Cover Image
close this bookCrucibles of Hazard: Mega-Cities and Disasters in Transition (UNU, 1999, 544 pages)
View the document(introduction...)
View the documentAcknowledgements
View the document1. Introduction - James K. Mitchell
View the document2. Natural disasters in the context of mega-cities - James K. Mitchell
View the document3. Urbanization and disaster mitigation in Tokyo - Yoshio Kumagai and Yoshiteru Nojima
View the document4. Flood hazard in Seoul: A preliminary assessment - Kwi-Gon Kim
View the document5. Environmental hazards in Dhaka - Saleemul Huq
View the document6. Natural and anthropogenic hazards in the Sydney sprawl: Is the city sustainable? - John Handmer
View the document7. Disaster response in London: A case of learning constrained by history and experience - Dennis J. Parker
View the document8. Lima, Peru: Underdevelopment and vulnerability to hazards in the city of the kings - Anthony Oliver-Smith
View the document9. Social vulnerability to disasters in Mexico City: An assessment method - Sergio Puente
View the document10. Natural hazards of the San Francisco Bay mega-city: Trial by earthquake, wind, and fire - Rutherford H. Platt
View the document11. There are worse things than earthquakes: Hazard vulnerability and mitigation capacity in Greater Los Angeles - Ben Wisner
View the document12. Environmental hazards and interest group coalitions: Metropolitan Miami after hurricane Andrew - William D. Solecki
View the document13. Findings and conclusions - James K. Mitchell
View the documentPostscript: The role of hazards in urban policy at the millennium - James K. Mitchell
View the documentAppendices
View the documentContributors
View the documentOther titles of interest

2. Natural disasters in the context of mega-cities - James K. Mitchell

"Give me a place to stand on, and I will move the earth.''

(Archimedes - 287 - 212 B.C.)

Introduction

Among the many global transformations that are occurring at the close of the twentieth century is a shift from a predominantly rural world to a predominantly urban one - a brave new world dominated by increasingly large metropolitan complexes of 1 million or more people.1 The societal and environmental implications of this transition are not yet fully understood, but it is already clear that the context of natural hazards management - and perhaps the very nature of hazards and human responses - will be profoundly altered as a result. At the same time, the entire universe of natural hazards issues and institutions is already changing in response to other factors. The combination of these two trends is producing a situation that is fraught with uncertainty and ripe with opportunities for new departures in mega-city hazard-reduction programmes and policies.

1 Most sources suggest that there are now between 300 and 400 such places, but it is not possible to provide a more precise count owing to limitations of population data and the wide range of criteria that are used to define and delimit urban areas.

Background

Urbanization - especially rapid large-scale urbanization - is a major contributor to the rising global toll of disaster losses (National Research Council, 1987; Science Council of Japan, 1989, pp. 13 - 14; US Office of Science and Technology Policy, 1992, p. 10). The world's big cities are rapidly becoming more exposed and more vulnerable to natural hazards and disasters (Havlick, 1986). Unfortunately, the special problems of urban disasters have not received much attention from scholars and professionals (Somma, 1991). "Too much of our efforts to assist... for too long have been focused on rural populations which are of ever diminishing relative magnitude and importance'' (Jones and Kandel, 1992, p. 70). Much of the most useful information on urban disasters is part of a "grey literature'' that is not readily available to potential users (Earthquake Engineering Research Institute, 1985; Organizing Committee of the International Seminar on Regional Development Planning for Disaster Prevention, 1986; IDNDR, 1990).

During the 1990s a number of researchers addressed the intersecting problems of cities and disasters. One of these was the disaster relief expert Frederick Cuny, whose strong convictions about the pivotal role of cities in international disaster operations began to appear in print just before his disappearance and probable murder in Chechnya (Cuny, 1994). Other notable contributions during this period include Sylves and Waugh's (1990) case-studies of emergency operations in American cities; Zelinsky and Kosinski's (1991) valuable compilation of data on urban evacuations; Berke and Beatley's (1992) analysis of earthquake planning in mid-sized US cities (Salt Lake City, Palo Alto, Charleston); a path-breaking edited collection of papers by geographers on housing and hazards in cities in less developed countries (Cairo, Bhopal, Mexico City, Agadir, Caracas, Hong Kong) (Main and Williams, 1994); a special issue of a European geography journal that focuses on geological hazards in mountain cities (Revue de Géographie Alpine, 1994); and the proceedings of a workshop on urban hunger in developing countries organized by the International Geographical Union's Commission on Vulnerable Food Systems (Bohle, 1994). The headquarters staff of the UN-sponsored International Decade for Natural Disaster Reduction (1990 - 2000) has also published a slim volume that reviews projects designed to improve urban disaster mitigation in 11 different countries and supplies information about organizations that are active in this field (IDNDR, 1996). Though all of these publications deal with hazards and disasters in exclusively urban settings, none focuses specifically or exclusively on large cities. Recent works with a more explicit focus on mega-city hazards include collections of papers prepared by contributors to a World Bank-sponsored conference (Kreimer and Munasinghe, 1992) and by members of the International Geographical Union's Study Group on the Disaster Vulnerability of Megacities (Parker and Mitchell, 1995). In addition, the UK Institution of Civil Engineers has published a report on mega-city vulnerability (Institution of Civil Engineers, 1995). A number of noteworthy individual papers have also appeared (e.g. Aguilar et al., 1995; Boone, 1996; Dando, 1994; Degg, 1993; Driever and Vaughn, 1988; Ezcurra and Mazari-Hiriart, 1996; Firman and Dharmapatni, 1994; Guterbock, 1990; Jairo-Cardenas, 1990; Kim, 1990; McGranahan and Songsore, 1994; Masure, 1994; Mitchell, 1993a; Reed, Tromp, and Lam, 1992; Setchell, 1995; Steedman, 1995; UNEP/WHO, 1994).

Curiously, although the published volume of mega-city hazards research is not large, certain mega-cities have attracted a disproportionate amount of attention. For example, several recent books by geographers, historians, and cultural analysts have dealt with the role of fire in the development of Chicago (Cronon, 1991; Miller, 1990; C. Smith, 1995) and Tokyo (Seidensticker, 1983, 1990). There has also been an unusually large number of articles on connections among rapid urbanization, flooding, civil war, and refugees in contemporary Khartoum (Abu Sin, 1991; Agmad, 1993; Bakhit, 1994; El-Bushra and Hijazi, 1995; Hansel, 1991; Ibrahim, 1991, 1994, 1995; Ruppert, 1991; Walsh et al., 1994; Woodruff et al., 1990; Yath, 1991, 1995). However, the bulk of mega-cities with admittedly serious hazards problems receive little or no focused coverage. Clearly, given the societal importance of mega-cities, the size of the threats, and the complexity of the associated problems, there is a need for more scholarly attention to this subject.

Because the likely impact of natural disasters on mega-cities is so great and so little understood, "the vulnerability of mega-cities'' was identified as a high-priority topic for research and action during the International Decade for Natural Disaster Reduction (IDNDR). A major effort in support of that theme was mounted by the global scientific and technology community under the auspices of the International Council of Scientific Unions (ICSU). At ICSU's invitation, the International Association of Engineering Geologists and the World Federation of Engineering Organizations were encouraged to bring together a variety of academic and professional groups with the objective of formulating a collaborative agenda for mega-cities and disasters. This included projects by the UK Institution of Civil Engineers and the International Geographical Union as well as the World Bank and Habitat, among others. Independent but related programmes were also organized by the global community of landscape architects and other professional bodies. Most of these activities involve organizations of scientists, engineers and urban managers. The conference sponsored by the United Nations University on which this book is based is distinctive because it is the only one that has adopted a broad approach to the subject of mega-cities and disasters based on human ecology and social science. The difference in approach has important philosophical and practical consequences.

The intellectual principle on which the IDNDR rests is a well-founded recognition that society already possesses a great deal of valuable information about the reduction of natural disasters but has largely failed to put that information to appropriate use. For most of the scientific and engineering community the main task is to apply existing knowledge about disaster reduction to a series of specific contexts - in this case, mega-cities. It is assumed that, like Archimedes, managers can stand on the stable base of accumulated technical knowledge and engage the lever of science to "move'' the world of urban disasters. Undoubtedly, if resolutely followed, this strategy will produce important benefits, especially when it results in genuine partnerships between the producers and users of scientific information (Mitchell, 1988). By itself, however, it is an approach that incompletely engages the subject of mega-cities and natural disasters.

A basic dichotomy

Cities and disasters belong to two distinct research and management traditions that are organized by different groups of scholars and professionals around different founding assumptions, frequently in pursuit of different objectives. Often there is little communication between these groups as well as a lack of mutual understanding. The differences are readily apparent in the professional literature of both fields. Few journals about urbanization or urban issues give much prominence to environmental topics - fewer still to natural disasters - and few of the journals on natural hazards and disasters highlight urban disaster issues.

Differences in outlook and action between the two groups exist in large part because the starting points for intellectual discourse on cities and on disasters are so far apart. Natural disasters are widely regarded as Nature's ultimate sanction on human behaviour - sharp reminders that we do not live in a world that is entirely of our own making. On the other hand, cities are popularly seen as supreme human achievements - affirmations of societal power to alter Nature and sustain an environment that is, to a very large extent, divorced from Nature. When earthquakes, floods, and storms strike cities they intrude on places that are monuments to the human control of Nature. Urban disasters are, in effect, affronts to civilization!2 This helps to account for the disproportionate importance, in popular culture, of such disasters compared with similar rural events. For example, the celebrated Chicago fire of 1871 took 250 lives on the same day that 1,200 people died during a fire in rural Wisconsin. The Chicago fire has become the stuff of popular legend but only aficionados of American disasters know what happened at Peshtigo - despite the fact that it was the country's most deadly and physically destructive conflagration.

2 Of course, the starting points for analysis are not necessarily reflected in contemporary theories and explanation. For example, most natural disaster specialists are uncomfortable with the label "natural'' because they realize that a hazard is an interactive process that involves both people and natural systems. Some might even go so far as to assert that all so-called "natural'' hazards and disasters are entirely human created. In the same way, many urbanists are well aware that natural environmental factors have affected the development of cities.

For a student of natural hazards, the city is but one among many physical settings of extreme events, albeit a setting that usually makes few concessions to the natural systems that are the agents of disaster. Conversely, from the perspective of urban managers, natural disasters are just one among many problems that must be confronted within city boundaries - and an infrequent one, at that! Hazard management has a place in the spectrum of urban issues but that place must be negotiated among competing demands for attention to other problems (e.g. new urban in-migrants, decaying infrastructure, poor housing, hazardous facilities, deindustrialization, pollution, homelessness, lack of mass transit, street crime, inadequate social services, ethnic and racial tensions).

Nor is it simply that natural disasters must compete with other urban issues for public attention; they are intimately bound up with those issues. For example, the flood and landslide disasters of Rio de Janeiro, Caracas, and Tijuana are closely connected to problems of obtaining affordable housing faced by poor migrants from rural areas. Prospects for earthquake hazard reduction in Tokyo are affected by the high cost of urban land, which encourages crowded neighbourhoods, restricts open space, and pushes new developments onto hazard-susceptible low-lying filled land in Tokyo Bay. The protection of Calcutta or Lagos against flooding is also tied up with inadequate means for disposing of solid wastes, antiquated or non-existent sewage systems, and budgetary conflicts between city and national governments. The fragmentation ("de-urbanization'') of metropolitan populations and governments has a significant bearing on the design and operation of hazard-management systems in mega-cities of the United States (e.g. Los Angeles, Miami) and other developed countries. The implications are clear. If natural disasters are to be reduced in a big-city setting, the Archimedean metaphor of science-driven solutions is not an appropriate guide. Urban disasters are not merely a kind of inert problem that will yield to the direct application of scientific knowledge - they push back! In other words, not only do disasters affect cities but urbanization affects the creation of vulnerability, the scope for mitigating action, and a wide range of related topics. In brief, urbanization provides an interactive context for disaster.

