Claire Flashcards
(1) Hazard-Risk cycle
Hazard: anything that can cause harm Risk: How great the chance that someone or something will be harmed by the impact Vulnerability and resilience Exposure Value of potential loss Impact
• Risk is very complicated to calculate – essentially a calculation – effects vary on parts of society (infrastructure or welfare) and person to person (wealth, social class etc)
(1) What is a (natural) hazard?
“A dangerous natural phenomenon that may cause loss of life, property damage and disruption.”
UNISDR Terminology on Disaster Risk Reduction, United Nations, Geneva, Switzerland
(2009)
• Intensity and recurrence of hazardous events is key to the ensuing risk and impact
- An event that can cause harm
- Volcanos produce many types of hazards
(1) volcanoes pose a potentially ‘low probability–high impact’ event:
- Eruptions are unevenly spaced in time and magnitude:
- Problem of human memory – people are underprepared for the threat that they live close to - Ethiopia
- Unlike other natural hazards:
- The range of hazards is large e.g., lava, ash, flows etc.
- The hazard may continue for decades e.g., lahars
- The hazard may continue whilst the volcano is quiescent e.g., gas, lahar, landslide
- Effects continue after eruption – sometimes for years to centuries and longer – lahars that flow in rainy seasons – degassing that continues for decades
(1) Types of volcanic hazards
– Lava – Pyroclastic flow (PDC) – Ash fall / tephra – Dome collapse – Explosions / lateral blasts – Gas – Acid rain – Earthquakes – Flooding & jökulhlaup - – eruption under glacier – melts glacier and causes a torrent of water that causes flooding Secondary hazards – after or due to eruption - – can still happen at same time as hazard but is not directly related – tsunamis can cause death, but it was due to volcano – Lahar – Landslide / debris avalanche / sector collapse – Ash remobilisation – Tsunami – Evacuation – Roof collapse – Floods – Fires
(1) What is vulnerability?
“The characteristic of an element that makes it susceptible to the effects of a hazard.”
UNISDR Terminology on Disaster Risk Reduction, United Nations, Geneva, Switzerland
(2009)
• High vulnerability, and exposure, are generally the outcomes of skewed development processes
• Environmental management
• Demographic changes
• Rapid movement of people – due to commuting, war, migration, etc
• Rapid and unplanned urbanization in hazardous areas
• Failed governance
• Corruption means the scientific research does not reach the government
• Scarcity of livelihood options
• If hazard destroys crops, then farmers will be severely affected
• Related to susceptibility, fragility, weakness, lack of capacity etc
(1) What is resilience?
The capacity to cope, absorb, adapt or recover quickly from the effects of a hazard Human responses to volcanic threats are influenced by many factors which impact on societal resilience: • Culture • Belief systems – society cohesion • Education (level and about specific hazards/risks) • Awareness • Trust in experts/authority • Indigenous knowledge • Past experience • Self-reliance, mental fortitude • Good support networks • Good lifestyle – health, fitness, job • Ability to protect yourself • Preparedness
- Building design
- Urban planning
- Variety of livelihoods
- Food resources
(1) How to build resilience?
- Community resilience (and disaster prevention) can be improved through preparedness and hazard & risk mitigation:
- Disaster risk reduction – e.g., high bridges on lahar routes, location of key infrastructure away from hazardous paths
- Hazard mitigation – e.g., diverting lava
- Preparedness – e.g., community education, familiarity with potential hazards and warning systems/messages
- Emergency response measures – e.g., evacuation plans
- Sustainable livelihoods, poverty reduction, good medical facilities (‘safe hospitals’)
(1) What is exposure?
“People, property, systems and other elements present in the hazard zone that are subject to potential loss.”
UNISDR Terminology on Disaster Risk Reduction, United Nations, Geneva, Switzerland
(2009)
• If populations and resources were not located in (i.e. exposed to) potentially dangerous settings, no problem of disaster risk would exist
• Exposure is a determinant of risk but not a necessary one – i.e. it is possible to be exposed but not vulnerable.
• But, to be vulnerable, it is necessary to be exposed.
