Environmental hazards have always existed; we may think of the biblical account of The Deluge where people saw the Wrath of God as a punishment for their sins. Nowadays, human beings mainly rely on physical explanations rather than “Acts of God”. However, there is a paradox between modernity, achievements in science and medicine and the human inability to minimize natural hazard impact.

Images shown in media give the impression that hazards cause more and more damage. What about scientific reports? Is natural hazard impact getting worse worldwide and what are the reasons that explain these trends”?A retrospective view over the past decades is necessary to compare hazard impact between past and present but how far should we go back in history? On the one hand, it is worth asking if the world is becoming a more dangerous place. The hypothesis of a physical evolution of natural hazards may explain such a trend. On the other hand, is the human society becoming somehow more vulnerable to hazard? The aim of this paper is to demonstrate that reasons are more “human” than “physical” since the ideas of “hazard impact” and “humans” are strongly linked.First, the focus will be on a physical explanation of hazard impact.

Global climate change and hazard characteristics will be developed. The idea of a single trend that would apply worldwide will be put to the question since there is a spatial variability of impact. Then, after explaining human interaction with its environment and its link with hazard impact definition, it will be determined which aspects are getting worse. Eventually, the emphasis will be on a failure in hazard management owing to either a lack of money or to human perception of hazard.A general increase of natural hazard impact?From a physical viewpoint, global climate change introduces new physical patterns such as increased erosion because of deforestation..

. Then, men have to cope with this modified environment. Global warning brings about more and more flooding along coasts and around islands since the sea level rises (Bangladesh, Pays Bas, Maldives).

Then, human beings have the impression that catastrophes are increasing. From October 2000 to April 2001, a large part of British Isles was affected by the most widespread flooding in over 50 years. Moreover, storm frequency has started to increase in Europe between Christmas and 1999-2000 New Year.However, John Whittow states, “this is not to say that natural hazard are proliferating” (p309). It means that there is not a quantitative increase of hazard occurrences. On the one hand, the number of natural hazards seems to be higher due to the rise of technology that enables men to detect disasters more frequently. Hazards were less registered in the past but it does not mean that they were less numerous. On the other hand, as soon as a natural event happens, it is displayed in media; so people get more information about hazard phenomena.

However, does a better communication reflect reality? For instance, television coverage does no always reflect the spatial variability of hazard occurrences. Moreover, media tend to stress the most spectacular events. Figures are used to point out the catastrophic aspect and people pay attention to the most striking ones. Media are powerful sources of information but people’s perception of hazard can be easily distorted, especially when direct experience of hazard is lacking.

In a nutshell, even if humans are more informed, it does not necessary imply that hazard impact is getting worse. So, the rise of hazard impact cannot be reduced to an increasing number of events.Moreover, qualitative characteristics of hazards such as speed of onset, magnitude (measured with Richter scale), areal extent or duration must be taken into consideration. Extreme events have already occurred in the past (Lisbon Earthquake in 1755..

.). So, they are not new phenomena.

But, is their intensity getting worse? It is important to consider the most striking ones such as earthquakes. Indeed, they provoke an extreme stress and very damaging consequences. For instance, the cost of Agnes earthquake in 1972 represented half the annual total. Loss seems to be bigger when stress is extreme as it is actually concentrated in a short time period.However, the impact of an event is not always linked to its size.

For instance, floods are less spectacular than earthquakes because stress is not so extreme. They are more regular and cost is actually spread through time but it does not mean that total cost is lower. Indeed, flooding “cause the most frequent disasters”1. Losses may also be spread unevenly between different sorts of hazards: in 1996, deaths caused by flooding represented more than half of death total. Yet, total cost of flooding may be uneasy to detect because indirect losses may be considerable.

Moreover, difference between the effects of drought and endemic seasonal hunger and malnutrition in LDCs is not obvious. To sum it up, a mere hazard classification is not enough to determine the most damaging hazards and which one presents the worst evolution. An impact is actually defined through a series of physical parameters combined together.Is hazard impact spread worldwide or does it apply to specific regions? For example, coasts are more and more affected by flooding. Hazard impact is spread unevenly over different parts of the world as the following graph highlights it (UNESCO website). Asia is the most affected.”1996 deaths outcome”, repartition according to continentsThe spatial variability of hazard impact has been brought into relief. It means that hazards are not getting worse equally from a geographical viewpoint.

