HIGH FLYERS THINK TANK

Extreme Natural Hazards

University of Melbourne, Tuesday 30 October 2007

Introduction

The purpose of these proceedings from the High Flyers Think Tank is to contribute to the understanding and treatment of emerging extreme natural hazard risks affecting Australia within a global context. The Academy's view is that researchers from organisations and universities, senior public and private sector decision-makers, and end users could be better informed to anticipate change. They could thereby respond to inherent risks more coherently and effectively.

It is hoped that such advances will ultimately lead to better preparedness for extreme disasters; minimising damage to environments, communities and infrastructure while fostering better public confidence in Australian risk governance. 'Governance' in this context refers to all levels of government – local, state and federal – as well as that of research, forecasting and emergency services organisations such as CSIRO, Bureau of Meteorology, Bureau of Rural Sciences and Geoscience Australia.⁵

In this report it is recognised that effective communications between all levels of governance is critical since effective dialogue between high-level government agencies and local agencies, including community representatives, is crucial for dealing with extreme natural hazards in a timely manner. Coupled with this framework, is a need for adequate resources to be delivered to the government levels responsible for enacting decisions, collecting data, monitoring, and further supporting decision-making processes.⁶

Recently, there have been some significant changes in the political perception of the natural environment. 'Government' is seen increasingly as responsible for failures in instances where public vulnerability thresholds have increased and there is a reduced tolerance for the impact of natural hazards. Government is also more often seen as responsible for anything denominated as a 'failure' with respect to either mitigating natural disasters or ensuring that recovery is as rapid as possible. This has introduced problems with how to measure success with respect to reducing vulnerability. For example, what represents a 100 per cent success rate for disaster recovery? Anything less than 100 per cent is regarded as not good enough by the electorate and governments are increasingly being held responsible.⁷

The Think Tank discussions, in the following sections, outline the need for research to better understand the causes and early warning indicators for different extreme natural hazards, to assist with preparedness and timely responses.

The discussions also support earlier findings from other more specialised research from the field of extreme natural hazards.^{8 9} For example, a recent study on natural catastrophe causes, trends and risk assessment found that the particular challenges and requirements which extreme events present could be summarised under hazard and risk assessment, risk assessment/vulnerability and risk control. In this context it is notable that the global insurance industries are key stakeholders and have already made considerable investments in this area for their own risk assessment purposes.

These findings also support the Think Tank PPRR matrix for examining extreme natural disasters in terms of prevention, preparedness, response and recovery (Figure 1).^{10 11}

The Think Tank specifically focused on the theme of PPRR as elements of risk assessment and management. While it should be recognised that this simplified framework departs from that given in the Australian Standard on risk,¹² it is a useful method for identifying the stages that are involved in dealing with extreme natural hazards.

Figure 1. Think Tank outcomes matrix (Click on image for a larger version)

The following descriptions are supplemented by a recent report by the Academy, The changing risk environment: ideas for a new Australian policy framework for handling risks.¹³

1. Prevention of emergency situations (sometimes also called mitigation), involves taking measures to eliminate or reduce the incidence or severity of emergencies. For hazards, seven major control interventions are possible: 1. modify wants; 2. choose alternative technologies; 3. prevent initiation of events; 4. prevent releases; 5. restrict exposure; 6. block consequences; 7. mitigate consequences. Within this framework, solutions to risk are not necessarily technical – sometimes it can be the increasing of public understanding or trust to accept the risk and what is being done about it.

2. Preparation for emergency situations involves taking measures to ensure that communities and services are capable of coping with the effects of emergencies. Keys to effective preparedness are: 1. sound analysis so that the risk manager knows what is being prepared for; 2. early detection of hazards before they become disasters; 3. a comprehensive survey of vulnerabilities where the nature of the risks cannot be foreseen, and steps taken to reduce those vulnerabilities; 4. effective response strategies able to cope with any disaster situation; 5. advance planning of strategies to help affected communities recover.
   