• Exposure risk relates to duration and intensity so must be assessed.
Differs from environmental concentrations because people can take protective actions to reduce their exposure (e.g. wear a facemask), whilst other actions might increase exposure locally (e.g. kids playing in ash)
(1) Exposure equation:
Intensity X Frequency x Duration
(1) Risk
- Key words: ‘chance’, ‘likelihood’, ‘probability’, ‘possibility’, ‘uncertainty’, ‘threat’
- Definitions:
- the probability that an adverse event will occur.
- the combination of the probability of an event and its negative consequences
- the likelihood that someone/ something will be harmed/lost by a hazard
- The uncertainty concerning the occurrence of loss
- Any threat which could have a positive or negative effect on an ‘objective’ or ‘element’
- A measure of the probability and severity of an adverse effect to health, persons, property or the environment.
- Effect: impact, damage, injury, liability, loss, harm
(1) Risk equations – depends on need
- Hazard probability
- Hazard probability x vulnerability
- Hazard probability x vulnerability x value of the element at risk (x exposure)
(1) The disaster risk equation:
‘The risk of disaster increases as the frequency or severity of hazards increases, people’s vulnerability increases and people’s capacity to cope with the consequences is decreased.’
• Hazard probability x vulnerability (± exposure)
Capacity to cope
(1) What is ‘impact’?
- We can also call it ‘effect’, ‘harm’ or ‘damage’
- To have a strong effect on someone or something.
- We can categorize impact in many ways:
- Primary and secondary impacts
- Acute and chronic impacts (immediate and long term)
- Proximal and distal impacts
- Cascading impacts (or ‘knock-on effects’)
(1) Impacts of volcanic hazards
• Immediate – Death – Injury – Damage to buildings – Damage to infrastructure/utilities • Roads • Water • Sewage • Electricity etc. – Damage to agriculture – Damage to aviation – Displacement of populations • Social problems • Vulnerability • Health – acute disease & epidemics • Famine – Longer term – Climate change – Chronic disease – Vegetation death
- Starvation - a secondary or even tertiary hazard
- 3 of the top 4 are secondary incidents
- Average number of fatal incidents per fatal incident is going down in time – better infrastructure – humanitarian aid – healthcare – hopefully means that 60,000 people wont starve again?
(1) Electricity infrastructure
Insulator Flashover: may occur with <3 mm of ash fall provided a significant portion of the insulator creepage distance (>50%) is covered in wet ash
Loading Damage: ash accumulation may overload lines, weak poles and light structures, and cause additional tree-fall onto lines. Precipitation will exacerbate the risk:
• Loading damage typically occurs with >100 mm ash accumulation
• Induced tree fall from ash load may occur with thicknesses >10 mm
Disruption to Control Systems: ash ingress into heating, ventilation and air-conditioning systems can block intakes leading to reduced performance, and affecting dependent systems:
(1) Relationship of damage and disruption
- Tephra (ash fall) causes little damage but huge disruption
- The remainder have a relationship
(2) Volcanic hazard assessments
• We need to answer the following questions:
– When and where did the volcano last erupt?
– How frequently has the volcano erupted in the past?
• These tell us about the return period i.e. when the volcano is likely to erupt again
– What sizes of eruptions occurred in the past?
• This tells us about possible magnitude
– What types of eruptions occurred in the past?
• This tells us about how explosive or effusive the eruption could be
– What hazards were produced and how far have they reached?
• This tells us about likely impact
– How long have eruptions lasted?
• Tells us the likely length of disruption
(2) Types of Hazard Assessment
Long-term assessment (likelihood of the future based on the past)
Short-term assessment (monitoring)
Dependent on:
• User of the assessment: civil protection, national government, international bodies e.g. UN
• Decision-making timescale
• Availability of data
The point – forecasting (not prediction)
to help to decide what to do in the future based on anticipated impacts
(2) Monitoring volcanos
• Volcano-monitoring techniques provide the data needed for issuing short-term warnings or predictions over timescales of hours, days, or perhaps weeks.