There is one further twist to the relationship between disasters and mega-cities - perhaps the most important factor of all. Both elements of the problem - cities and disasters - are now changing in complex ways. (Some of these changes are treated at length later in this chapter and can be only previewed here.) The nature of disasters is changing: new kinds are emerging - some of them unprecedented. The management of disasters is changing: in many countries, existing institutions and policies are coming under increased criticism and there are calls for new ones. The ways we think about disasters are changing: for example, a post-modernist dialogue is beginning to appear in the professional literature. Equally, the nature of cities is changing: not only are they more numerous, bigger, and non-Western, but their economic and cultural functions are in flux, casting previous models of urban development into doubt. The management of cities is changing: pressed from below by burgeoning demands from their rapidly growing and diversifying populations, governments of mega-cities are also shifting to adjust to other pressures such as the changing role of national governments in the face of global economic integration and the reluctance of those governments to honour existing commitments to assist the cities. Urban mass movements and emergent neighbourhood organizations are also exerting pressure for the "reinvention'' of mega-city governments. As big cities become increasingly important actors on the world stage, their populations are drawing on the experience of peer cities beyond national boundaries. As a response to such changes, it is no surprise that the intellectual basis of urban analysis is also in flux.

In these circumstances, the Archimedean metaphor requires further modification. Not only are the problem spheres of disasters and urbanization interactive (i.e. urban disasters "push back''), but given the speed and scope of changes that are now afoot in both spheres there may no longer be a truly stable base from which the leverage of scientific knowledge can be pivoted. This means that taken-for-granted assumptions about scientific information and disaster reduction need to be critically examined in light of the highly changeable contexts that bracket both disasters and mega-cities. Until these contexts are better known, the transfer of knowledge and technology among countries and cities should proceed with caution. Mega-city disaster-reduction initiatives will have to be carefully tailored to local conditions and specific settings. It is one of the tasks of this book to show how this process might proceed in different mega-cities.

A contextual model

In coming to grips with the complexities of a dynamic set of relationships between mega-cities and natural disasters it may be helpful to think in terms of a contextual model (fig. 2.1; Mitchell, Devine, and Jagger, 1989). Such a model is intended to show how a more or less discrete process relates to the broader environment of which it is a part. For the purposes of this chapter, the central process is one of hazard and the larger context is that of urbanization.3 Both components of the model can be considered separately and in conjunction. The hazard component includes four main elements: physical processes, human populations, adjustments to hazard, and net losses. As people seek to adjust to hazardous natural environments, these four elements modify each other through seven endogenous feedback relationships. A change in one or more elements may set in motion a cascade of reflexive changes in the others. The stimulus for change may be some or all of the following: an extreme natural event such as an earthquake or a storm; community encroachment into an area of known hazard; urban technologies that inadvertently alter existing physical risks; the development of vulnerable groups, institutions, or technologies; the adoption of specific new hazard-management measures and the decay of old ones; and a broad-based commitment to raise general thresholds of public protection against hazards of various kinds.

3 There is nothing to prevent the reader from reversing this order, making urbanization the process and disasters the context. Indeed, it can be argued that such an analysis is probably essential to a complete understanding of the urban disaster nexus. However, for convenience and because my own primary experience lies in the field of natural hazards research, I have placed hazard at the centre of the model. The background to this model is discussed in greater detail in Mitchell, Devine, and Jagger (1989).


Figure 2.1 A contextual model of hazard (Source: Mitchell, Devine and Jagger, 1989)

Change can also come from outside the hazard system - in this case from the urban context in which the hazards are situated. That context includes exogenous factors or processes that interact with components of hazard, but are largely independent of them. These include variables such as the stages of city development (including the growth of mega-cities), the internal structure and functioning of urban areas under different sociocultural, political, and economic systems, and the urban implications of a wide range of contemporary changes in science, technology, environment, economics, politics, and culture. Among the latter are: the flow of new scientific information and the diffusion of new technologies; the growing recognition that humanity now has the capability to affect environmental change on a global scale; worldwide economic restructuring and integration; the collapse of an existing international geopolitical order and the development of replacement arrangements (e.g. continental trading blocs, ethnically fragmented states, "reinvented'' governments, non-governmental organizations, urban mass movements); and the adoption of "post-modern'' modes of thought and action both within the intellectual community and beyond.

These two components (i.e. hazard system and urban context) are a bit like the "figure'' and "ground'' of visual illusions. In the same way that a viewer can construct different interpretations of an illusion by focusing on either the figure or the ground, it is possible to arrive at an explanation of urban disasters that depends on disaster agents and proximate human responses or to fashion an explanation around the distal structuring characteristics of urban systems and urban lifeways. The latter approach may redefine the meaning of disaster in an urban setting and predispose cities to "disasters'' that are different from those natural scientists customarily think of.4 But urban disasters are not just amalgams of disaster and urbanization; they are the products of a set of changeable relations between both components. This feature makes it difficult to fashion effective programmes of urban disaster reduction by employing only the perspectives on hazards management or urban management. The permutations of natural hazard and urbanization produce different outcomes in different settings at different times. Moreover, the set of possible permutations is significantly larger than the set of outcomes to date; an understanding of this fact opens up a vast untapped potential for intervention to make cities safer places.

4 The American humorist Garrison Keillor touched on this phenomenon in a piece about the various imagined disasters that arise in the minds of people who are trapped in malfunctioning high-rise elevators: "In the city of New York, you go out your door in the morning, you take your life in your hands. You may not get off an elevator the same person you get on. Choose your elevator wisely'' (New York Times, 19 December 1993, p. E13). He might just as readily have said "choose your city wisely''!

Mega-cities

Because mega-cities are a relatively recent phenomenon, the literature on mega-city disasters is sparse - though suggestive of problems to come. Much more is known about the relationship between natural disasters and urbanization in general. This section begins with a long caveat about attitudes to city hazards by urban analysts and other social commentators. It continues with a brief historical overview of mega-city disasters and then surveys the changing balance of safety in mega-cities. This is followed by a discussion of the disaster-forcing attributes of urbanization, a review of the evolution and status of mega-cities, an examination of different types of urban structure that may be reflected in mega-cities, and an exploration of the contrasting disaster issues of rich mega-cities and poor ones.

The intention is to show that, despite some tendencies toward uniformity and homogenization, there is a great deal of variety among mega- cities and considerable uncertainty about their eventual dimensions. Different types of mega-city are likely to have quite different disaster-susceptibilities and different capabilities for response. Whether this will continue is another matter. Mega-cities may grow, becoming both more complex and more alike. Or there may be perhaps two or three basic variants. For example, if the organization of the global economic and political system continues to be dominated by a relatively few countries, it may well be that the natural hazard profiles of mega-cities in the dominant (rich) countries will continue to diverge from those of the (poor) rest, thereby rendering problematic the exchange of hazard experience and hazard-reduction technologies between both. On the other hand, there is evidence that some of the characteristics of third world cities are beginning to show up in places such as Los Angeles and Miami, thus raising the possibility of eventual reconvergence.

The best of places; the worst of places

He compared Los Angeles and Mexico City (which he knew well) to volcanoes, spilling wreckage and desire in ever-widening circles over a denuded countryside. It is never wise, he averred, to live too near a volcano.

(Davis, 1992)

Recent urban scholarship is reconceptualizing the city and the process of urbanization. But there is little consensus about what is emerging. For some analysts, cities are simply enduring features of the human landscape whose public images wax and wane (Hall, 1988; Jones, 1990). According to Peter Hall, during the past century people have passed through a great cycle of belief about cities. In the late nineteenth century, cities were widely regarded as places of physical decay and spiritual desolation - "The City of Dreadful Night'' (Hall, 1988; Jones, 1990). Thereafter, especially during the period from 1920 to 1970, more positive and hopeful images appeared, partly in response to an array of efforts at urban improvement. But, towards the close of the twentieth century, the city - especially in Western societies - is again being viewed as a place of decay and despair. On the other hand, different commentators have challenged these views by suggesting that urbanization has begun to take on entirely new forms that call into question the very notion of the city. In the United States, commentators have pointed to the emergence of vast suburbias (Fishman, 1987) or "edge cities'' (Chudacoff and Smith, 1994) or to "cyburbias'' that are plugged into an information revolution and disconnected from particular places (Sorkin, 1992). Similar trends have been reported - at least in embryonic form - among cities of the third world such as Bangalore in India (New York Times, 21 December 1993). Yet other writers have identified cities whose purposes and forms are so deeply and continually contested by subgroups of inhabitants that it is difficult to form a coherent picture of the whole (Davis, 1992).

Against this background of critical scholarship, the dominant view of the city seems to be that of a place riven by problems (Brunn and Williams, 1993, pp. 32 - 41; Jones, 1990, pp. 163 - 166; McHarg, 1969; Teune, 1988). Natural hazards and disasters are usually fitted readily - all too readily - into these problem sets. Both in rich and poor countries, the growth of such giant urban agglomerations poses major metropolitan planning problems which must be solved in a broad regional context. Food, fuel, electric power, and drinking water must be brought in from considerable distances, and massive quantities of waste materials must be disposed of. The giant metropolis needs very sophisticated transportation systems if it is to function effectively, and city dwellers may have to travel considerable distances to recreational facilities and surviving farm and wilderness areas. Most major cities also have some special problems resulting from the peculiarities of their sites, for example subsidence, earthquakes, and air and groundwater pollution buildups in Mexico City, and flooding and typhoon damage in Dacca. (Bromley et al., 1989, p. 372; emphasis added)

No one would contest that many cities do indeed face serious problems. However, the reality of urban living - and the role of natural hazards in mega-cities - is more complex. This is aptly recorded in Alexander Frater's book Chasing the Monsoon (1991). At one point in a journey around India following the monsoon, Frater asks a Bombay poet whether the sudden arrival of torrential rains at the end of the dry season is a romantic event. The reply comes back:

"Not in the big cities.... Where's the romance in mud, slush, rats and floods? Or in the anguish of watching those who live in the streets trying to cope?... For many of my fellow-citizens it is a period of misery and hardship.'' (Frater, 1991)

But, a few days later in Calcutta, Frater concludes that:

despite the excess rain that kept sousing and rotting its foundations... the city was indestructible.... For the people on the streets it was a haven. They were in Calcutta by choice, victims of starvation or persecution elsewhere, and would rather be here, living rough, than in the places from which they had fled. (Frater, 1991)

These two, quite different, observations capture the ambiguity of living in a hazardous big city; such places may be exposed to serious physical hazards but they are also centres of human opportunity (Tuan, 1988). Both aspects of the situation need to be kept in mind by those wishing to address the hazards of mega-cities. The two contrasting characterizations of the city are freighted with policy implications for the reduction of natural hazards and disasters. A pathological perspective on cities is readily paired with an emphasis on disaster vulnerability; a more benign assessment of cities underscores their potential for resilience (Sachs, 1988). An emphasis on vulnerability tends to narrow the scope for individual actions or informal group actions by populations at risk; they are regarded as passive victims in need of protection. Conversely, an emphasis on resilience points hazard analysts in a different direction. We are reminded that laypeople are often capable of acting in their own best interests, and that this may mean finding ways to live with hazards, rather than seeking to control them or the populations they affect.

Mega-cities affected by natural disasters

Large cities of antiquity certainly experienced heavy losses as a result of natural disasters, but the data are highly incomplete.5 Chandler and Fox (1974) show that at least a score of cities each recorded tens of thousands of deaths in the wake of earthquakes, floods, volcanic eruptions, and other events. But few of these places had populations that exceeded 200,000 and most were considerably smaller.

