Keywords

Disasters happen with appalling frequency in our world, resulting in death, injury, destruction, disruption and economic loss that can set back the development efforts of affected countries by decades. Disasters from natural hazards are growing at a rapid rate for several reasons. Population growth, migration and urbanisation are all leading to the poorest, most vulnerable, people occupying land that is more exposed to severe and frequent threats, while urbanisation and migration are also producing communities that are both more dependent on infrastructure and lacking in local knowledge of their environment. At the same time, climate change is increasing the severity and frequency of many of those threats. In 2015, the world community met together to sign the Sendai Framework for Disaster Risk Reduction (UNDRR 2015), aimed at building a safer world for everyone.

The signing of the Sendai Framework was a critical step towards mobilising resources to counter these trends towards more frequent, more costly disasters. Much of its content deals with planning and policies that will reduce exposure to hazards, especially in new developments, and reduce vulnerability in existing ones. However, threats will always occur that exceed the protection in place, and the Sendai Framework also promotes the critical role of early warnings in enabling people to survive and recover from disasters. Since the cost of disasters can severely set back progress in development, the implementation of more effective responses to weather-related hazards is a theme in many of the Sustainable Development Goals, also agreed by the world community in 2015.

The objective of this book is to save lives and livelihoods and to reduce injury, damage and disruption from weather-related hazards in all parts of the world by helping those who create policy and plan, design and operate warning systems to make use of the latest research on what makes a good warning, so that their warnings may be more effective in our rapidly changing world. The material for this book has been gathered as part of the World Meteorological Organisation’s High Impact Weather (HIWeather) project under the auspices of the World Weather Research Programme (Golding et al. 2019).

While the frequency and severity of threats are growing for the reasons given above, the ability to avoid or reduce their impacts is also growing as a result of scientific research and technical innovation. Weather forecasting has seen spectacular advances in prediction accuracy in the last half century, with 5-day forecasts now more accurate than 1-day forecasts were then. These improvements have come as a result of technical achievements in computing and satellite observation as well as from the application of new science. The ability to gather and communicate information has always been at the heart of weather forecasting and warning, but the revolution in mobile communication of the past 20 years has enabled warning messages to reach a much greater number of people, even in remote areas of developing countries, so that there is a much greater awareness of the approach of weather-related hazards.

We have written this book for professionals and trainees who are involved in setting policy and planning, implementing and operating warning systems or parts of them. They may be in central or local government, in emergency management, in the management of businesses or utilities, in international aid agencies or in community response groups. The material is also a suitable introduction for those planning research on aspects of the warning chain especially those undertaking transdisciplinary research across the physical and social sciences.

In this book, we refer to anyone who acts on a warning as a decision-maker . They may be an individual acting to protect themselves by deciding whether to evacuate or to postpone a journey. They may be a manager responsible for the staff, customers and plant of a business. They may be an emergency manager responsible for the safety of a community. Or they may be a government minister responsible for the safety of a nation. Each has different levels of responsibility and will take different decisions in order to exercise the power they have been given.

This book is focused on the production and use of warnings. We distinguish a forecast , which produces information about the future state of the weather or some other aspect of the environment, without consideration of its use, from a warning which provides information about a threat so as to enable a response. The response will be different according to the lead time, the confidence level, the severity of the threat, the vulnerability of those threatened and other factors. For instance, a long range, low confidence warning may be used to initiate training or other “no regrets” preparatory activities, whereas a short range, high confidence warning may be used for more costly responses such as closing down a factory.

A decision-maker is a user of a warning but may also be a producer of a warning for someone else. For instance, an emergency manager in a city will receive a warning and initiate city-wide activities to protect citizens from the threat. At the same time, they may themselves issue a warning to citizens, advising them to take specific action. Similarly, a weather forecast centre in the path of a storm may need to respond to warnings issued by emergency services in order to protect its staff and to maintain operations.

