Previous Page Table of Contents Next Page

Strategies for specific hazards

1. Hurricanes
2. Drought and desertification
3. Geologic hazards
4. Floods
5. Landslides

The natural hazards of principal concern to development practitioners in the region are:

- Hurricanes
- Drought and desertification
- Geologic hazards (earthquakes, volcanic eruptions, tsunamis)
- Floods
- Landslides

While hazards often materialize as discrete events, their occurrence may also overlap. For example, hurricanes and tsunamis may produce floods; earthquakes may trigger landslides; and erosion and sedimentation are frequently the result of flooding, desertification, or unsound land management practices rather than hazards in their own right. The natural hazards listed above are of greatest concern to development agencies, not only because they cause the most harm to human life and property, but also because they may be exacerbated by development practices. But the most important thing about these hazards is that means of reducing their impact are now available.

At the start of a development study, the planner should attempt to determine from the information available whether any particular hazard constitutes a problem in the study area. In the absence of sufficient information for these preliminary decisions, the planner usually decides by default: the hazard in question will not be considered.

Because of the availability of new techniques in hazard assessment, it is no longer necessary to make these default determinations. The information required to evaluate a natural hazard can often be obtained as a part of the planning process and it is possible to make hazard assessments part of the study without incurring unreasonable costs or sacrificing other aspects.

The availability of information determines the strategy for treating a natural hazard in a development study. The crucial question is: is the existing information sufficient to determine whether the hazard poses a significant threat in the study area? If not, additional information must be generated, fast enough and cheaply enough to be commensurate with the rest of the study. For hurricanes, desertification, and geologic hazards, the available information is generally adequate; for flooding and landslides, it is not (see box below).

1. Hurricanes

In the Caribbean island countries hurricanes cause more damage and disturb the lives of more people than any other natural hazard. In Mexico and Central America they are second only to earthquakes. From 1960 to 1989 hurricanes killed 28,000 people, disturbed the lives of 6 million people, and destroyed property worth US$16 billion in the Greater Caribbean Basin (excluding the United States and U.S. possessions). Small countries are particularly vulnerable to hurricanes, since they can be affected over their entire area, and major infrastructure and economic activities may be crippled in a single event.

More significant, however, is the record of reducing this impact. Hurricane intensity has not abated. Thus, with population density increasing, the number of deaths would be expected to increase over time. In fact, it has decreased. In 1930 three people were affected by hurricanes for each person killed. By 1989 that ratio had risen to 100,000 to one. The ratio of dollar value of damage to people killed rose from 5,000 to 20,000,000 in the same period. These reduced death rates are due almost entirely to improved warning systems and preparation. Some progress has been made toward reducing damage, but that is a more difficult issue.

A hurricane is defined as a large non-frontal tropical depression or cyclone with wind speeds that exceed 119 km/hr (a tropical storm has wind speeds of 63 to 119 km/hr). The hurricane season of the Greater Caribbean Basin is June through November, although 84 percent occur in August and September. Hurricanes cause damage by their high winds, heavy rainfall, and storm surge. Winds up to about 162 km/hr cause moderate damage such as blowing out windows. Above that velocity winds begin to cause structural damage. Heavy rainfall can cause river flooding, putting at risk all structures and transportation facilities in valleys, and can also trigger landslides.

Storm surge is a rise in sea level due to on-shore winds and low barometric pressure. Storm surges of 7.5 meters above mean sea level have been recorded, and a surge of over 3 meters is not uncommon for a large hurricane. Storm surges present the greatest threat to coastal communities. Ninety percent of hurricane fatalities are due to drowning caused by storm surges. If heavy rain accompanies a storm surge, and the hurricane landfall occurs at a peak high tide, the consequences can be catastrophic. The excess water inland creates fluvial flooding, and the simultaneous increase in sea level blocks the seaward flow of rivers, leaving nowhere for the water to go.