5 Appendix 2 provides a list of known pre-twentieth-century urban disasters that killed more than 10,000 city residents.

The Tokyo - Yokohama metropolitan area was probably the first mega-city to be devastated by a natural disaster. The Great Kanto earthquake (1923) killed approximately 100,000 people in Tokyo and destroyed more than 700,000 houses in a city of over 2 million (Seidensticker, 1983, p. 259). In the half-century between 1945 and 1995, around 100 natural disasters affected large cities. The most physically destructive single events during this period took place in: Nagoya (typhoon, 1959), Tangshan (earthquake, 1976), Bucharest (earthquake, 1977), Mexico City (earthquake, 1985), Miami/Dade County (hurricane, 1992), and Kobe (earthquake, 1995). However, a focus on great catastrophes may be misplaced, because the aggregate impacts of lesser events can be much larger. This is particularly so in third world mega-cities, where small storms, floods, and seismic shocks, events that would be considered inconsequential elsewhere, often wreak disproportionately heavy losses. In such places, natural hazard is a more or less chronic everyday problem. Earthquakes, severe storms, and wildfires appear to be the most frequent catastrophic urban events.

It is worthwhile noting that urban areas are the sites of many different types of disaster: natural, technological, biological, and social. Thus far, most of these have been simple events of one type, but compound events that involve two or more of these components are beginning to appear (e.g. floods that disperse toxic materials, earthquakes that rupture oil pipelines and storage tanks, droughts that uncover flooded waste disposal sites). The potential for further "surprises'' is obvious.

A shifting balance between hazard and opportunity

Evidence of increasing disaster-susceptibility in mega-cities bespeaks a change in the balance of hazard and opportunity in urban life. Of course changes in the mix of urban hazards and shifts in the hazard - opportunity balance have gone on throughout history. For several centuries the balance has tilted steadily towards higher levels of safety and opportunity for most city residents, while hydrological, meteorological, and geological hazards have declined relative to other types of threat. Urban concentrations of wealth, knowledge, talent, and power have justified - and made possible - extraordinary efforts to protect people and property against the more obvious extremes of Nature. The avoidance, prevention, and mitigation of natural hazards were a primary responsibility of those who ruled the ancient cities of Egypt, China, and Central America (Tuan, 1979; Wu Qingzhou, 1989). Dyking, drainage, and irrigation schemes were among the earliest urban public works in these places. The suppression of fire was a high priority in the crowded cities of medieval Europe and elsewhere at other periods (Rosen, 1986). A widely shared commitment to protect urban lives and property against natural extremes continues today. Contemporary Japanese cities are showplaces of hazard engineering, the metropolitan areas of Rotterdam and London are flanked by some of the world's most modern and technologically advanced flood defences, Mexico City has invested in the world's only public earthquake alarm system (New York Times, 19 December 1993), and the vast drainage network of greater Los Angeles has been so extensively modified by humans that it is now an almost totally artificial construct (Cooke, 1984; McPhee, 1989).

Heightened urban sensitivity to hazards and the benefits of hazard reduction do not mean that big cities have been - or are - entirely safe from extreme natural events. Nor do they mean that cities have been free of other types of hazard. (The historical record of urban epidemics, riots, fires, industrial accidents, and similar social ills is salutary.) But, over the long run, it is clear that a quite remarkable record of success has attended efforts to provide many urban dwellers with improved security against geological, meteorological, and hydrological hazards.

Now there is growing concern that cities are becoming so numerous, so large, so complex, and so volatile that the balance between risk and reward will tilt back towards conditions of increased natural hazard in the absence of determined efforts to reverse that trend. A world of mega-cities may well be more risky, more exposed, and more vulnerable than the world we currently inhabit. The scale of potential mega-city losses imposed by floods, storms, earthquakes, and other phenomena compels attention to the problem. So too does the timing, for we are fast becoming a predominantly urban world, a world where half the urban population will be citizens of mega-cities.

Urbanization: An agent of disaster

Before examining the links between mega-cities and disasters it is necessary to identify the general hazard-intensifying and disaster-forcing aspects of urbanization. Urban development increases disaster-susceptibility in a number of ways. First is the frequent association of cities with naturally risky locations such as seacoasts and floodplains because such places also confer important benefits (e.g. buildable land, well-appointed sites for the collection and transhipment of goods, fertile hinterlands). Initial settlements may take advantage of available safe sites but subsequent growth typically spills over into adjacent high-risk areas. Coastal metropolises of the eastern United States (Boston, New York, Miami, Houston) and the seaward-expanding cities of Japan are good examples.

Secondly, the physical process of building cities often creates or ex acerbates existing environmental risks. For instance, paving over watersheds reduces infiltration, speeds runoff, and increases flood volumes; constructing coastal defences may reduce supplies of beach sand and facilitate erosion during storms. As the leading edge of urban development marches across the landscape, the incidence of natural disasters tends to keep pace (Cooke, 1984). The human role in creating conditions for disaster is clearly visible. Bangkok klongs (canals) that used to accommodate overflow from the Chao Phraya River have been filled in to create streets that are now chronically flood prone, while the city continues to subside owing to pumping of water from underlying aquifers.

Thirdly, cities increase disaster potential by concentrating people and investments. A disproportionate amount of material wealth is bound up with cities in the form of buildings (ceremonial, commercial, industrial, and residential) and infrastructure (i.e. the complex and expensive networks of lifelines that sustain urban populations and make it possible for them to interact with each other and the outside world). When an extreme event occurs, urban losses are often very heavy. In a matter of hours, hurricane Andrew inflicted over US$ 20 billion of property damage on the Miami metropolitan area, whereas it took about six weeks of heavy flooding in mostly rural sections of nine Midwestern states to produce approximately half as much material loss (Myers and White, 1993).

Fourthly, the built environment is continuously wearing out but the rate of urban replacement rarely matches the rate of urban obsolescence. As a consequence, most cities contain large concentrations of old buildings that fail to meet present standards for hazard-resistant construction. Differential ageing and uneven replacement of the physical stock typically produce a complex patchwork of disaster-susceptibilities.

Fifthly, many urban areas contain populations that are particularly vulnerable to disaster. For example, metropolitan areas often attract large numbers of in-migrants, most of them poor and all of them separated both from the familiar landscapes of home, whose risks were known, and from traditional support networks or customary behaviours that provided a modicum of security in the event of disaster.

Finally, few governments of rapidly growing cities have been able to allocate significant resources to hazard reduction when they are already stretched to breaking point by the task of providing basic support services for their expanding populations. In short, cities often contain all of the ingredients for disaster: heightened risks, concentrated exposure, and increased vulnerability. In light of the available evidence about intensified urbanization associated with mega-cities, the potential for a quantum leap in disaster-susceptibility is clear.

Changing patterns of urbanization

Some time during this decade there will be more people living in urban areas than anywhere else on Earth. In 1990 urban dwellers accounted for 47 per cent of the world's population but that number is expected to rise to around 63 per cent by 2020. Some 6,000 years separated the earliest urban centres from the first cities of more than 1 million people. Now there are about 300 such places and their numbers are growing rapidly. At the beginning of the twentieth century approximately 2 per cent of all humans lived in just 14 mega-cities.6 Today the proportion is close to 20 per cent and it will probably rise to around 30 per cent by the year 2020 (Fellman et al., 1992).

6 London, Manchester, Birmingham, Glasgow, Berlin, St. Petersburg, Moscow, New York, Chicago, Philadelphia, Boston, Calcutta, Peking, and Tokyo.

Moreover, the size of mega-cities is increasing rapidly. In 1900 there were no cities of more than 10 million and only three that exceeded 3 million (London, Paris, and New York). Now there are about 60 cities with populations greater than 3 million and 14 super-mega-cities of more than 10 million (table 2.1).7 What is, by some counts, the world's largest city (Mexico City) currently contains at least 15 million people. The entire contiguous area of Mexico City (i.e. the central city plus adjacent smaller cities, towns, and suburbs) may hold more than 20 million. It is projected that there will be two or three urbanized areas of 20 million by the year 2000 (Mexico City, Sa o Paulo, and perhaps Tokyo - Yokohama) and another 20 of between 10 and 20 million (Jones and Kandel, 1992, p. 75).

7 Fellmann, Getis, and Getis (1992) identify the 10+ million cities as: Beijing, Bombay, Buenos Aires, Cairo, Calcutta, London, Los Angeles, Mexico City, Moscow, New York, Osaka - Kobe, Paris, Rhine-Ruhr, Rio de Janeiro, Sa o Paulo, Seoul, Shanghai, and Tokyo - Yokohama.

Mega-cities now occur in almost every culture and region on Earth. They are absent or scarce only in Oceania and high latitudes of the northern hemisphere. Countries that are notably rich in mega-cities include: the United States (36), China (31), India (21), Russia (15), Brazil (13), Germany (10), and Japan, the United Kingdom, and Indonesia (7 each). Conversely, most African states (20) have none. Coastal locations predominate, especially in Latin America and Australia.

Table 2.1. The world's 20 largest metropolitan areas ranked by population, estimated for 1995

Rank

Location

Population (million)

Area (km2)

Density (inh/km2)

1

Tokyo-Yokohama

27.2

2,819

9,664

2

Mexico City

20.9

1,351

15,465

3

São Paulo

18.7

1,168

16,017

4

Seoul

16.8

885

18,965

5

New York

14.6

3,298

4,434

6

Osaka-Kobe-Kyoto

13.9

1,281

10,825

7

Bombay

12.1

246

49,202

8

Calcutta

11.9

541

21,990

9

Rio de Janeiro

11.7

673

17,364

10

Buenos Aires

11.6

1,385

8,416

11

Moscow

10.4

981

10,646

12

Manila

10.2

487

20,867

13

Los Angeles

10.1

2,874

3,525

14

Cairo

10.1

269

37,509

15

Jakarta

9.9

197

50,225

16

Teheran

9.8

290

33,726

17

London

9.1

2,263

4,028

18

Delhi

8.8

357

24,570

19

Pans

8.7

1,110

7,797

20

Karachi

8.0

492

16,292

Sources United Nations (1987), Jones and Kandel (1992)

The spatial distribution of mega-cities has also undergone a major shift in recent decades. Most of the early twentieth-century mega-cities were located in Europe and North America (Berry, 1990, pp. 109 - 110). Today, by far the great majority - and most of the super-mega-cities - are in less developed countries of Asia, Africa, and Latin America (Gilbert and Gugler, 1992; Yeung, 1990). To a greater degree than ever before, mega-cities will be composed of large numbers of poor people who depend on informal economic institutions and systems for support (Perlman, 1987). As we shall see, this change has important implications for the management of urban hazards.

How far will the trend toward global mega-cities proceed? Is it essentially unilinear - towards more, bigger, and increasingly complex places? Might deconcentration and dispersal be an alternative future pattern? Are all mega-cities likely to evolve in the same way? How might expected changes affect prospects for disaster and disaster reduction? If current trends continue, and mega-city development follows the pattern of North America and Europe, most of the signs point in the direction of increased population concentration at the global and national scales, with the possibility of regional and local deconcentrations and dispersals. In the United States, the aggregate pattern of population and investments has, for some decades, been shifting from the interior to the coasts, from the north towards the south, and from the east towards the west. Massive growth of cities such as Los Angeles and Miami follows in the wake of these trends. Similar processes are at work in many other countries. If Doxiadis's vision of the future is correct, the eventual outcome a century hence may involve the coalescence of linear mega-cities to form a more or less continuous urban "ecumenopolis'' comprising a population of around 12 billion people (Jones, 1990)!

At the same time, in developed countries, there is a substantial amount of deconcentration (counter-urbanization) at the scale of the individual mega-city. Central-city growth has slowed dramatically or gone into reverse and the populations of many such cities are emptying out into a far-flung web of low-density suburbs and exurban developments. Some observers argue that there is no longer a need for urban concentration and that future large cities will be spread out over such large areas at such low densities that they will not resemble the historical city at all. Whether or not this occurs, it is clear that third world mega-cities have a long way to go before reaching this stage. In such places, the process of deconcentration is slow, indeed barely detectable in most.