We characterise the production of warnings as a value chain whose aim is to provide the information that enables the best decisions to be taken, both by individuals and by those with responsibility to protect others. In a perfect warning chain, the warning received by the end user would contain precise and accurate information that perfectly met their need, contributed by each of the many players in the chain. In real warning chains, information, and hence value, are always lost as well as gained at each link in the chain. In business, the term “valley of death” was coined as a metaphor of the failure of research to lead to successful innovation. NRC (2001) adopted this to represent the failure of research to translate into operational weather forecasting improvements. In Fig. 1.1, we use it more generally to represent the failure of the expert information generated in warning organisations to lead to the desired responses due to inadequate communication along the warning chain. The height of each mountain may be interpreted as the maturity of the expertise available for use in weather warnings. Successful communication of information from one contributor of expertise to the next is represented by spanning the valleys with bridges, whose height can represent the success of the communication between those contributors in avoiding the loss of information. Without a bridge, there is no communication, and the expertise of a particular contributor is completely lost. This representation of the warning process is, of course, a gross oversimplification of reality. Real warnings are created from a complex web of interactions taking place continuously among a wide variety of people more or less involved in the core activities shown. At the same time, distinct activities may in some cases be combined in a single person. There are also professionals whose expertise lies in being one of these bridges. Nevertheless, the concept is a useful one that highlights the very broad range of disciplines involved – and the need for those disciplines to communicate with each other effectively.

Fig. 1.1
figure 1

The valleys of death concept of a warnings value chain. (© Crown Copyright 2020, Met Office)

Full size image

This conceptual value chain can be read in either direction – there are no arrows on the diagram! Indeed, it is important that it is read in both directions. In designing a warning system , the starting point must be the decisions that need to be taken to protect life, property, infrastructure and livelihoods. These should be the basis for deciding what information is required and how it should be delivered – sometimes referred to as the “first mile paradigm”, because the user is at the start of the process. These decisions will vary according to the hazard that is being responded to, the person making them and the environment in which that person sits. In some cases, they will be so distinctive that a specific tailored form of warning is needed for that person. In other cases, a generic form of warning may need to be designed to meet the common requirements of a variety of people. The need to take a decision demands specific information, which can be traced up the value chain to define the expertise and resources required to produce it. However, if we do this without considering the capabilities of the upstream contributors, the warning system will certainly fail. Not only are there limitations to what science and technology make possible, but there may be capabilities available that enable more effective decisions to be taken that hadn’t been thought possible. Thus, design of a warning system is an iterative process, starting from the decision-maker, progressing up the value chain, then continually returning to the decision-maker in a process of mutual adjustment. On the other hand, when the warning system is in operation, the flow of information is predominantly down the value chain from producer to user – and this needs to happen quickly, accurately and reliably to maintain the value of the information. Nevertheless, the return flow of information is still crucial, providing updates from those involved “on the ground” so that the warning producers can maintain their situational awareness. Since this book is aimed particularly at those designing a new or improved warning system, we adopt the first mile paradigm for the ordering of the chapters – starting with the decision-maker and moving up the value chain to the information producers.

In this book communication and partnership are key words. Communication takes place between people, between institutions and between data systems. Limitations in any of these can inhibit the effectiveness of the warning system of which they are a part. Institutional communication is particularly important in creating an environment for successful partnership. At the highest level, government is responsible for creating the legislative framework that facilitates partnership working between organisations. However, while a good governance framework is necessary, it is not sufficient in itself. Shared personal knowledge of the aims, culture and language of the partners is critical to building the trust that enables outstanding performance when the threat is real.

Warnings of weather-related hazards have a long history going back to the foundations of national weather services in the late nineteenth century, largely in response to the implementation of the telegraph as the first telecommunication network, enabling instant communication of observations and warnings. The first applications were in maritime safety, at a time when most propulsion was by sail and shipwrecks were commonplace. With the growth of aviation in the early twentieth century, the extreme sensitivity of early flying machines required great care to avoid dangerous weather conditions. As a result, much of the modern structure of weather services was put in place to support the safety of aviation. Moving to the second half of the twentieth century, the massive growth in personal transport led to increased requirements for warnings of adverse road conditions. Weather prediction advanced rapidly through the application of satellite observing and computer prediction. At the same time, rapidly expanding populations in locations exposed to hazardous weather generated the need for a wider range of warnings on land, particularly of storms and floods. These developments exposed limitations in the ability of weather services to meet the needs of decision-makers. Whereas a sailing ship’s captain would know what the impact of a gale force wind would be and an aircraft pilot would know the safe visibility for landing at his destination airport, users of this wider range of weather warnings were less likely to understand how the predicted hazard would affect them. More recently, the ability to get warning information to people has advanced rapidly with the availability of dedicated radio and television services and the mobile phone. As the range of warnings has extended into aspects of the weather that are inherently less certain, it has become more important to communicate the confidence that the hazard will occur at the location and severity predicted. Taken together these challenges have led to the development of a range of new warning paradigms incorporating impact and risk as well as hazard. While not yet adopted universally, these newer types of warning will become the norm over the next few years and their core capabilities are described here, drawing on the latest capabilities in probabilistic weather and hazard prediction at ever finer geographical scales. With an increasing diversity of responsibility in responding to warnings, the study of how people receive and react to a warning has become a critical input to warning design, and educational programmes have been developed to grow familiarity in the warnings and the desired responses. However, weather services and emergency managers have often lagged behind commercial business in applying the science of behavioural psychology to help ensure positive rather than negative responses to messages. This is now changing, as more weather services set up social science groups to advise them on warning design. As a result of these changes, an ever-increasing range of expertise is being brought to bear on warning production. While some weather services continue to employ the full range of expertise in a single organisation, it is increasingly recognised that the achievement of a critical mass of expertise in these newer disciplines may be best achieved through the building of partnerships between complementary organisations.