   The important point about preparedness, perhaps the overriding point, is the need to conduct solid planning before the event to know what to do during an event. For example, there was little planning for the April 2007 Solomon Islands tsunami, which caused a major disturbance in Queensland and was a humanitarian disaster in the Solomons. So there is a need for adequate planning before hazard events actually happen, even for events which do not actually cause a catastrophe.
   
   Forecasting is important but there are questions regarding when and where. As an example, during the 1999 Sydney hailstorm an earlier alert would have helped. Car yards could have moved cars under cover, which could have reduced losses. In terms of human safety, there could have been fewer injuries if there had been more warning. However, for things such as roof damage, there would be little benefit because the design of buildings couldn't be altered within a short time.
   
   There are many examples where preparation has significantly improved resilience. For example, 'Goyder's Line' which declares a zone beyond which cropping should not occur helped those who adhered to this rule avoid disastrous crop failures.¹⁴

3. Responding appropriately to emergency situations involves taking measures in anticipation of, during and immediately after emergencies to ensure that effects are minimised. Required elements are: 1. good integration of all the different forces – fire services (urban and rural), ambulance, State Emergency Service and police. Agreed command systems need to be in place well in advance so that there is no need for negotiation after an event has taken place; 2. availability of specialists to cope with particular hazard situations, such as HAZMAT teams trained and equipped to cope with exposures of hazardous materials; 3. personnel available who are trained.
   
   With regard to response, there is a general problem with communicating warnings to the public in that false alarms can give rise to complacency. There are practical problems in creating physical communication networks that can remain functional throughout various emergency conditions. Again, it is uncertain how people in Australia would actually respond to warning systems (there is a need for such social science studies). It would also be very useful to make quality spatial data available in real time, so people can respond better, and work out where the damaged areas are located.

4. Recovering from emergency situations involves taking measures that support emergency-affected individuals and community efforts to reconstruct physical infrastructure, as well as restore emotional, economic and physical wellbeing. Factors to be considered are: whether insurance covers damage and losses suffered; whether compensation may be claimed; whether (and under what conditions) government loans may be justified; and whether counselling services are required and who should pay for their provision.¹⁶

Participants at the Think Tank recognised the value of the PPRR framework but also identified the need to combine some of these elements with the broader approach of risk management. The following quote encapsulates their point:

The terminology used in hazard and risk language is not yet standardised and so it is widely acknowledged by researchers that terms are often used inconsistently. This reflects the fact that practitioners and researchers come from a wide range of disciplines.

Thus a special focus of dealing with extreme natural hazards necessarily lies in the definition and applicability of the terms natural hazard, vulnerability and risk – especially as they apply to planning practices. Risk concepts are so complex that their application in planning continues to be analysed and improved as a major subject of research. Therefore it is necessary to carefully define terms and, for the purpose of this report, key terms are used as follows (mostly following a Guidance Note series¹⁸).

These definitions were considered by participants to be appropriate for this report because they focus on information that planners and other stakeholders require when discussing natural hazards. They also necessarily involve a broad understanding of hazards and their potential impact on human developments which are considered vital to the discussion of mitigation and adaptation measures.

While such definitions are useful, it must be recognised that all definitions can inadvertently constrain the possibilities identified as being available. For instance, some problems may be exaggerated while other important factors may be completely missed. Therefore the present language used by researchers and hazard managers to describe and discuss natural hazards will need to become more refined and standardised in the near future.¹⁹

One trend that may influence the use of this language is the increased application of risk-based models or resilience models as overarching frameworks.²⁰ An advantage of such frameworks of risk is that they help focus the debate about extreme natural hazards on social responses, preventative actions and preparedness for events – since danger is something individuals attempt to control.²¹