Effects of city form and structure on natural disaster potential

Urban analysts recognize a number of different cultural variants of city form and structure, each with their own characteristic patterns of land use and population distribution. For example, Latin American cities tend to be made up of concentric zones, with old but well-maintained housing near a single central business district and poor shanty towns on the periphery. Industrial areas and the élite residential zone cut across these rings in the form of wedge-shaped blocks of territory. In contrast, the typical North American city more or less reverses this pattern, with a ring of decayed housing surrounding the central business district and affluent suburbs on the periphery. Also the central business district may be in decline, its functions challenged by many secondary shopping and industrial nodes scattered throughout the metropolis. West European cities often have well-marked historic cores with many preserved public buildings, surrounded by old houses - newly upgraded by middle-class gentrifiers - side by side with the cramped homes of working-class residents. Segregated urban ethnic neighbourhoods are less marked than in the United States. Suburbs may be either rich or poor and are often flanked by green belts with dormitory villages in the countryside beyond. African and south Asian cities tend to contain substantial tracts of market gardens within the built-up area. Middle Eastern cities often preserve distinct religious "quarters.'' The former socialist cities of Russia and Eastern Europe have a high percentage of their land given over to industrial districts and relatively few socio-economic contrasts among neighbourhoods. For much of the post - Second World War period, new investment was denied to Shanghai and other coastal cities of China, as national leaders sought both to hold down the growth of all large cities and to spread economic development more widely into the interior. The contrast between outmoded facilities and today's resurgent economic boom in the coastal mega-cities of China is striking (Gaile, 1983; Sit, 1988; New York Times, 22 December 1993).

These examples do not exhaust the catalogue of cultural and regional variations of urbanization but suffice to make the point that cities are functionally organized in different ways that have distinctive areal expressions. These patterns are then draped over a varied physical landscape and adjusted to it, a process that helps to disguise the underlying organizing principles. When a natural disaster strikes, it disrupts - and may destroy - not just the lives of citizens and the city's physical fabric but also the functional organization of the metropolis. However, the same organizing principles typically reassert themselves during post-disaster reconstruction.

Different types of organization have different loss potentials, different vulnerabilities, and different prospects for recovery. Likewise, different types of city require different types of preparedness-planning and different hazard-mitigation strategies. For example, a single central business district plays an irreplaceable role in Lima (Peru), a city that contains 30 per cent of the country's population and 70 per cent of its industry; the several business districts of a typical multi-nucleated North American city share economic functions and rarely serve the entire nation. Likewise, the Lima periphery would hardly be a suitable location for the replication of suburban watershed management schemes borrowed from an American mega-city. In similar vein, the in-city agricultural lands of African and South-East Asian cities provide amounts of open space for hazard refuges and post-disaster relief staging areas that could only be dreamed about in the typical large Japanese city.

The preceding variants notwithstanding, there are also significant similarities among mega-cities. For example, the great majority are ports - very often the most important ports in the respective countries. Therefore protection against waterborne hazards and disasters is likely to be a shared concern, and safeguarding marine facilities, ship channels, and water-dependent land uses may receive high priority. Furthermore, if, as some studies suggest, the physical fabric and organizational structure of mega-cities are becoming more uniform, then existing differences may gradually disappear in the future.

Mega-city disasters: Polycentre and periphery perspectives

A global trend towards ever-larger cities is unmistakable. So too is a trend towards increasingly similar urban forms. But it is equally clear that there remain vast differences between the mega-cities of developed countries and those of the developing world. Chief among these is the contrast between wealth and poverty. That dichotomy has important implications for the management and reduction of urban natural disasters. A brief review of the contrasting perspectives makes this clear.

Much of the concern about natural disasters in mega-cities of developed countries stems from the fact that such places contain a large proportion of the world's material investments and economic wealth. A disaster that takes many human lives is clearly possible in cities such as London, Los Angeles, or Tokyo, but such places are much more likely to suffer heavy economic and material losses in the wake of extreme natural events. As hurricane Andrew demonstrated for Greater Miami and the Hanshin earthquake for Kobe, a single metropolitan disaster in a rich country can produce economic losses that rock the global reinsurance industry.

However, the issue of mega-city disasters in rich countries is not simply one of losses that are inflicted on the affected communities themselves; it is increasingly connected with the roles that such places play in the global economy and the capacity of disasters to disrupt global economic functioning (Sassen, 1991). Several years ago the world's business community was jarred by reports that a major earthquake in Tokyo might precipitate a collapse of the global economic system (M. Lewis, 1989). This revelation followed on the heels of a record 1987 windstorm in southern England that had shut down the London Stock Exchange and may have helped to trigger the worst international stock market crisis since the Great Depression (Mitchell, Devine, and Jagger, 1989). Since then a series of other events has further underlined the disaster-vulnerability of mega-cities that anchor the global finance and trading network. They include: an underground flood that immobilized much of Chicago's financial district; riots and the possibility of a great earthquake in Los Angeles; an earthquake that disrupted the San Francisco Bay Area; and the bombing of the World Trade Center in New York. Between them these six mega-cities house the headquarters of more than 60 per cent of the world's leading private corporations (Berry, 1990).

Now there is evidence that the urban-based new information technologies on which the global economy increasingly depends are highly vulnerable to disruption by storms, floods, earthquakes, and other unexpected events. Nineteen mega-cities together make up a global "polycentre'' that directs and controls the international entrepreneurial system (Berry, 1990).8 Fifteen of these places account for 70 per cent of all electronic data flows in the contemporary world (J. Lewis, 1990). Not only are mega-cities of the polycentre full of expensive buildings and infrastructure and strategic economic functions, their sophisticated but demanding technologies that make for economic (and perhaps geopolitical) dominance may be susceptible to new kinds of vulnerability. In these cities, as well as the familiar consequences of natural disasters, there is considerable potential for disaster "surprises'' in the future (Mitchell, 1996).

8 New York, Chicago, Los Angeles, San Francisco, Philadelphia - Wilmington, Dallas - Fort Worth, Houston, St. Louis, Detroit, Pittsburgh, Tokyo, Osaka - Kobe, Seoul, London, Paris, Ruhrgebiet, Frankfurt, Randstadt, Rome.

What do we know about the ability of mega-cities in developed countries to respond to severe natural events? The Loma Prieta earthquake and other recent mega-city disasters provided evidence that some of the measures that have proven to be effective in reducing losses in smaller cities may not work quite so well in emerging Western mega-cities (Mitchell, 1993a). Prediction, warning, and evacuation systems that depend on sophisticated technology and highly effective public bureaucracies are particularly open to question. In addition, big-city disasters possess features that have not been common in smaller communities and that may raise entirely new problems of disaster management. For example, disaster impacts in mega-cities that dominate mass media markets are likely to be extensively, continuously, and obsessively reported whereas impacts on other communities that have less access to these channels are likely to be overlooked. The consequences for skewing post-disaster relief are considerable. Secondly, the complex societal mixes of mega-cities pose new problems for the delivery of emergency response services and disaster relief; linguistic, ethnic, and other differences are often marked in such places. Thirdly, the sheer size and complexity of mega-city infrastructure networks make them particularly liable to disruption. Finally, recovery is apt to occur more slowly than in smaller places. In short, past lessons of disaster management may no longer be applicable in the mega-cities of the polycentre.

Of course, most of the world's mega-cities are not part of the polycentre. Instead they serve as primary contact points between the polycentre and regional or local markets on the global periphery. Tijuana (Mexico) is a good example. Once a small regional town, it is now the fourth-largest city in Mexico with a population of well over 1 million. Tijuana's recent growth has been fuelled by investments of multinational corporations in maquiladora firms near the US border. As more shanty towns go up in the steep semi-arid valleys of the city fringe, and more people crowd into the riverside lowlands, the incidence and severity of floods and landslides in Tijuana are also accelerating.

In places such as Manila, Dhaka, Ankara, or Lima there is the potential for heavy loss of life during disasters as well as catastrophic material destruction. The situation in Lima is typical. This is a city that has suffered severe earthquakes at least five times in the past 300 years. At the end of the Second World War, just over half a million people lived in the metropolitan area. Today, there are more than 5 million. Vast numbers of poor rural peasants have flooded into Lima. Not all groups are equally exposed to hazard. Indeed, the pattern of hazard-susceptibility is a complex one that has evolved in response to changes in demography, economics, land ownership, building practices, and other factors. Middle- and upper-income groups live in well-constructed houses that often comply with antiseismic codes and are sited in neighbourhoods with wide streets and ample open spaces. If affected by an earthquake there are enough resources to ensure quick recovery. The peripheral shanty towns (pueblos jovenes) are also low-density settlements, this time composed of light bamboo structures that do not collapse when the ground moves. People are poor, but levels of social organization are high. In contrast, seismic vulnerability is high in the inner-city slum areas. Here many poor families are crowded into old adobe brick structures, adjoining streets are narrow, and open spaces are non-existent. There are few neighbourhood organizations or other local institutions that might be called on in the event of a disaster. Here earthquake protection measures are minimal or, more often, non-existent. As summarized by one observer, the situation is full of bleak prospects:

The inhabitants of critical areas would not choose to live there if they had any alternative, nor do they neglect the maintenance of their overcrowded and deteriorated tenements. For them it is the best-of-the-worst of a number of disaster-prone scenarios such as having nowhere to live, having no way of earning a living and having nothing to eat. Given that these other risks have to be confronted on a daily basis, it is hardly surprising that people give little priority to the risk of destruction by earthquake. (Maskrey, 1989)

In summary, there is a high degree of uncertainty about the future of mega-cities. Their growth seems assured, but at what density? New ones may spring up in unexpected places under the influence of changing geo-economic forces. Increasingly similar in outward form, mega-cities in different cultures and continents may still retain distinctively different internal structures. The divisions between rich mega-cities and poor ones may become wider and their disaster-susceptibilities may also diverge. But, at the same time, the differences between all mega-cities and their rural hinterlands may become sharper. It would be foolhardy to assume that the disaster-susceptibility of any one mega-city will be quite like that of any other. This is a period of great urban flux; it bears close watching by students of hazards and disasters.

Hazards and disasters

As the previous section shows, the process of large-scale urbanization is complex and changing. So too are the study and management of natural hazards and disasters. This section focuses on some of the policy and management-related changes of recent years. Although the US experience is highlighted, the changes noted apply to many other countries. At the same time, new ways of thinking about hazards and disasters are emerging, whose long-run implications are difficult to foresee. A brief introduction to some of these changes is provided; the reader is referred to other sources for more exhaustive coverage (Mitchell, 1993b).

Emerging crises of natural hazards management

The adequacy of existing means for managing natural hazards and other types of environmental hazards is increasingly being called into question in the United States and the global community. This is illustrated by a sampling of the issues that have recently emerged in professional and lay forums.

First are problems that are posed by new types of hazard. These come in several varieties. Some are amalgams of natural and technological hazards (Showalter and Myers, 1994). When a storm or a tsunami affects a chemicals manufacturing or storage facility it is not just the threat of high water and strong winds that is of concern; it is also the possibility that toxic materials may be dispersed throughout surrounding areas (e.g. Nagoya, 1959; Times Beach, Midwest floods, 1993). If an earthquake affects a nuclear reactor site, radioactive materials may be released. The flooding of old mines can cause surface collapses.

Given the expanding variety of technological hazards, the possibilities for new or unusual combinations of natural and technological hazards are spiralling upwards. Three classes of technological hazard pose quite different sets of problems when combined with natural hazards:

(a) Unsuspected hazards involve substances or activities that were regarded as harmless or benign until scientific evidence or human experience showed otherwise (e.g. DDT, asbestos).

(b) Improperly managed hazards involve failures of various kinds of hazard-control systems (e.g. nuclear facilities such as Windscale, Three Mile Island, Chernobyl; chemical plants such as Seveso, Basle, Bhopal; transportation systems such as the US space shuttle Challenger and supertankers such as the Exxon Valdez; storage and disposal sites for toxic materials such as Kyshtym, Times Beach, Love Canal, Minamata).

(c) Instrumental hazards are intended to cause harm and are consciously employed towards that end; they include sabotage, arson, and warfare. Military industrial technologies belong to this group (e.g. nuclear, biological, and chemical weapons such as defoliants and nerve agents; deliberate oil-spills and oilfield conflagrations).