The justification for implementing warnings as part of a risk reduction strategy is that (a) the cost of a warning system is much less than most other risk reduction options and (b) the benefits to society far outweigh the costs. When considering the benefits, there is often a focus on economic benefits, and these are certainly important. In the most-developed countries, economic costs of weather-related hazards are large and increasing, due to growth in hazardous events due to climate change, growth in economic vulnerability due to increased wealth and growth in exposure due to the spread of populations into more hazard-prone areas. The relatively small number of deaths and injuries remains important, however, as the cost of these to society is high. The impact of indirect hazard impacts on people’s health, well-being and productivity is hidden but of increasing interest to researchers and potentially a significant additional cost. By contrast those affected in the least-developed countries have much less wealth to destroy, and hazard impacts are primarily measured in fatalities and injuries. Thankfully these are reducing, though they remain far too high. However, the smallness of the economic losses hides the fact that a person who loses an uninsured house is destitute regardless of whether their house was worth $10 or $10 million, and such losses need to be measured against an appropriate comparator such as a country’s GDP, to understand their significance.

Fig. 1.2
figure 2

Aspects of a working partnership. (Source – Rob Honch, ECCC)

Full size image

This book is distinctive in the literature on warnings, in that it concentrates on the partnerships connecting experts across the “valleys of death” (Fig.  1.2). The need to build partnerships is not unique to effective warning systems. Consideration of how they work in other contexts can provide useful pointers to those attempting to make progress in warnings. There is a whole literature on building successful business partnerships (Rosen 2007; Morten 2009; Swientozielskyj 2016), which is increasingly being applied more widely (de Bruijn and Tucker 2002, Bang and Frith, 2017, Stibbe and Prescott 2020, Bucher et al. 2020, WISER 2020). The lessons learned there are equally relevant to the production of warnings (Golnaraghi 2012). Every partnership has a formal and an informal aspect. The formal provides the legal basis that ensures the informal can proceed without the danger of being derailed by accidental ignorance or misunderstanding. Non-disclosure agreements, memoranda of understanding, contracts, etc. all have their place, enabling the respective organisations to be open with each other. Part of the formal process must also be regular reviews that challenge the justification and effectiveness of the partnership and, if necessary, dissolve it without delay. However, the real work lies in building the relationship that these represent, and that takes time, patience, lots of listening, strong leadership and hard work. A successful partnership requires that each partner understands the culture and can speak the language of their counterparts. They must know their organisational viewpoints and aims. Ultimately, a partnership works if the partners trust each other, something that can only be achieved by actively working together over a long period of time. It can easily be interrupted by changes of personnel, so must be actively maintained, both when it is going well and when it is not. It is in the pressurised situation of an incipient disaster that a failure of trust is most likely to surface – with potentially fatal results.

In the main part of this book, each chapter describes one of the bridges in our conceptual value chain (Fig. 1.3). It first describes the expertise and methods of those at the decision-maker’s end of the bridge, highlighting the information they need to make their contribution. It then moves across the bridge to investigate the constraints and limitations of the provider of that information. Then the nature of the bridge itself is addressed, identifying the characteristics that inhibit communication and showing how these may be overcome through the building of strong partnerships. Each chapter includes some examples of real partnerships that have addressed these issues, the challenges they encountered and the outcomes that were achieved. Finally, the key points for success are summarised for reference.