There are at least two kinds of risk: (1) measurable uncertainty (probability); and (2) unmeasured uncertainty (where numerical probabilities may not apply). The key difference relates to ambiguity of information. Scientific risk is relatively quantitative, objective and unambiguous whereas lived risk is relatively qualitative, subjective and highly ambiguous.²²

Infrastructures, environments, population densities and societal resources are some of the components of risk frameworks that influence the degree of potential harm. In such a context, 'risk' is considered as a measure of this constantly changing potential.^{23 24}

One problem with some of these frameworks is that many environmental risks exceed simple, mathematical measures. They instead involve complex problems and feedbacks.²⁵

One type of risk-based approach that may avoid some of these difficulties is capabilities-based.

All of these tools would benefit from an improved integrating philosophy. One area that has much potential involves concepts of building resilient ecosystems. This is because these concepts explicitly acknowledge the potential for internal conflicts for resources and competing demands within systems. For example, agricultural systems have conflicts between natural and social elements. Within these frameworks, many disciplines of research can work towards examining socio-ecological resilience phenomena to build around objectives and community targets with the types of characteristics that people need and want in resilient systems. Resilience reflects the degree to which the system can build capacity for learning and adaptation.²⁷

This capacity for learning has sometimes been overlooked. For example, scenarios for managing extreme natural hazards do not always account for adaptive responses; the Y2K scare is an example of an adaptive response to a potential breakdown in critical infrastructure.

Adaptation and mitigation should not be considered polar opposites; for things like climate change, Australia will need to carry out both measures. For example, when responding to the problem of the ozone hole, people carried out mitigation by reducing CFCs in the atmosphere, and adaptation – by 'slip, slop, slap' to reduce personal exposure to UV radiation. This two-pronged approach is also required for extreme natural hazards.

At the Think Tank, participants discussed scientific, social and technological options for investigation and implementation with the objective of better dealing with extreme natural hazards in Australia. While many possibilities and recommendations were identified, participants noted that most of the options would need to be implemented in close consultation with the community. This would give the community confidence to accept decisions by emergency workers and to not feel distant from the decision-making processes.²⁹

At the time of the Think Tank, participants observed that anxiety among public and scientific research communities was compounded by concerns about the increasing speed of technological change, the burgeoning volume of data and pace of development. Then there were additional difficulties seen to be faced by governments, regulators and others involved in risk governance. These are associated with fully assessing, controlling and dealing with hazardous events and their associated risks (Figure 2).

Figure 2. Losses from great natural disasters (far exceeding 100 deaths or US$100m in losses), 1950–2005. Figures are adjusted for inflation.30 (Click on image for a larger version)

Participants questioned whether outliers and spikes should be considered important when examining hazard and risk or loss curves. For instance, should attention be given to events that occur 1 in every 50 years? Or should research attention concentrate on the 1 in 5 or 1 in 10 year events? In terms of bushfires, Australia seems to be coping much better with the smaller events, but still suffers large losses from the less frequent 1 in 50 year events.

They recognised that in temporal terms, there are examples of extreme natural hazards that have a relatively slow onset to occurrence, such as droughts and associated pest outbreaks. However, there are also other events that have a rapid onset, including natural hazards like landslides, tsunamis and tornadoes.

At a spatial level, they recognised that these events can occur in largely uninhabited areas with most damage inflicted on natural ecosystems and agricultural production, and in many cases with limited economic or community impact. However, these events may also occur in urban areas with immense impact on large numbers of people. Similarly, hazards may affect small areas of the continent (a tornado in Lismore) or occur over extensive areas (drought over the Australian continent).

Therefore Think Tank participants identified that time-sequenced maps of natural hazard occurrence and risk are useful tools for policy-makers and international development organisations. For example, the World Bank 'Hotspots study' combined hazard occurrence and severity, population and economic data, and loss data to derive relative risk estimates.^{31 32} The Hotspots study estimates global relative risk using proxies based on the geographic distribution of population density and economic productivity sampled on a latitude and longitude grid.