• Core methods
– Seismic
– Ground deformation
• Supportive methods
– Microgravity, electrical & magnetic studies
– Geochemical monitoring (gas and water)
– All detect and measure changes in the state of a volcano caused by magma movement beneath the volcano.
(2) Seismic monitoring
• Vital tool for monitoring and prediction
• Baseline monitoring and seismic array (multiple stations) essential
• >25 successful forecasts in 20 years
• Vital tool for monitoring and prediction
• Baseline monitoring and seismic array (multiple stations) essential
• >25 successful forecasts in 20 years
• Measure seismic waves produced by deformation through movement of gas/magma.
• Earthquake activity beneath a volcano almost always increases before an eruption
• Different types of earthquakes represent different physical processes
o e.g., harmonic tremor, fracturing, magma movement
- Volcano tectonic earthquake – cracking and fracturing
- Harmonic tremor – movement of magma up conduit
(2) Volcano deformation overview:
- Upward pressure from rising magma deforms volcano
- The ground can change shape by rising up, subsiding, tilting, or forming bulges
Taal 2020 deformation
• Usual deformation are mm scale over a few months/years
• At Taal it has been very rapid – the deformation has become a hazard itself
(2) Volcano deformation breakdown:
Tilt measurement: tiltmeter, ‘dry’ tilt
Lateral displacements: EDM (electronic distance meter), GPS
Vertical displacements: Precise leveling
Space-based: radar interferometry (InSAR), lasteraltimetray
Interferometric Synthetic Aperture Radar. Satellites record images of the Earth’s surface, and these images can be combined to show subtle movements of the ground surface, called deformation.
(2) Geochemical data
Determination of magmatic composition to gain insights into:
- Eruption style and magnitude
- Depth of magma chamber
- Rates of magma ascent
- Magmatic processes
(2) Hazard Assessment Methods
- Probabilistic / Stochastic – determine probability distributions for range of behaviours and outcomes.
- Deterministic - outcomes are precisely determined through known relationships among states and events, without any room for random variation e.g. scenario.
Probabilistic – determine all the possible outcomes
Deterministic – take one scenario and run with it
Volcanos are stochastic – but it is difficult to try to understand and plan for all outcomes
(2) Characterising uncertainty
Defining and addressing the ‘known unknowns’ and the ‘unknown unknowns’ (Donald Rumsfeld, 2002)
- Aleatoric uncertainty – irreducible, ‘unknowable’ random system - stochastic approach
- Epistemic uncertainty – reducible through knowledge, ‘unknown to me’ things we don’t understand yet - statistical approach
(2) Scenario-based approaches
Scenarios are deterministic (determined by previous events) and are used to identify a range of conditions / events / hazards / impacts on which statistical approaches and modelling can be based to generate quantitative outputs.
If identified with decision-makers, then scenarios map to scenario-planning activities within government.
Common scenarios include: worst case scenario and a most likely scenario approach and include characteristics:
• Magnitude / frequency
• Hazards
• Impacts e.g. threats to life, damage to critical infrastructure
(2) Volcanic eruption scenarios
- A small magnitude short-lived eruption producing lava flows and depositing ash locally, potentially affecting operations and quality of life
- A moderate magnitude eruption producing an eruption column up to 17km high (tropospheric), ash fall and gas emissions, ash dispersal disrupting transport
- A large magnitude eruption producing an eruption column ~27km high (stratospheric), column collapse pyroclastic flows posing a potential threat to life, ash fall and gas emissions, ash dispersal disrupting transport and evacuation routes
Laki eruption – 8 months in 1783. Holuhraun eruption 2014-15 – 6 months
Laki – risk of sulphur gas eruption, carried by wind, could kill many
(2) Probabilistic Approaches
- Probability of an outcome depends on a compound series of events all occurring.
- Frequentist
- Independent events with independent probabilities – defines event probability as the limit of frequency in a large number of trials, e.g. tossing a coin.
- Bayesian
- Series of events linked by conditional probabilities – reasonable expectation represents state of knowledge or quantification based on personal belief.
- Frequentist – frequency of observed results. Lots of data for similar events, can take an average and say that’s what’s going to happen in the future.