The UN Department of Humanitarian Affairs, created in 1992, has begun to examine a different but related set of problems that they call complex emergencies. These refer to events such as those occurring in the former Yugoslavia, Kurdistan, southern Sudan, Mozambique, and Somalia, where political conflicts, drought, famine, and other problems are intertwined (DHA News, January - February 1993). Hazards of global environmental change constitute a separate but related class of events that are now making their way onto the public policy agenda (Mitchell and Ericksen, 1992). It is widely accepted that a build-up of greenhouse gases in the atmosphere might trigger climate changes and other repercussions such as sea-level rise. Some of the industrial hazards are sufficiently well known to be classifiable as "routine'' hazards, but others - including most of the hazards connected with global environmental change - are entirely unprecedented in the human experience. They are best considered "surprises'' (Mitchell, 1996). How should public policies be changed to take account of the widening range of threats to human survival?

A second way in which natural hazards are changing grows out of the first. It is that there are now strong pressures to expand the legal definition of natural disasters. In the past, only the victims or potential victims of events triggered by natural phenomena (somewhat erroneously labelled "acts of God'') were considered eligible for public assistance to upgrade preparedness or provide relief. However, in recent years there has been an unmistakable trend towards broadening the range of technological and social phenomena that are eligible for aid. In the United States this began with natural gas shortages in the cold and snowy winter of 1977 and later included the community of Times Beach, Missouri - a notorious case of contamination by the toxic chemical dioxin. More recently, the collapse of an old, disused, and deteriorating underground railroad system was treated as a "natural'' disaster when water from an adjacent canal inundated the basements of high-rise buildings in downtown Chicago. In 1992, civil unrest in Los Angeles also qualified for disaster status, as did the 1993 bombing of the World Trade Center in New York city. These events suggest that distinctions between different kinds of disasters are waning in the public policy arena. What accounts for such changes? Perhaps they reflect the growing impact of socio-technical hazards and the decline of natural phenomena in the intensely human-made environments of a rich country? Maybe they are linked to further politicization of public decisions about disasters, or to the political influence of specific interest groups that place a high premium on predictability and stability (e.g. business corporations)? It is also possible that they are products of a general shift in public attitudes towards risks of all kinds.

A third type of change is swelling public dissatisfaction with hazard-management agencies. Criticism of disaster management in developing countries such as Bangladesh or the states of the African Sahel is not new. Mass media reports about the poor performance of national government organizations and international agencies are legion. Natural hazards and disasters can be volatile political issues in developed countries and a certain amount of controversy about governmental responses is the norm - as anyone who has observed the aftermath of Italian earthquakes, or Australian wildfires, or American hurricanes can attest to. But recently there has been a sharp escalation of complaints about the effectiveness of hazard-management agencies in major developed countries such as the United States, the United Kingdom, Italy, Australia, and Russia.

The US Federal Emergency Management Agency (FEMA) has been a particular target. It has been accused of providing insufficient and inappropriate relief to disaster victims. It has also been criticized for encouraging the occupation of hazardous lands by offering low-cost insurance to rich investors; and it has drawn fire for devoting too much effort to cleaning up after past disasters and too little effort to reducing the probability of future disasters. FEMA's mishandling of relief in the wake of hurricane Andrew (Florida, 1992) triggered a major investigation by the US Congress. Critics called for the nation's armed forces to replace FEMA, and large numbers of military personnel have, in fact, been deployed after recent disasters. The military is commonly in charge of disaster management in third world nations because it is often the only institution capable of providing aid during disasters and one of the few organizations that can be counted on to enforce government policies at other times. But in the United States and other Western nations proposals for a larger military role in civilian affairs are often controversial. Advocates of civil authority and legal due process are concerned that increased military involvement in disasters may signal an erosion of citizen rights and responsibilities, while others point to the reduction in international tensions and the need for more cost-effective national institutions as grounds for making creative use of military expertise in new roles.

Without going into detail, it is worthwhile to note that there is a widespread loss of faith in the ability of national public agencies to combat natural and technological hazards in many other countries. The failures of Soviet agencies in connection with the Armenian earthquake (1988) and the Chernobyl nuclear power station fire (1986) have been well documented and they are believed to have contributed to the collapse of the Soviet government. British civil defence agencies have also been roundly criticized for inadequate preparedness and lack of attention to hazard mitigation (Mitchell, 1989; Parker and Handmer, 1992).

Partly because government agencies have come under attack, there has been a determined effort to shift the burden of disaster management onto private individuals and institutions. In countries such as the United States, Germany, and the United Kingdom, this began with a conservative revolution in politics led by people such as Margaret Thatcher, Ronald Reagan, and Helmut Kohl. In the context of hazard management, policy reforms commonly took the form of insurance schemes (flood insurance, earthquake insurance, crop insurance, etc.), limitations on central government spending for disaster relief and recovery, an end to public subsidies for building in hazardous areas, and penalties for people who wilfully build - or rebuild - in such places.

Now there is a growing body of evidence that such policies may not work as intended. For example, insurance is not the panacea it was once proclaimed to be. Many potential victims are uninsured or underinsured and those who have adequate insurance often experience serious problems securing reimbursements. Not all perils are covered by insurance, and major problems arise when hazards involve several perils (e.g. hurricanes bring floods, erosion, wind damage, landslides, and other events). Cut-backs in government funding of social services have become common throughout the developed world in recent years, and spending on disasters is no exception. As a result, policies that emphasize private responsibilities for hazard management may help to widen the gap between richer and better-educated victims - that is, those who can afford to make provisions for their own security - and the poor or disadvantaged groups that lack such a capacity. In short, a hazard-protection system that relies primarily on market mechanisms may well be detrimental to broader public interests.

British experience with the great storm of 15 October 1987 illustrates several of these problems (Mitchell, Devine, and Jagger, 1989). Before the storm, local governments and private individuals in England had been encouraged to be self-reliant and not to expect the national government to provide recovery funds in the event of a disaster. But the storm - which recorded the highest wind speeds in 250 years - blew down some 15 million trees and inflicted economic losses greater than any natural disaster in Britain since the end of the Second World War. In the process it exposed the limitations of local resources for coping with disasters and it compelled a major reversal of national policies that would have left local governments to take care of natural disasters.

If there is concern about the general effectiveness of disaster management by the private sector, there is deep consternation about the future of hazard insurance systems. Lack of insurance coverage and inadequate reimbursements are continuing problems, but the central issue is that very large disasters may bankrupt the entire international insurance system. Insurance and reinsurance companies in Germany, Japan, and the United States are all deeply troubled by this prospect. These include Lloyd's of London, which ran up losses of US$ 4.2 billion in 1993 as a consequence of heavy pay-outs to victims of earthquakes, hurricanes, and pollution emergencies (New York Times, 30 April 1993). In 1989, the Loma Prieta earthquake (northern California) and hurricane Hugo (Puerto Rico, Virgin Islands, South Carolina) inflicted severe losses on the American insurance industry and the global reinsurance market. But these losses were nothing compared with the losses sustained in 1992 as a result of three "superstorms'' that occurred within a few weeks: typhoon Omar (Guam), hurricane Andrew (Florida), and hurricane Iniki (Hawaii). The total bill for insured storm losses was at least US$ 20 billion. Yet worse is possible - indeed likely. It is estimated that a single large hurricane could cause US$ 30 - 40 billion in insured losses and that a major earthquake in Tokyo or southern California might wreak even worse damage. These are the kinds of event that could disrupt investment markets throughout the world and trigger a collapse of the global financial system. Such circumstances underline the increasing interdependence of global society and its vulnerability to strategic loss.

Failure to make use of available hazard-reduction information and measures of known effectiveness constitutes another general policy issue. It is one that helps to stimulate the ongoing UN-sponsored International Decade for Natural Disaster Reduction (Mitchell, 1988). In many places it would be possible to mitigate losses simply by putting what is known into effect. For example, the value of warning and evacuation systems has been proven repeatedly; yet such systems are often underused. Likewise, hazard-mitigation schemes offer reliable paths toward reducing the long-term costs of disasters but they are often resisted in favour of immediate post-disaster relief, insurance, and compensation programmes. Why do individuals and governments fail to make optimal use of available knowledge? There is no single answer to this question. A large number of factors are involved.

· lack of agreement about definition and identification of problem;

· lack of awareness of hazards;

· misperception or misjudgement of risks;

· lack of awareness of appropriate responses;

· lack of expertise to make use of responses;

· lack of money or resources to pay for responses;

· lack of coordination among institutions;

· lack of attention to relationship between "disasters'' and "development'';

· failure to treat hazards as contextual problems whose components require simultaneous attention (i.e. reciprocity);

· lack of access by affected populations to decision-making;

· lack of public confidence in scientific knowledge;

· conflicting goals among populations at risk;

· fluctuating salience of hazards (competing priorities);

· public opposition by negatively affected individuals and groups.

Underlying all of these specific reasons is a larger problem. It is this: society fails to treat natural hazards as complex systems with many components that often require simultaneous attention. We tinker with one or another aspect of these systems when what is required are system-wide strategies. Perhaps even more important, we fail to address the direct linkage between natural hazard systems and economic investment decisions that drive the process of "development'' and affect the potential for disasters. That such links exist has been known for a very long time:

If a man owes a debt, and the storm inundates his field and carries away the produce, or if the grain has not grown in the field, in that year he shall not make any return to the creditor, he shall alter his contract and he shall not pay interest for that year. (The Code of Hammurabi, King of Babylon, c. 2250 B.C.)

But most of the decisions that are taken to build new facilities or redevelop old ones, or to adopt new production and distribution processes, or to develop new land, or to effectuate a myriad of other development goals are not currently very sensitive to considerations of natural hazards. They should become so. And that is a task that will require a great deal of effort by natural hazard scientists to go beyond the laboratory and the research office or the field study site to acquire an understanding of how best to apply their expertise in public settings. It will also require the users of scientific information about hazards (architects, engineers, planners, banks and mortgage companies, international development agencies, and investment financiers) to nurture a mutually interactive relationship with the scientists who are producers of that information.

"Development'' is only one of the major public issues that overlaps with natural hazards reduction. Others include: environmental management; public health; security (personal, social, and national); and urbanization. All of them are major problem sets in their own right, each patterned by philosophical and managerial disputes and unresolved issues. Efforts to work out mutually supportive policies and programmes raise entirely new sets of contextual issues for hazards experts. How shall the different kinds of policies for these broad problem sets be articulated? Can they be organized around a single concept such as sustainable development, or will it be necessary to make tradeoffs among each of them separately on an ad hoc basis? Some of the experience gained from attempts to work out sustainable development programmes in the World Bank and elsewhere might prove useful here (Cohen and Leitmann, 1994; Goode, 1996; Jones and Ward, 1994; Stren et al., 1992; White, 1994; and Wikan, 1995). So might the experience of "harmonizing'' laws and regulations among countries that join together in new continental-sized free trade blocs (e.g. the European Union or the North American Free Trade Agreement).

The difficulties of merging hazards reduction in the overlapping context of sustainable development are instructive and require further comment. The rising incidence of disasters throughout the world is one of the most commonly cited indicators of non-sustainable development. (The other major villains are loss of biodiversity, impairment of natural systems, and global economic crises.) As Odd Grann, Secretary General of the Norwegian Red Cross, has said:

All major disaster problems in the Third World are essentially unresolved development problems. Disaster prevention is thus primarily an aspect of development, and this must be a development that takes place within the sustainable limits.

The evidence that disasters and development are connected is simply too overwhelming to be ignored. In the case of problems such as drought and famine in the African Sahel or the deaths of poor slum dwellers in landslides that affect the unstable slopes of Latin American cities, the connections between disasters and development are clear. But it is another matter entirely to believe that most or all natural disasters will disappear if only we adopt sustainable development policies and programmes. "Sustainable'' development does not necessarily equal "safe'' development.

It will probably always be necessary to support and nourish institutions that have the capability of responding to unexpected natural disasters. Much of what our existing hazard-management organizations do could be described as preparing for and coping with "routine disasters,'' that is, problems that were bound to occur sooner rather than later and for which we have developed reliable management responses. These are the sorts of things that it is hoped will diminish if sustainable development is properly implemented.