STRATEGIES FOR INCORPORATING NATURAL HAZARD CONSIDERATIONS INTO DEVELOPMENT PLANNING STUDIES BASED ON INFORMATION AVAILABILITY

Hurricanes. If hurricanes are found to be a threat, the study can go directly to local-level vulnerability reduction strategies. Mitigation actions using established structural and non-structural techniques can be undertaken as soon as it is established that the project falls within the hurricane-prone belt because there is currently no practical way to relate mitigation strategies to different hurricane intensities.

- The section on hurricanes discusses how to prepare for storms so as to reduce the damage they cause, with emphasis on procedures for small towns and villages.

Desertification. The information available on desertification for the region is very general, but it can be augmented for a study area easily and at low cost to the level needed for policy orientation and project identification and formulation. Desertification potential must be refined in the context of a development study precisely because the degree of that potential is directly related to the impact of development activities on natural conditions.

- The section on desertification gives the desertification potential for each political subdivision (state, department, province) subject to the hazard and tells how to prepare a desertification analysis using only four universally available parameters.

Geologic Hazards. The information available on seismic, volcanic, and tsunami hazards is sufficient to determine in the preliminary mission whether they constitute a significant threat in the study area. But vulnerability reduction is a site-specific matter, with emphasis on micro-level zonation and determination of severity and probability characteristics. These determinations require relatively elaborate and expensive techniques that are suitable only for feasibility and engineering design studies.

- The appendix lists the political subdivisions subject to seismic and tsunami hazards and gives the location and brief history of all active volcanoes in Latin American and the Caribbean. The lists are sufficient to determine whether these hazards pose a significant threat in a study area.

Flooding and Landslides. The information available on flooding and landslide hazards is generally spotty or nonexistent, but with a combination of historical event studies and new mapping techniques, these hazards can be evaluated at costs and scales appropriate to the corresponding stage of a development study.

- The section on floods describes a technique for quick mapping of flood-prone areas at scales up to 1:50,000 by interpretation of satellite imagery.

- The section on landslides describes alternative methods, depending on the source materials available, for mapping landslide threat.

To assess risk as a step in the process of preparing a hurricane hazard mitigation plan, a planner first determines whether the study area lies within the belt of commonly occurring hurricanes. If it is located in "Hurricane Alley" (see Figure 9), the planner studies records of past storms and land uses and correlates them with probable future land use and population changes. Most cities in the West Indies are in low coastal zones threatened by storm surge, and population movement to these high-risk zones greatly increases vulnerability. The economic sectors most affected by hurricanes are agriculture and tourism. Bananas, one of the most important Caribbean crops, are particularly vulnerable. The tourism sector in the Caribbean is notorious for its apparent disregard of hurricane risk. Not only does a hotel built with insufficient setback risk damage by wave action and storm surge, the building also interferes with the normal processes of beach and dune formation and thus reduces the effectiveness of a natural protection system.

Figure 9 - OCCURRENCE OF TROPICAL STORMS AND CYCLONES IN THE WESTERN HEMISPHERE 1/

1/ Wind strength of Beaufort 8 and above

Source: Munchener Ruck. Mapa Mundial de los Riesgos de la Naturaleza. (Munich, Federal Republic of Germany, Munchener Ruckversicherungs: 1988)

Once the risks are defined and quantified, planners and engineers can design appropriate mitigation mechanisms. Obviously, these are most cost-effective when implemented as part of the original plan or construction. Examples of effective mitigation measures include avoiding areas that can be affected by storm surge or flooding, the application of building standards designed for hurricane-force winds, or the planting of windbreaks to protect crops. Retrofitting buildings to make them more resistant is a more costly but sometimes viable option, but once a project is built in a flood-prone area it may not be feasible to move it to safer ground.