However, the growing interdependence between individuals spread the impact “far outside the immediate area of impact”2. So, even if regions are more or less affected, hazard impact tends to become a worldwide concern.The discrepancy between developing and developed countries puts to the fore a physical explanation of hazard impact. It mainly concerns infrastructures in DCs whereas it also causes deaths and injuries in LDCs. Some countries or regions are more affected than others (flooding in Bangladesh, volcanoes in Philippines…

). In the structural paradigm of hazard, disasters are seen as a factor in the “growing gap between rich and poor” countries3. Hazard impact is then seen as a reason for underdevelopment. But the nature of an event remains the same wherever it occurs and the impact crosses frontiers. Moreover, California is one the wealthiest region in the USA despite San Andreas Fault. Consequently, physical determinism is not a means of explaining hazard impact. Impact consists in a complex system that involves interactions between Nature and Man.

Environmental change exists because of human interference.Indeed, there is a change of risk pattern due to human interference. Mankind control of environment (resource use, deforestation) modifies geophysical and biophysical systems. The worldwide urbanisation and industrialisation phenomena intensify land pressure.

Thus, constructions and settlements in low valued hazard-prone areas such as floodplains are proliferating. Growing population heightens the impact because the number of people and buildings concerned by risk is larger.Human interferenceBurton and Kates’s definition of natural hazards links them to human systems: “those elements of the physical environment harmful to Man and caused by forces extraneous to him” (1994)4.

A natural event that occurs in a desert area is seen as a physical process whereas it is perceived as a “hazard” when its threat concerns humans. Hazard impact is assessed according to human presence, settlements and activities. It is not seen as getting worse in a desert area because men are not involved. Indeed, if a hazard-prone area is inhabited, nobody cares about the impact since its consequences are likely to be low. Moreover, media, scientists and insurers take hazard impact less in consideration when man is not affected by a disaster.In addition, men define the degree of hazard impact since they give values to things. So, a hierarchy of the consequences of an extreme event is established.

It implies that the statement of hazard impact is not necessary objective. The following diagram shows the severity of an event according to environment, goods or life5. Risk is more important concerning life and goods and consequences on the environment are put in the background. So hazard impact is getting worse according to the type of loss considered.

Relationships between the severity of environmental hazards, probability and risk (source: After Moore, 1983)Human life seems to be the main criteria of hazard impact. Yet, as far as DCs are concerned, natural event impact is still said to be higher than before even if the number of victims has been reduced over the past decades. Property market value losses, costs of recovery are burning issues. Does it imply that hazard impact is mainly an economic concern? Indeed, when it is asserted that the impact is getting worse, it is often connected with money it requires: insurance and compensation rates, taxes..

. The cost of damage is said to be low if the property market value is low.Moreover, hazard impact is synonymous with the negative idea of “loss” or “cost” when it is getting worse (loss of deaths, money..

.). Indirect losses such as deaths, disease after an event must be included in a Cost Benefit Analysis. In addition, trauma felt by individuals could be considered as an indirect loss. It is more difficult to estimate since they concern humans’ perception of impact; they are different from economic cost.Nonetheless, it is sometimes difficult to study the evolution of hazard impact. Moreover, data may be uncertain, inaccurate or unavailable.

Data established before the beginning of the century can be incomplete and useless to compare trends between past and present. So, it is essential to gather reliable database. For instance, there are no available details about flooding on the Chesil Beach in 1903 in the report by Leland (1546) and Camden (1590). Moreover, figures can be under or over-estimated according to interests. For example, insurers may exaggerate them in order to less indemnify policyholders. Insurance is a means of lowering the cost by sharing the burden: a rate is determined according to the cost of an event.

Moreover, hazard impact varies according to the space scale considered (local, regional, national, worldwide). The impact seems to be lower at national scale rather than at the local one since the cost may be shared. The choice of a similar timescale is also essential in order to compare hazard impact between two time periods. The most relevant one could be decades.Nevertheless, as far as Western world is concerned, the evolution of trends reveals that the impact is worse from an economic viewpoint. Recurrent damage and increasing reconstruction costs are observed throughout years. According to the UNESCO, economic costs caused by hazards doubled every 10 years and total cost could reach 150 milliard dollars in the next decade. The following graph underlines the rise of economic losses.

Decadal losses from weather and climate hazards in USA: economic losses6However, there seems to be a fall of fatalities in wealthiest countries. Indeed, if we consider the number of victims in American coastal and tornado prone areas, it is lower than before. It is explained by measures such as building regulation or by an improvement of weather forecasting. The following graph underscores the decline of deaths for both tornadoes and hurricanesDecadal losses from weather and climate hazards in USA: life losses6So, hazard impact is not always getting worse. Firstly, consequences may be positive: it depends on CBA outcome. Are benefits higher than cost? Secondly, the cost may decrease throughout years. Thirdly, the impact is different between goods, lives or environment; so, there are different types of impacts.

