responsible Planning
• Altered States

15
Single Hazard Model
Hazard
Impact
Risk
Action
Knowledge areas incomplete, research required
16
Single Hazard Model
Tectonic Stress Model
Fracture Mechanics
Earthquake Process Model
Fault Topology
Event Genesis
Wave Propagation Model
Anelastic structure
Near-surface structure
17
Single Hazard Model
Ground Motion
Structural Design
Impact Process Model (Built Environment)
Soil-Structure Interactions
Structure Dynamics
Primary and Secondary System Response
18
Single Hazard Model
Categorization
Tolerance Levels
Risk and Risk Perception
Asset Concentrations (space and time)
Resiliency
19
Multiple Hazard Model(1st Generation)
20
Multiple Hazard Model(2nd Generation)
21
Multiple Hazard Model(3rd Generation)
Integrated Risk System
22
Improving ResilienceKnowledge System Approach

• Product orientation widely used to match
  scientific output to end-user needs.
• Product design at provider level often assumes
  end-user knowledge is capped.
• Elevating community knowledge base will produce
  higher-level outcomes
• User-mediated science and technology outputs

23
Third Generation Multiple Hazard Information
Technology
Action
Provider/User Community Interaction Model
Integrated Hazard System
Integrated Impact System
24
Community Interaction Model

• Links observational, modeling, simulation,
  prediction science
• Links science to technology (engineering
  solutions)
• Links solutions to risk management community
• If near-real-time, links response and emergency
  management communities.

25
Community Feedback Loops

• Feedback between sectors important for modulating
  basic ST output.
• Knowledge base kept high by combination of push
  and interactive technologies.
• Heightened curiosity hastens application of new
  ST advances in the field.
• Move beyond virtual knowledge product generation.

26
IT Requirements for Implementation

• Spatial data integration
• Monitoring and near-real-time data assimilation
  of time series
• Data QC, preliminary analysis, archiving,
  management for research and products
• Physical descriptions Characterizations and
  models
• Modeling and simulation codes and results
• Scenario building, description, and dissemination
• Data integration from other components
• Servicing decision pathways and community
  interactions
• Mitigation planning
• Emergency response
• Capturing feedbacks
• Formal and informal education
• Community outreach
• Products (push, interactive, user-mediated)

27
Education and Outreach

• Certificate and Degree Programs
• Student and investigator exchange
• Integrated Project studios
• Professional awareness
• Public/political awareness
• Identifying stakeholders
• Links with digital media.

28
Highlights TurkeyIncreased earthquake risk in
Istanbul

• The 1999 Earthquake Sequence in Northwest Turkey
  points to the high likelihood of even more
  damaging earthquakes in the Marmara Sea
  (Istanbul) region, during the next few decades.
• The Center for Disaster Management (CENDIM) of
  Bogazici University and Columbia are initiating
  an alliance in research and education dealing
  with earthquake hazard risk reduction. Will
  extend Columbias current scientific work in the
  Marmara region and expand into structural and
  socio-economic risk assessment and risk
  management activity.

29
From Parsons, Tom, Toda, Shinji, Stein, Ross S.,
Barka, Aykut, Dieterich, James H., Heightened
Odds of Large Earthquakes Near Istanbul An
Interaction-Based Probability Calculation Science
2000 288 661-665
30
Fig. 1. (A) Stress change caused by earthquakes
since 1900. Shown are the maximum Coulomb stress
changes between 0 and 20 km depth on optimally
oriented vertical strike-slip faults (44). The
assumed friction coefficient is 0.2, as has been
found for strike-slip faults with large
cumulative slip (45, 46). A 100-bar deviatoric
tectonic stress with compression oriented N55W
(47) is used, under which optimally oriented
right-lateral faults strike E-W except along the
rupture surface. The 1993 to July 1999 seismicity
recorded since installation of IZINET (7) has
uniform coverage over the region shown.
Calculated stress increases are associated with
heightened seismicity rates and with the future
epicenter of the 17 August 1999 Izmit earthquake
(indicated by star) sites of decreased stress
exhibit low seismicity. (B) Izmit aftershocks
are associated with stress increases caused by
the main rupture first 12 days from IZINET (7),
such as the Yalova cluster southeast of "Y," and
the occurrence of the 12 November 1999 Düzce
earthquake. Faults Y, Yalova P, Prince's
Islands M, Marmara I, Izmit.
From Parsons, Tom, Toda, Shinji, Stein, Ross S.,
Barka, Aykut, Dieterich, James H., Heightened
Odds of Large Earthquakes Near Istanbul An
Interaction-Based Probability Calculation Science
2000 288 661-665
31
From Parsons, Tom, Toda, Shinji, Stein, Ross S.,
Barka, Aykut, Dieterich, James H., Heightened
Odds of Large Earthquakes Near Istanbul An
Interaction-Based Probability Calculation Science
2000 288 661-665 Figure 4. (A) Observed and
modeled transient response to stress transfer.
The 13 M 6.8 North Anatolian earthquakes for
which the stress at the future epicenter was
increased by 0.5 bars are plotted as a function
of time. The earthquake rate decays as t-1 in a
manner identical to aftershocks, as predicted by
(29-32).
(B) Calculated probability of a M 7 earthquake
(equivalent to MMI VIII shaking in greater
Istanbul) as a function of time. The probability
on each of three faults is summed (43). The large
but decaying probability increase is caused by
the 17 August 1999 Izmit earthquake. "Background
tracks the probability from earthquake renewal
"interaction" includes renewal and stress
transfer. Light blue curve gives the probability
had the Izmit earthquake not occurred.