- Bayesian – you can change the probabilities as you learn more about the system. E.g. what is the likelihood of an eruption happening? The likelihood is X but once you have observations, e.g. ground deformation, likelihood changes to Y as more likely.
(2) Probabilistic hazard assessment
- Most volcanic systems are too complex and our understanding of them too rudimentary for precise, unequivocal predictions of eruptions.
- We can assess the likelihood of an event/series of events occurring.
- Simulations of single hazard models can produce probabilistic hazard footprints
- Probabilistic event trees: A graphical, tree-like representation of events - branches are logical steps from an event through increasingly specific subsequent events (intermediate outcomes) to final hazardous outcomes.
- Graphically displays all relevant possible outcomes of volcanic unrest in progressive detail
- Focus on the range of outcomes that could result from different combinations or sequences of events.
- Calculate the resultant probabilities of different outcomes
- Can be solved probabilistically, using discrete probability values and/or probability distributions
(2) Bayesian Event trees
- Useful logical frameworks for discussing probabilities of possible outcomes at volcanoes showing unrest or erupting.
- Each branch leads from a necessary prior event to a more specific outcome and is allocated a conditional probability.
- Probability is estimated through modelling for each potential outcome, using data (from the tools discussed in previous slides).
(2) Expert elicitation
• Expert elicitation is a method for quantitatively characterizing the state of knowledge about an uncertain quantity through eliciting expert judgement.
• Fills a gap where there are no data or data are highly uncertain.
Aims:
1. To produce statistical probability distributions that describe the uncertainty around frequency-magnitude relationships
2. Discuss assumptions for modelling hazard ‘footprints’ (areas of inundation - spatial & temporal)
(2) Hazard maps
- Answers questions:
- What type of activity may occur?
- How far will effects reach?
- Where are the safe areas located?
- Each map is unique:
- takes into account volcano behaviour, local topography, inhabited areas etc.
- Critical for planning:
- Emergency procedures
- Long-term land-use
- Individual risk
- Limitations:
- Boundaries must be regarded as approximate and conservative
- Cannot predict some variables e.g. eruption duration, vent location, changes to landscape by eruption
(2) Global Volcanic Hazard Assessment
Volcanic Hazard Index - An index-based approach involves assigning scores to a series of indicators, which are then combined to give an overall hazard score.
• Scores are combined to give a volcano-specific hazard score using:
[eruption frequency × (‘frequent’ characteristics of volcano’s eruptions)] + extreme characteristics
• This can be expressed in terms of the indicators as:
[frequency status score × (modal VEI + PF score + mudflow score + lava flow score)] + maximum recorded VEI
Where the score for each hazard is defined by the potential for fatalities
Population Exposure Index - total population for circles of radius 10, 30 and 100km of each volcano.
Hazard Levels plotted with PEI for each volcano provides a visualisation of a risk matrix and Risk Levels I, II, III with increasing risk shown by the warming colours.
(2) Global Volcanic Hazard Assessment
Overall threat in a country obtained by equation:
Overall threat = mean VHI x number of volcanoes x population within 30km
Top 5 countries with highest overall volcanic threat (% represents the country’s threat as a % of total global threat)
(3) Eruption forecasting (prediction)
- Most volcanic systems are too complex and our understanding of them too rudimentary for precise, unequivocal predictions of eruptions.
- Eruption prediction is short term. It often relies on application of physical laws governing the failure of materials which includes an observable, monitored component such as seismicity, gas emissions or ground deformation.
- Extrapolation of a fitted line (linear or curved) of monitoring data on an ‘inverse-rate’ plot, to a pre-determined intercept.
• A combination of methods should be used, rather than relying on single techniques.
• Can be validated with real-time data.
• Not fool proof
– Mt St Helens
– Lateral blast
– Not predicted
• Probabilistic forecasting also used
– Bayesian event tree models for eruption forecasting (BET-EF)
– For long-term planning and crises
– Merges all types of info from monitoring to theoretical models
(3) Warning systems
- Warning systems are vital for telling the population about the hazard and associated risk.
- Warning message carries results of authority’s risk decisions to citizens.
- People tend to deny danger, or that any deviation from routine is necessary
- Messages must be precise, leave no room for interpretation.