Clearly, the roster of hazard-management issues is packed with still unresolved old issues and emerging new ones. Among the topics that require attention are the following:

· the changing contributions of people, natural systems, and technology to the creation of hazard;

· measures to encourage improved use of available information about hazards (including scientific knowledge and folk wisdom);

· global interdependence and the vulnerability of large institutions (e.g. economies, cities) to major disruptive events;

· the relative human impacts of cumulative small-scale hazards and single large disasters;

· innovative procedures for coping with unprecedented hazards (i.e. surprises);

· attitudes toward risk and hazard;

· equity in the distribution of hazard costs and benefits;

· the illumination of polarizing debates about appropriate hazard-management strategies (e.g. "top - down'' versus "bottom - up,'' centralization versus decentralization, rights versus responsibilities, anticipation versus reaction);

· alternative means for sustaining stakeholder involvement in decision-making beyond periods of acute crisis; and

· coalition-building between hazards interest groups and others that address overlapping problems (e.g. sustainable development).

This is by no means an exhaustive catalogue of potential research opportunities and needs. Note that urbanization or urban hazards and disasters, as such, are not explicitly identified in this list, but they are interwoven with most of these issues and topics.

Changes in the theory of hazards and disasters

The theory of research on natural hazards and disasters has been extensively reviewed elsewhere (Emel and Peet, 1989; Hewitt, 1983; Kates, 1978; Mitchell, 1984, 1987, 1989, 1990, 1991, 1993b; O'Riordan, 1986; White, 1973; Whyte, 1986; Whyte and Burton, 1980). I wish to focus only on recent (i.e. circa post-1988) changes. Two major research paradigms were in existence by the late 1980s: an older, human ecology one, which focused on decision-making by individuals and groups in the face of extreme events, and a newer, political economy paradigm, which emphasized underlying political, economic, and social structures that create conditions of hazard and prevent populations at risk from taking appropriate safety measures. Since 1988, debates between proponents of each paradigm have continued (Emel and Peet, 1989; Hewitt, 1992; K. Smith, 1992; Blaikie et al., 1994), but the two groups have tended to focus on different topics. Human ecologists have carried their ideas forward in a specialized literature on the social amplification of risk and risk communication, particularly in the context of major technological hazards (Greenberg et al., 1989; Wynne, 1989). Political economists have preferred to examine issues of vulnerability and equity, although others have also addressed those themes (Anderson and Woodrow, 1991; Dow, 1992; Kates and Weinberg, 1986; J. Lewis, 1990; Liverman, 1990; Taylor, 1990).

There have been some attempts at intellectual synthesis, including the creation of a new subfield - political ecology (Blaikie and Brookfield, 1987; Dalby, 1992; Little and Horowitz, 1987). Such syntheses incorporate elements of both paradigms and sometimes also include other perspectives (Alexander, 1993; Bryant, 1991; Palm, 1990). At the same time, a number of scholars have sought to forge links between hazards research and overlapping fields, in the hope of providing explanations for a wider range of problems. For example, medical researchers have called for attention to biological and disease hazards that were formerly neglected, and students of geopolitics have underscored the growing importance of environmental disasters as threats to national and global security (Dalby, 1992; Lewis and Mayer, 1988; Mintzer, 1992).

Overall there has been a recognition that broader interpretive frameworks are necessary - frameworks that incorporate both society and nature and a variety of contextual variables (Mitchell, Devine, and Jagger, 1989). However, whereas some have approached the task of extending hazards theory by means of synthesis among existing approaches, others have subjected long-accepted foundational concepts to critical analysis. For example, students of industrial disasters have questioned assumptions about a return to "normal'' that underlie cyclical equilibrium models of disaster (Couch, 1996; Edelstein, 1988). They suggest that some disasters are essentially open-ended processes that may persist indefinitely. Other have criticized a widespread assumption among hazards researchers that public leaders want scientists to help them reach good decisions by eliminating uncertainty. Waterstone counters that the manipulation of uncertainty is a core political device that few leaders willingly surrender (Waterstone, 1993, p. 150). Wynne (1992) has also expressed doubts about the applicability of "uncertainty'' as a basic concept in hazards studies. Drawing on a variety of evidence about recent surprises, he argues that many types of hazards may be better understood as indeterminate phenomena. Such risks as stratospheric ozone depletion, the build-up of greenhouse gases in the atmosphere, and the hazards of new industrial technologies are unprecedented in human experience; they represent a break with past knowledge. By invoking the concept "indeterminacy,'' Wynne and others imply that hazards research is pressing against the limits of scientific methods and scientific philosophies (Wynne, 1992; P. J. Smith, 1990). Funtowicz and Ravetz (1990) have come to similar conclusions by examining the quality of scientific knowledge about global environmental hazards. When assessing emergent types of hazard they judge that we may need to think of several orders of scientific knowledge, each characterized by different degrees of certainty and reliability. O'Riordan and Rayner (1991) carry this analysis further and call for the creation of risk-management institutions that explicitly seek to combine science and politics. O'Riordan (1986) has also gone on record in support of creative use of knowledge from the arts and humanities to inform hazards research and education.

Thoughts like these lie at the interface between science and other forms of knowing. They are part of a general questioning of established knowledge that constitutes "postmodern discourse'' in the social sciences and, especially, the humanities. In a surprising turn of events, hazards research - an interdisciplinary social science with ties to the natural sciences - is now beginning to be fertilized by ideas from the humanities. It is not easy to characterize either postmodernism or the special attributes of postmodernist theory (see Harvey, 1989; Hayles, 1991; Rosenau, 1992). Postmodernism is both a cultural condition and an intellectual construct. As a cultural condition, it is identified with certain architectural and artistic styles that involve the unexpected juxtaposition of forms and characters, transitory states, fluidity, and eclecticism. As an intellectual construct, postmodernism can be viewed as one outcome of a long-running border war between science and the humanities, wherein the allegiance of social scientists has been divided. The basic ideas of postmodernism have been fashioned by literary theorists and other humanists who have been concerned about the excesses and limitations of rational analysis and the neglect of alternative forms of knowing. Their thesis is as follows. The dominant global society is a product of the European Enlightenment. This "modern'' society has endured for several hundred years, founded on beliefs about rationality in the organization and functioning of both the natural and social worlds and on the capacity of people to discover these truths and to act in light of them for their own benefit (i.e. to achieve "progress''). It is now argued that modernity has run its course and is being replaced by a fundamentally different "postmodern'' era.

For social scientists, the salient characteristics of postmodern thought are: its rejection of grand theory and so-called metanarrative; its adherence to pluralist interpretations; its characterization of change as discontinuity; its celebration of difference, contingency, and the primacy of local experience; its use of paradox and metaphor as investigative tools; and its non-pejorative assessment of alternative (i.e. non-scientific) forms of knowing. In the words of one commentator:

Post-modernists rearrange the whole social science enterprise. Those of a modern conviction seek to isolate elements, specify relationships, and formulate a synthesis; post-modernists do the opposite. They offer indeterminacy rather than determinism, diversity rather than unity, difference rather than synthesis, complexity rather than simplification. They look to the unique rather than the general, to intertextual relations rather than causality, and to the unrepeatable rather than the re-occurring, the habitual, or the routine. Within a post-modern perspective social science becomes a more subjective and humbling enterprise as truth gives way to tentativeness. Confidence in emotion replaces efforts at impartial observation. Relativism is preferred to objectivity, fragmentation to totalization. Attempts to apply the model of natural science inquiry in the social sciences are rejected because post-modernists consider such methods to be part of the larger techno-scientific corrupting cultural imperative, originating in the West but spreading out to encompass the planet. (Rosenau, 1992)

Few hazards researchers have self-consciously placed their work in a postmodern context, but there are clear echoes of postmodern sensibility in many recent critiques of the field (Cooke, 1992; Curry, 1988; Taylor and Buttel, 1992). The following statement by Hewitt is a good example.

[The] elements of today's dominant view of hazards... are fully symptomatic [of ] the abstracted, gender-blind, class-blind, secular and amoral, mechanistic and technocratic style of work that prevails in international "Big Science'' and the preoccupations of agencies and publics in the wealthier and more powerful centers.... They not only display the Enlightenment desire to generalize and command "rational'' interpretation by reducing everything to a common language and set of concepts. They are repeatedly beguiled by the most "advanced'' or fashionable concepts, the worries of "fast-breaking'' technologies in their own places of origin, and a belief these define and alone can solve "problems'' elsewhere. (Hewitt, 1992)

Hewitt's statement is an indictment of current theory as well as current practice in the management of hazards. He is not alone in holding such views. This is not to say that contemporary hazards research is riven by postmodern critiques. The challenges that I have reported here are not yet a dominant presence in the field. But they are beginning to tap into a distinct undercurrent of disquiet among many who are conscious of the limitations of existing theory and practice (Browning and Shetler, 1992).

Whether postmodernist critiques will fundamentally alter the nature of hazards research is an open question. At first glance there are grounds for scepticism. The existence of close links between research and praxis suggests that there will be strong counterpressures to resist any theory that calls for thoroughgoing institutional and social change - and evinces deep suspicion of the very organizations that are most involved with the making and execution of policy. Of course, current debates about hazards need not lead to postmodern theories of hazard that blend the contributions of scientists and humanists. Other purely scientific ("modern'') approaches are possible, including those based on theories of evolution, chaos, and complexity (Argyros, 1991; Waldrup, 1992). (For example, some Russian colleagues are developing ideas about processes that limit the disturbance of large systems that are based on the application or extension of the well-known Le Chatelier Principle in chemistry.) But hazards researchers have been slow to link up with this more formal work, most of which is firmly in the tradition of Enlightenment thought. It is too soon to assess the advantages that a "science of chaos and complexity'' might offer to hazards scholars.

Mega-cities and natural disasters reconsidered

In this chapter I have attempted to take a very broad conceptual stance towards the subject of mega-cities and natural disasters. This is deliberate because I am convinced that the task of reducing mega-city disasters is not simply one of applying existing practical knowledge that scientists, engineers, and hazards professionals - myself among them - have gained from years of working with those events. If we are to reap success, a commitment to mutual understanding and collaboration among academics, professionals, and laypersons who are hazard specialists and academics, professionals, and laypersons who are urban specialists will have to be made. Because of the changes that are now taking place in the study and management of natural hazards and the study and management of cities, and also because new forms of hazard and new types of urban phenomena (i.e. mega-cities) are emerging, the road ahead is not well signposted for either set of interest groups. So we must proceed like two well-founded cultures that are coming into contact for the first time.

This is a process with which Japanese society has had considerable experience from the late nineteenth century onwards. And the Japanese experience offers instructive parallels, both in a general sense and in a more restricted one that applies specifically to issues of hazard and disaster. After the Meiji restoration, contacts between Japan and the rest of the world were managed in such a way as to permit Japan to benefit while at the same time preventing Japanese society from being inundated by foreign cultural influences. A successful transition to a new social order was obtained. As part of that transition, Western science came to Japan and with it an interest in the scientific and technical analysis of natural hazards and disasters. Out of this arose leading institutions of disaster research and a distinctive style of hazards management that attracts visitors from all over the world. Whether one learns about the means by which Kagoshima beneficially coexists with its hyperactive backyard volcano, or the gated rivers of the Japanese coast, or Nagoya's three-tiered elevated highway, ground-level park, and underground flood reservoir scheme, or Tokyo's firewall high-rise complex, there can be no mistaking the fact that a very distinctive blend of scientific knowledge, technological innovation, and societal cooperation has been achieved in a variety of big-city contexts.9

9 Many critics might contend that this blend is not readily transferable elsewhere without major modifications.

In coming to grips with the burgeoning disaster potential of the world's mega-cities, the hazards community and the urban community are embarking on a somewhat similar exercise in cultural contact, this time among professional and academic cultures as well as national ones. We cannot know how the contact will turn out, but we can make sure that the best thinking from each community is incorporated into the process.

REFERENCES

Abu Sin, M. E. 1991. "Migration from Eastern Gezira into Greater Khartoum - A case study in rural - urban migration and population integration processes.'' GeoJournal 25(1): 73 - 80.