Before proceeding to the core chapters of the book, which are organised according to the “first mile paradigm”, described above, Chap. 2 describes an effective risk management framework in the context of societal drivers of risk and the responses in the United Nations 2030 agenda, introducing the main components, the roles of the main actors and the need for evaluation. We introduce some key definitions and emphasise the contribution of early warnings within this framework.

In Chap. 3 we explore the challenges of achieving a level of risk perception, in each decision-maker, that is commensurate with the most cost-effective action while being consistent with the warning producer’s capabilities. Firstly, we look at the evidence for how people respond to warnings and how the nature and delivery of the warning affect that. Then we look at the aims of the person providing the warning, the constraints within which they must act and the judgement process by which they decide when the level of confidence needed for a warning is reached. Then we address the connection between the two, provided by the delivery of the warning, and how a partnership between warner and receiver can produce a more effective response.

Chapter 4 looks at the range of actors who produce warnings in the public and private sectors, the sources of information they draw on to comprehend the nature of the hazard, its impacts and the implications for those exposed and the process of drawing that information together to produce a warning. We consider the wide range of experts who provide the tools to assess the impacts of the predicted hazards and the challenges and limitations of these tools and the information they produce. Then we look at the diverse ways in which these tools need to take account of the way their outputs will feed into warnings and the nature of the partnerships that can facilitate this.

Chapter 5 focuses on translation of the hazard into its impact which is at the heart of current efforts, within the WMO HIWeather project and elsewhere, to improve the effectiveness of warnings by incorporating impact information into the warning process. At the same time, it presents some of the most difficult and demanding challenges in contrasting methodology and language. In general, the hazards we are concerned with can be described well by repeatable processes that may be couched in mathematical language. By contrast, their impacts depend on the social and economic characteristics of the communities that they affect; repeatable characteristics can only be discerned statistically and often cannot be related to any mathematically describable process. While the hazard can usually be defined quite precisely – if not always accurately – descriptions of its impact may depend substantially on the perceptions of the observer. As a result, the experts on each side of this bridge may have quite different and conflicting views as to what information it is appropriate to exchange. Here we explore the needs of the impact scientist first, remembering that relevant impacts are those of interest to the end user. A key challenge is in obtaining historical information on impacts, especially where the raw data are confidential, and then of matching suitable hazard data to them. We then consider the constraints on the hazard forecaster, who may have access to large volumes of model predictions but cannot easily relate them to the times and locations of those being impacted and who has limited knowledge of model accuracy in hazardous situations. Creating a bridge between these two requires an open and pragmatic approach from both sides, with relationships built up over time, through joint working, so that the different ways of thinking can be absorbed.

Chapter 6 looks at the aspects of forecasting systems required to achieve consistency between the prediction of the state of the atmosphere and of related environmental hazards. We first look at the different approaches to hazard prediction and then consider the limitations in predicting their meteorological drivers. We note that different modelling structures are adopted in different hazard forecasting disciplines and consider how these relate to the user requirements for those hazards. We identify the benefits of seamless approaches to hazard prediction and the challenges of achieving them in a multi-institution situation.

Chapter 7 addresses the problem of monitoring and predicting the weather. We look at how atmospheric modellers use observations to initialise their forecasts. Effective use of data in models places specific requirements on the observations. We then consider the application of basic physics and engineering in producing sensors and the observing platforms that carry them. There is a long history of close working between sensor and platform designers and meteorologists that has produced spectacular advances in forecast accuracy. However, the latest high-resolution forecasting models require data that cannot be obtained with conventional approaches to either in situ or remotely sensed observing. At the same time, new capabilities in manufacturing and communication have made available a vast amount of relatively lower-quality observations that will require new collaboration models to bring into effective use.

Finally, in Chap. 8, we take a step back and consider the warning chain as a whole system whose aim is to avoid loss by delivering the needs of those taking decisions in response to the warnings. We emphasise that, within the chain, every actor is both a user of information coming from upstream actors and a provider of information to downstream actors and that the effective definition, communication and use of that information depends on partnerships.

An effective warning system saves lives and cost, builds trust and makes people more confident of their safety. Building such a system is worthy of the time and effort that it will take.

Fig. 1.3
figure 3

Contents of the book in summary

Full size image