- Mostly done through an alert level system …
(3) Volcanic Alert Level Systems
- So … scientists’ decisions to change an alert level have great consequences for the people affected by the action and often involve social and political factors that scientists feel obliged to take into account.
- RESPONSIBILITY may be beyond their level of expertise and beyond their mandate.
- Papale believes that volcano science should be the only focus for volcanologists during a crisis. Decision making for societal benefit is a political action.
- The move from one level to another does not allow for uncertainties to be expressed.
(3) Mitigation (reducing vulnerability)
the action of reducing the severity or seriousness of something
Can be primary, engineered interventions:
• Diverting lava flows
• Cooling lava with water sprays
• Channelizing rivers down which lahars may flow
• Artificial draining of crater lakes (water!)
Can be secondary interventions:
• City planning (e.g., building hospitals away from valleys)
• Building housing with roofs resilient to ashfall
• Strengthening and increasing the height of bridges over lahar-prone rivers
• Removing populations from the at-risk area
Everything we have talked about so far is a form of mitigation
• Hazard, risk, vulnerability, exposure assessments and communication of key findings of those studies
(3) Preparedness
Planning – Areas to be evacuated – Evacuation routes – Shelters – Emergency hospitals – Lifelines (utilities – water, electricity etc.) – Infrastructure – Ash clearance and dumping Communications • Warning messages to go with alert levels • Education
(3) Communication
• Communication is CRUCIAL in saving lives
• Education of local communities and public awareness.
• Important to engage with community leaders who can disseminate information e.g. priests, doctors, local government.
• Example:
– Nevado del Ruiz, Colombia, 13 Nov. 1985
– Deadly lahars killed 23,000 in Armero 72 km from summit
– Caused by melting snow
– Scientists warned authorities … who failed to take action
– Police informed after lahar began … no warning system in place
– Scientists now believe they did not communicate effectively.
– Political problems too of sending in the army to evacuate an area threatened by guerrillas.
Case Study:
Kīlauea, Hawai’i
The 27 June 2014 Pahoa Flow
(3) Response
Evacuation & exclusion
• If scientists can effectively predict an eruption, exclusion zones not needed
• Sufficient lead time necessary to implement warning system for evacuation
NOT AN EASY DECISION!
• Economic expense of mass evacuation is high
• Displacement of populations leads to psychological disorders such as stress, depression etc.
• May also lead to epidemics, starvation and violence.
Decisions on evacuation
- Easiest in small communities
- Potential economic loss is small
- Little blame if false alarm
- Hardest in densely-populated regions
- Evacuation process is logistically complex
- Economic burden is great
- False alarms are disastrous
- ‘Mental calculations’ on whether to evacuate, based on little science, leave the decision makers open to litigation if wrong decisions are made
- A cost-benefit analysis is a rigorous, systematic and transparent method
- Woo, Natural Hazards (2008) 45:87-97 lays out how to conduct a probabilistic analysis https://link.springer.com/article/10.1007/s11069-007-9171-9
(3) Risk perception
• Risk perception studies (social surveys) can appraise success of public education effort:
• Whether people have knowledge of hazards/risks
• Whether people have knowledge of emergency plans
• Whether people know how to prepare
• Whether there are other risks that they consider to be more important
• Whether people trust the agencies responsible for the plan
• Excellent example:
Barberi et al. Volcanic risk perception in the Vesuvius population. JVGR (2008) 172, 244-258. https://www.sciencedirect.com/science/article/pii/S0377027307004179
(3) Risk management cycle
- The work we do can be represented in the context of the well-known risk management cycle.
- The data/knowledge base and monitoring underpin everything.
- We have a number of international collaborative projects that are taking a holistic approach to hazard and risk through preparation, early warning, response, forensic analysis and communication (eg STREVA part of the NERC ‘Increasing Resilience to Natural Hazards’ initiative; FUTUREVOLC an EU FP7 project focusing on the highest risk volcanoes in Iceland)
- If communication and preparation throughout the cycle are effective then real-time response when an eruption begins (green star) should also be effective….see next slide.