Agmad, Adil Mustafa. 1993. "The neighborhoods of Khartoum: Reflections on their functions, forms and future.'' Habitat 16(4): 27 - 45.

Aguilar, Adrian Guillermo, Exequiel Ezcurra, Teresa Garcia, Marisa Mazari-Hiriat, and Irene Pisanty. 1995. "The basin of Mexico.'' In Jeanne X. Kasperson, Roger E. Kasperson, and B. L. Turner II (eds.), Regions at risk. Tokyo: United Nations University Press, pp. 304 - 366.

Alexander, David. 1993. Natural disasters. New York: Chapman & Hall.

Anderson, Mary B. and Peter J. Woodrow. 1991. "Reducing vulnerability to drought and famine: Developmental approaches to relief.'' Disasters 15(1): 43 - 54.

Argyros, Alexander J. 1991. A blessed rage for order: Deconstruction, evolution, and chaos. Ann Arbor: University of Michigan Press.

Bakhit, Abdel Hamid. 1994. "Availability, affordibility and accessibility of food in Khartoum.'' GeoJournal 34(3): 253 - 255.

Berke, Philip R. and Timothy Beatley. 1992. Planning for earthquakes: Risk, politics, and policy. Baltimore, MD: Johns Hopkins University Press.

Berry, Brian J. L. 1990. "Urbanization.'' In B. L. Turner II, William C. Clark, Robert W. Kates, John F. Richards, Jessica T. Mathews, and William B. Meyer (eds.), The earth as transformed by human action: Global and regional changes in the biosphere over the past 300 years. Cambridge: Cambridge University Press, pp. 103 - 119.

Blaikie, P. and H. Brookfield. 1987. Land degradation and society. London: Methuen.

Blaikie, Piers, Terry Cannon, Ian Davis, and Ben Wisner. 1994. At risk: Natural hazards, people's vulnerability and disasters. London: Routledge.

Bohle, Hans-Georg. 1994. "Cities of hunger: Urban food systems in developing countries.'' GeoJournal 34(3): 243 - 244 (see also additional papers, pp. 245 - 304).

Boone, Christopher G. 1996. "Language politics and flood control in nineteenth- century Montreal.'' Environmental History 1(3): 70 - 85.

Bromley, Ray, Peter Hall, Ashok Dutt, Debnath Mookherjee and John E. Benhart. 1989. "Regional development and planning.'' In Gary L. Gaile and Cort J. Willmott (eds.), Geography in America. Columbus, OH: Merrill Publishing Company, pp. 351 - 386.

Browning, Larry D. and Judy C. Shetler. 1992. "Communication in crisis, communication in recovery: A postmodern commentary on the Exxon Valdez disaster.'' International Journal of Mass Emergencies and Disasters 10(2).

Brunn, Stanley D. and Jack F. Williams. 1993. Cities of the world: World regional development, 2nd edn. New York: HarperCollins.

Bryant, E. A. 1991. Natural hazards. Cambridge: Cambridge University Press.

Chandler, Tertius and Gerald Fox. 1974. 3000 years of urban growth. New York: Academic Press.

Chudacoff, Howard P. and Judith E. Smith. 1994. The evolution of American urban society. Englewood Cliffs, NJ: Prentice-Hall.

Cohen, Michael A. and Josef L. Leitmann. 1994. "Will the World Bank's real 'New Urban Policy' please stand up?'' Habitat International 18(4): 117 - 126.

Cooke, Ronald U. 1984. Geomorphological hazards in Los Angeles. London: Allen & Unwin.

Cooke, Ronald U. 1992. "Common ground, shared inheritance: Research imperatives for environmental geography.'' Transactions of the Institute of British Geographers 17: 131 - 151.

Couch, S. R. 1996. "Environmental contamination, community transformation, and the Centralia mine fire.'' In James K. Mitchell (ed.), The long road to recovery: Community responses to industrial disaster. Tokyo: United Nations University Press, pp. 60 - 85.

Cronon, William. 1991. Nature's metropolis: Chicago and the Great West. New York and London: W. W. Norton.

Cuny, Frederick C. 1994. "Cities under seige: Problems, priorities and programs.'' Disasters 18(2): 152 - 159.

Curry, Michael. 1988. "Commentary on 'Geographers and nuclear war: Why we lack influence on public policy'.'' Annals of the Association of American Geographers 78(4): 720 - 724.

Dalby, Simon. 1992. "Ecopolitical discourse: 'Environmental security' and political geography.'' Progress in Human Geography 16(4): 503 - 522.

Dando, W. A. 1994. Urban famines in a future world of mega-cities: A commentary. Professional Paper No. 20. Indiana State University, Terre Haute, Department of Geography and Geology, pp. 81 - 87.

Davis, Mike. 1992. City of quartz: Excavating the future in Los Angeles. New York: Vintage Books.

Degg, Martin. 1993. "The 1992 'Cairo Earthquake': Causes, effect and response.'' Disasters 17(3): 226 - 238.

Dow, Kirstin. 1992. "Exploring differences in our common future(s): The meaning of vulnerability to global environmental change.'' Geoforum 23(3): 417 - 436.

Driever, Steven L. and Danny M. Vaughn. 1988. "Flood hazard in Kansas City since 1880.'' Geographical Review 78(1): 1 - 19.

Earthquake Engineering Research Institute. 1985. Proceedings - U.S. - Japan Workshop on Urban Earthquake Hazards Reduction. Publication No. 85 - 03. Berkeley: EERI.

Edelstein, M. R. 1988. Contaminated communities: The social and psychological impacts of residential toxic exposure. Boulder, CO: Westview Press.

El-Bushra, El-Sayed and Naila B. Hijazi. 1995. "Two million squatters in Khartoum urban complex: The dilemma of Sudan's national capital.'' GeoJournal 35(4): 505 - 514.

Emel, J. and J. R. Peet. 1989. "Resource management and natural hazards.'' In J. R. Peet and N. Thrift (eds.), New models in geography: The political-economy perspective. London: Unwin Hyman.

Ezcurra, Exequiel and Marisa Mazari-Hiriart. 1996. "Are megacities viable? A cautionary tale from Mexico City.'' Environment 38(1): 15, 26 - 35.

Fellmann, Jerome, Arthur Getis, and Judy Getis. 1992. Human geography: Landscapes of human activities. Dubuque: Wm. C. Brown.

Firman, Tommy and Indira Dharmapatni. 1994. "The challenges to sustainable development in Jakarta Metropolitan Region.'' Habitat International 18(3): 74 - 94

Fishman, Robert. 1987. Bourgeois utopias: The rise and fall of suburbia. New York: Basic Books.

Frater, Alexander. 1991. Chasing the monsoon. New York: Alfred A. Knopf.

Funtowicz, S. O. and J. R. Ravetz. 1990. Uncertainty and quality in science for policy. Dordrecht: Kluwer Academic Publishers.

Gaile, Gary L. 1983. "Reanalyses of Chinese spatial inequality.'' Professional Geographer 35: 281 - 283.

Gilbert, Alan and Josef Gugler. 1992. Cities, poverty and development: Urbanization in the third world. Oxford: Oxford University Press.

Goode, Judith. 1996. "On sustainable development in the mega-city.'' Current Anthropology 36(1): 131 - 133.

Greenberg, Michael R., Peter M. Sandman, David B. Sachsman, and Kandice L. Salmore. 1989. "Network television news coverage of environmental risks.'' Environment 31(2): 16 - 20, 40 - 44.

Guterbock, Thomas M. 1990. "The effect of snow on urban density patterns in the United States.'' Environment and Behavior 22(3): 358 - 386.

Hall, Peter. 1988. Cities of tomorrow: An intellectual history of urban planning and design in the twentieth century. Oxford: Basil Blackwell.

Hansel, G. 1991. "Rural - urban migration and the problem of town planning - The case of Um Badda, Omdurman.'' GeoJournal 25(1): 27 - 30.

Harvey, D. 1989. The condition of postmodernity: An enquiry into the origins of cultural change. Oxford: Blackwell.

Havlick, Spencer W. 1986. "Third world cities at risk: Building for calamity.'' Environment 28(9): 6 - 11, 41 - 45.

Hayles, Katherine, ed. 1991. Chaos and order: Complex dynamics in literature and science. Chicago: University of Chicago Press.

Hewitt, Kenneth. 1983. Interpretations of calamity. Boston: Allen & Unwin.

- 1992. "Mountain hazards.'' GeoJournal 27(1): 47 - 60.

Ibrahim, F. N. 1991. "The southern Sudanese migration to Khartoum and the resultant conflicts.'' GeoJournal 25(1): 13 - 18.

Ibrahim, F. N. 1994. "Hunger-vulnerable groups within the metropolitan food system of Khartoum.'' GeoJournal 34(3): 257 - 261.

Ibrahim, F. N. 1995. "Cultural change and the chances of reintegration after reimmigration to home areas - The case of the Southern Sudanese dislocated population in Greater Khartoum.'' GeoJournal 36(1): 103 - 107.

IDNDR (International Decade for Natural Disaster Reduction). 1990. International Conference 1990 Japan: Toward a less hazardous world in the 21st century: Proceedings. September 27 - October 3, 1990, Yokohama and Kagoshima.

IDNDR (International Decade for Natural Disaster Reduction). 1996. Cities at risk: Making cities safer... before disaster strikes. Supplement to No. 28, Stop Disasters. Geneva: IDNDR.

Institution of Civil Engineers. 1995. Megacities: Reducing vulnerability to natural disaster. London: Thomas Telford.

Jairo-Cardenas, J. 1990. "Natural disasters and urban struggle for housing.'' Architecture et Comportement/Architecture and Behaviour 6(2): 177 - 191.

Jones, Barclay and William A. Kandel. 1992. "Population growth, urbanization, disaster risk, and vulnerability in metropolitan areas: A conceptual framework.'' In Alcira Kreimer and Mohan Munasinghe (eds.), Environmental management and urban vulnerability. World Bank Discussion Paper 168. Washington, D.C.: World Bank, pp. 51 - 76.

Jones, Emrys. 1990. Metropolis. New York: Oxford University Press.

Jones, Gareth A. and Peter M. Ward. 1994. "The World Bank's 'new' Urban Management Programme: Paradigm shift or policy continuity?'' Habitat International 18(3): 33 - 51.

Kates, Robert W. 1978. Risk assessment of environmental hazard. SCOPE (Scientific Committee on Problems of the Environment) Report No. 8. New York: Wiley.

Kates, Robert W. and Alvin M. Weinberg, eds. 1986. Hazards: Technology and fairness. Washington, D.C.: National Academy Press.

Kim, Kwi-Gon. 1990. "Risk assessment in urban planning and management: A metropolitan example.'' Habitat 14(1): 177 - 190.

Kreimer, Alcira and Mohan Munasinghe, eds. 1992. Environmental management and urban vulnerability. World Bank Discussion Paper 168. Washington, D.C.: World Bank.

Lewis, James. 1990. "The vulnerability of small island states: The need for holistic strategies.'' Disasters 14(3): 241 - 249.

Lewis, Michael. 1989. "How a Tokyo earthquake could devastate Wall Street and the world economy.'' Manhattan, inc., June: 69 - 79.

Lewis, Nancy D. and Jonathan D. Mayer. 1988. "Disease as natural hazard.'' Progress in Human Geography 12(1): 15 - 33.

Little, Peter D. and Michael M. Horowitz, eds. 1987. Lands at risk in the Third World: Local level perspectives. Boulder, CO: Westview Press.

Liverman, Diana M. 1990. "Drought impacts in Mexico: Climate, agriculture, technology, and land tenure in Sonora and Puebla.'' Annals of the Association of American Geographers 80(1): 44 - 72.

McGranahan, Gordon and Jacob Songsore. 1994. "Wealth, health, and the urban household: Weighing environmental burdens in Accra, Jakarta, and Sao Paulo.'' Environment 36(6): 4 - 11, 40 - 45.

McHarg, Ian. 1969. Design with nature. New York: Natural History Press (especially "The City: Health and pathology,'' pp. 187 - 195).