Finally, hazard impact does not have the same intensity everywhere.The importance of hazard managementThe severity of an event is mostly determined by human response to hazard. So, impact is not similarly getting worse because of a wide range of hazard approaches. These are more or less efficient. Responses are sometimes unsuited or hazard policies are inexistent due to a lack of resources. Not only hazards create damage but also response has a cost. Keith Smith’s analysis of flooding in Bangladesh reveals that fewer inhabitants of Sonadia island in the Bay of Bengal were killed in 1991 floods compared to 1970: shelters had been constructed.

But, it represented only 3 % of population at risk because the Bangladesh government did not have enough currency to build more. So, financial factor is essential in hazard mitigation. It may explain the divergence of hazard impact between LDCs and DCs.

Despite achievements of science and technology, man seems to be unable to protect lives, goods and environment against hazards. Natural events are sudden but most of the time, it is possible to forecast some of them thanks to advanced techniques in hazard prediction. Yet, human beings remain surprised by hazards when they occur. Subsequently, their decisions may be unsuited as they are taken too quickly. If people had thought on potential risks before, they would have been able to cope better with hazard. So, the impact is getting worse due to emergency unpreparedness.

So, hazard impact is getting worse because of a failure in human response. Stages such as recovery after an event are essential since they enable people to lower the impact of a natural event. For instance, basic supplies such as food, water, clothing, shelter or medical care are provided during relief period to avoid further losses of life. Rehabilitation and reconstruction include “longer-term activities which attempt to return an area to normality after severe devastation”7.However, not only the degree of impact depends on response during and after an extreme event but it also relies on measures adopted before it. A comprehensive hazard management that would include Environmental Impact Assessment, hazard zoning and prediction, land-use capability (RHS survey), enforcement of building codes, public safety measures is always needed but most of the time it is missing. “Environmental impact assessment” is relevant to manage hazard threat because it sets priorities. In general, there is a lack of “hazard-mitigation” whereas it is a relevant approach that aims at reducing losses through long-term strategies.

“Hazard mitigation” is also required because it enables people to be committed and to be more aware of risk. Indeed, it involves individuals, private businesses and industries, state, local and federal government. To sum up, hazard impact is getting worse owing to a lack of pro-active responses.Moreover, awareness of risk is essential to avoid unpreparedness. For instance, “HAZUS Loss estimation models” forecast human and economic impacts. In addition, knowledge of magnitude and probable frequency (records and statistical methods) may help policy-makers to take adequate damage-reducing adjustments.

Individuals have also to be aware of hazard threat. However, despite increased information and warning about the degree of risk, people do not change their settlements: individual interests are predominant and the risk is still denied. Otherwise, either people prefer to modify the physical event through large-scale “environment control” or they accept the loss. So, the impact of an event depends on people’s perception.It leads to the idea that hazard assessment is not only a rational and scientific knowledge of impact.

It is uneasy to determine objectively whether or not consequences are worse. Indeed, hazards are seen through a “perception box” that may lower or increase the impact. For example, people think that the impact is getting worse only because insurers take time to reimburse damage caused by extreme catastrophes. Moreover, “Willingness To Pay” reflects the importance humans give to think. If they are not willing to make efforts, the impact will remain high. The feeling that hazard impact is getting worse varies either from individual to groups or between experts and non-expert. If the event took place a long time ago or is far in space, people do not feel concerned.

To conclude, there is obviously a change in physical mechanisms. The so-called “global climate change” is seen as an explanation of a growing hazard impact. But it is actually getting worse as far as human beings are concerned.

Interactions of human beings with their environment play an important part in global warning. Moreover, men do not take these new mechanisms into consideration. On the one hand, they tend to ignore physical risks (settlements in floodplains). On the other hand, new policies are needed with growing population since previous ones are not adapted yet.

Most of the time, hazard management concentrates on response and recovery instead on focussing on forecasting and pre-planning policies. The degree of hazard impact relies on the efficiency of human management. That’s why hazard impact varies according to “hazard mitigation” strategy adopted. So, the impact is not equally getting worse because people cope with hazards in different ways. In addition, the hazard impact on environment, goods or lifes varies because people do not give them the same importance.

Yet, even if consequences of extreme events vary around the world, natural events remain a worldwide concern. Hazard impact physically crosses frontiers. It is also spread from an economic viewpoint because of globalisation and economic interdependence. But it is a pity that legislation and administration of the management response is not applied everywhere.

Moreover, increasing international relations are not used to develop common policies such as European WFD. Hazard impact on LDCs will remain severe as long as international aid and humanitarian action after a catastrophe remain the only response.