McPhee, John. 1989. The control of nature. New York: Farrar, Straus & Giroux.

Main, Hamish and Stephen Wyn Williams, eds. 1994. Environment and housing in Third World cities. Chichester: Wiley.

Maskrey, Andrew. 1989. Disaster mitigation: A community based approach. Development Guidelines No. 3. Oxford: Oxfam.

Masure, P. 1994. "Risk management and preventive planning in mega-cities: A scientific approach for action.'' Regional Development Dialogue 15(2): 183 - 188.

Miller, Ross. 1990. American apocalypse: The great fire and the myth of Chicago. Chicago: University of Chicago Press.

Mintzer, Irving M. 1992. Confronting climate change: Risks, implications and responses. Cambridge: Cambridge University Press.

Mitchell, James K. 1984. "Hazard perception studies: Convergent concerns and divergent approaches during the past decade.'' In Thomas F. Saarinen, David R. Seamon, and James L. Sell (eds.), Environmental perception and behavior: An inventory and prospect. Department of Geography Research Paper No. 209. Chicago: University of Chicago, pp. 33 - 59.

Mitchell, James K. 1987. "Boba so stikhiynymi bedstviyami'' [Human responses to natural hazards]. Geograficheskie aspetky vzsimodeystviya khozyaystva i okruzhayushchev sredy, akademiya nauk SSSR, Moscow, pp. 178 - 187.

Mitchell, James K. 1988. "Confronting natural disasters: An international decade for natural hazard reduction.'' Environment 30(2): 25 - 29.

Mitchell, James K. 1989. "Hazards research.'' In Gary Gaile and Cort Willmott (eds.), Geography in America. Columbus, OH: Merrill Publishing Company, pp. 410 - 424.

Mitchell, James K. 1990. "Human dimensions of environmental hazards: Complexity, disparity and the search for guidance.'' In Andrew Kirby (ed.), Nothing to fear: Risks and hazards in American society. Tucson: University of Arizona Press, pp. 131 - 178.

Mitchell, James K. 1991. "Hazards geography.'' In Gary S. Dunbar (ed.), Modern geography: An encyclopedic survey. New York: Garland, p. 76.

Mitchell, James K. 1992. "Major natural disasters in the Pacific Basin: Status, trends and emerging issues.'' International Journal of Mass Emergencies and Disasters 10(2): 269 - 279.

Mitchell, James K. 1993a. "Natural hazards predictions and responses in very large cities.'' In J. Nemec et al. (eds.), Prediction and perception of natural hazards. Dordrecht: Kluwer Academic Publishers, pp. 29 - 37.

Mitchell, James K. 1993b. "Recent developments in hazards research: A geographer's perspective. In E. L. Quarantelli and K. Popov (eds.), Proceedings of the United States - Former Soviet Union Seminar on Social Science Research on Mitigation for and Recovery from Disasters and Large Scale Hazards. Moscow, April 19 - 26, 1993. Vol. I: The American participation. Newark: University of Delaware, Disaster Research Center, pp. 43 - 62.

Mitchell, James K. 1996. "Improving community responses to industrial disasters: In James K. Mitchell (ed.), The long road to recovery: Community responses to industrial disaster. Tokyo: United Nations University Press.

Mitchell, James K. and Neil Ericksen. 1992. "Effects of climate changes on weather-related disasters.'' In Irving Mintzer (ed.), Confronting climate change: Risks, implications and responses. Cambridge: Cambridge University Press, pp. 141 - 152.

Mitchell, James K., Neal Devine, and Kathleen Jagger. 1989. "A contextual model of natural hazard.'' Geographical Review 89(4): 391 - 409.

Myers, Mary Fran and Gilbert F. White. 1993. "The challenge of the Mississippi flood.'' Environment 35(10): 6 - 9, 25 - 35.

National Research Council. 1987. Geophysical predictions. Washington, D.C.: Geophysics Study Committee.

Organizing Committee of the International Seminar on Regional Development Planning for Disaster Prevention. 1986. Planning for crisis relief: Towards comprehensive resource management and planning for natural disaster prevention. Vol. 3. Planning and management for the prevention and mitigation from natural disasters in metropolis. Nagoya: United Nations Centre for Regional Development.

O'Riordan, Timothy. 1986. "Coping with environmental hazards.'' In Robert W. Kates and Ian Burton (eds.), Themes from the work of Gilbert F. White, vol. 2 of Geography, resources, and environment. Chicago: University of Chicago Press, pp. 272 - 309.

O'Riordan, Timothy and Steve Rayner. 1991. "Risk management for global environmental change.'' Global Environmental Change 1(2): 91 - 108.

Palm, Risa. 1990. Natural hazards: An integrative framework for research and planning. Baltimore, MD: Johns Hopkins University Press.

Parker, D. J. and J. W. Handmer, eds. 1992. Hazard management and emergency planning: Perspectives on Britain. London: James & James.

Parker, Dennis J. and James K. Mitchell. 1995. "Disaster vulnerability of mega-cities: An expanding problem that requires rethinking and innovative responses.'' GeoJournal 37(3): 295 - 301 (see also additional papers, pp. 303 - 388).

Perlman, Janice E. 1987. "Megacities and innovative technologies.'' Cities, May: 128 - 136.

Reed, Stuart, Fred Tromp, and Alain Lam. 1992. "Major hazards - Thinking the unthinkable.'' Environmental Management 16(6): 715 - 722.

Revue de Géographie Alpine. 1994. "Croissance urbaine et risques naturels dans les montagnes des pays en développement,'' No. 4.

Rosen, Christine Meisner. 1986. The limits of power: Great fires and the process of city growth in America. New York: Cambridge University Press.

Rosenau, P. M. 1992. Post-modernism and the social sciences: Insights, inroads, and intrusions. Princeton, NJ: Princeton University Press.

Ruppert, Helmut. 1991. "The responses of different ethnic groups in the Sudan to rural - urban migration.'' GeoJournal 25(1): 7 - 12.

Sachs, Ignacy. 1988. "Vulnerability of giant cities and the life lottery.'' In Mattei Dogan and John D. Kasarda (eds.), A world of giant cities, vol. 1 in The metropolis era. Newbury Park, CA: Sage Publications, pp. 337 - 350.

Sassen, Saskia. 1991. The global city: London, New York, Tokyo. Princeton, NJ: Princeton University Press.

Science Council of Japan. 1989. International Decade for Natural Disaster Reduction: Proposals by Japanese scientists. Committee for Disaster Research.

Seidensticker, Edward. 1983. Low City, High City: Tokyo from Edo to the earthquake. New York: Alfred A. Knopf.

Seidensticker, Edward. 1990. Tokyo rising: The city since the great earthquake. New York: Alfred A. Knopf.

Setchell, C. A. 1995. "The growing environmental crisis in the world's mega-cities: The case of Bangkok.'' Third World Planning Review 17(1): 1 - 18.

Showalter, Pamela S. and Mary F. Myers. 1994. "Natural disasters in the United States as release agents of oil, chemicals or radiological materials between 1980 - 1989: Analysis and recommendations.'' Risk Analysis 14(2): 169 - 182.

Sit, Victor F. S. 1988. Chinese cities: The growth of the metropolis since 1949. Oxford: Oxford University Press.

Smith, Carl. 1995. Urban disorder and the shape of belief: The great Chicago fire, the Haymarket bomb and the model town of Pullman. Chicago: University of Chicago Press.

Smith, Keith. 1992. Environmental hazards: Assessing risk and reducing disaster. London and New York: Routledge.

Smith, P. J. 1990. "Redefining decision: Implications for managing risk and uncertainty.'' Disasters 14(3): 230 - 240.

Somma, Mark, 1991. "Ecological flight: Explaining the move from country to city in developing nations.'' Environmental History Review, Fall: 1 - 26.

Sorkin, Michael, ed. 1992. Variations on a theme park: The new American city and the end of public space. New York: Hill & Wang.

Steedman, S. 1995. "Megacities: The unacceptable risk of natural disaster.'' Built-Environment 21(2 - 3): 89 - 93.

Stren, Richard, Rodney White, and Joseph Whitney, eds. 1992. Sustainable cities: Urbanization and the environment in international perspective. Boulder, CO: Westview Press.

Sylves, Richard T. and William L. Waugh, Jr., eds. 1990. Cities and disaster: North American studies in emergency management. Springfield, IL: Charles C. Thomas.

Taylor, Antony J. 1990. "A pattern of disasters and victims.'' Disasters 14(4): 291 - 300.

Taylor, Peter J. and Frederick H. Buttel. 1992. "How do we know we have global environmental problems? Science and the globalization of environmental discourse.'' Geoforum 23(3): 405 - 416.

Teune, Henry. 1988. "Growth and pathologies of giant cities.'' In Mattei Dogan and John D. Kasarda (eds.), A world of giant cities, vol. 1 in The metropolis era. Newbury Park, CA: Sage Publications, pp. 351 - 376.

Tuan, Yi-Fu. 1979. Landscapes of fear. Minneapolis: University of Minnesota Press.

Tuan, Yi-Fu. 1988. "The city as a moral universe.'' Geographical Review 78(2): 316 - 324.

UNEP/WHO (UN Environment Programme/World Health Organization). 1994. "Air pollution in the world's megacities.'' Environment 36(2): 4 - 13, 25 - 37.

United Nations. 1987. The prospects of world urbanization. Revised as of 1984 - 85. ST/ESA/SER.A/101. New York: United Nations.

US Office of Science and Technology Policy. 1992. Science and technology: A report to Congress. Washington, D.C.: Executive Office of the President.

Waldrup, M. Mitchell. 1992. Complexity: The emerging science at the edge of order and chaos. New York: Simon & Schuster.

Walsh, R. P. D., H. R. J. Davies, and S. B. Musa. 1994. "Flood frequency and impacts at Khartoum since the early nineteenth century.'' Geographical Journal 160(3): 266 - 279.

Waterstone, Marvin. 1993. "Adrift on a sea of platitudes: Why we will not resolve the greenhouse issue.'' Environmental Management 17(2): 141 - 152.

White, Gilbert F. 1973. "Natural hazards research.'' In Richard J. Chorley (ed.), Directions in geography. London: Methuen.

White, Rodney R. 1994. Urban environmental management: Environmental change and urban design. Chichester: John Wiley.

Whyte, Anne V. 1986. "From hazard perception to human ecology.'' In Robert W. Kates and Ian Burton (eds.), Themes from the work of Gilbert F. White, vol. 2 of Geography, resources, and environment. Chicago: University of Chicago Press, pp. 240 - 271.

Whyte, Anne V. and Ian Burton. 1980. Environmental risk assessment. SCOPE (Scientific Committee on Problems of the Environment) No. 15. New York: Wiley.

Wikan, Unni. 1995. "Sustainable development in the megacity: Can the concept be made applicable?'' Current Anthropology 37(1): 635 - 655.

Woodruff, Bradley A. et al. 1990. "Disease surveillance and control after a flood: Khartoum, Sudan, 1988.'' Disasters 14(2): 151 - 163.

Wu Qingzhou. 1989. "The protection of China's ancient cities from flood damage.'' Disasters 13(3): 193 - 226.

Wynne, Brian. 1989. "Sheep farming after Chernobyl: A case study in communicating scientific information.'' Environment 31(2): 33 - 39.

Wynne, Brian. 1992. "Uncertainty and environmental learning: Reconceiving science and policy in the preventive paradigm.'' Global Environmental Change 2(2): 111 - 127.

Yath, A. Y. 1991. "The effect of differential access to accommodation on the Dinka migrants in Khartoum - The example of Gereif West.'' GeoJournal 25(1): 19 - 26.

Yath, A. Y. 1995. "On the expulsion of rural inmigrants from Greater Khartoum - The example of the Dinka in Suq el Markazi.'' GeoJournal 36(1): 93 - 101.

Yeung, Y. M. 1990. Changing cities of Pacific Asia: A scholarly interpretation. Hong Kong: Chinese University Press.

Zelinsky, W. and L. Kosinski, L. 1991. Emergency evacuation of cities. London: Unwin Hyman.