In the past three decades the ability to forecast and monitor these storms has increased greatly, which has had a dramatic effect on saving lives. The time and location of landfall and the resulting damage can be estimated. The U.S. National Hurricane Center uses this information to issue track prediction and intensity forecasts every six hours for tropical storms and hurricanes in the Atlantic/Caribbean region. The U.S. Oceanographic and Atmospheric Administration (NOAA) has developed the model Sea Lake Overland Surge from Hurricanes (SLOSH) to simulate the effects of a hurricane as it approaches land. This makes it possible to determine which areas should be abandoned and to plan evacuation routes. Tailored to specific areas, operational SLOSH is available in the United States and Puerto Rico and is being developed in the Virgin Islands. It can be expanded to other Caribbean and Central American countries.

At the national level non-structural mitigation strategies include campaigns to create a public awareness of warning services and protective measures, since informed citizens are more likely to check the condition of their roofs and other structures at risk. Good examples of such campaigns can be seen in The Bahamas, Barbados, and Jamaica. Taxation of investment in high-risk land is a potentially important strategy that has not been tried widely. Insurance can also be structured to encourage sound land use and structural mitigation actions. Among the important structural strategies are codes that control the design and construction of buildings and, in public works, the construction of breakwaters, diversion canals, and storm surge gates and the planting of tree lines to serve as windbreaks.

All these approaches may be effective in the largest urban settings where communications are good and institutional arrangements are firmly in place. But national emergency preparedness offices usually do not have the resources to function effectively in areas of low population density when faced with widespread catastrophes such as hurricanes. An alternative is to prepare small towns and villages to respond to emergencies by their own means. The approach followed by the OAS in collaboration with the Pan Caribbean Disaster Preparedness and Prevention Project (PCDPPP) in several Eastern Caribbean countries involves training local disaster managers and community leaders of both urban and rural settlements in organizing disaster risk assessment and mitigation in their communities. A training manual and accompanying video produced for this purpose are available. The focus is on lifeline networks (transportation, communications, water, electricity, and sanitation) and critical facilities (health and education facilities, police and fire stations, community facilities, and emergency shelters). Combining the disaster preparedness efforts of the PCDPPP program with disaster prevention through the integrated development planning of the OAS clearly illustrates the interface of disasters and development.

The process of preparing community leaders to cope with hurricanes consists of six steps:

- Preparing an inventory of lifeline networks and critical facilities.

- Learning the operation of these networks and facilities and their potential for disruption by hurricanes.

- Checking the vulnerability of the lifelines and facilities through field inspection and investigation.

- Establishing an effective working relationship with the agencies and companies that manage the infrastructure and services of the community.

- Developing an understanding of the total risk to the community.

- Formulating a mitigation strategy.

Communities can use the OAS-PCDPPP training manual and video to train themselves, but often they find it more effective to have outside help. The best approach is to set up a unit for local training in the national government which then travels regularly to each small community first to train the local leadership and then to give updates and practice sessions.

2. Drought and desertification

Droughts are prolonged dry periods in natural climate cycles. In arid and semi-arid regions, dry periods that are much drier than average and wet periods that are much wetter are common, and these variations cause serious problems. When the wet period is unusually wet, pastoralists increase the size of their herds and farmers extend their cultivation into areas normally too dry for agriculture. Then when a dry period comes, these expansions must be cut back. If the uses that exceed carrying capacity are not cut back, the vegetative cover can die, and the unprotected soil is subject to rapid erosion, one of the indicators of desertification.

Desertification is the spread of desert-like conditions induced by human activities with consequent decrease in biomass production. It is manifested by loss of productive soils, water and wind erosion, creation and movement of dunes, waterlogging, reduced quantity and quality of surface and subsurface water, and rapid depletion of vegetative cover. Figure 10 classifies the status of desertification by country and subdivision (province, department, state) in South America and Mexico.

Desertification is a result of many interrelated phenomena, with human-induced erosion and salinization often exacerbating natural drought. Soil erosion by moving water occurs on any sloping land but can be accelerated by overgrazing, deforestation, certain agricultural practices, road construction, and urban development. Erosion by wind can take place on flat land lacking vegetative cover. Erosion results in loss of soil nutrients, downstream damage by the deposition of sediments generated by the erosion, and depletion of water storage capacity.

Salinization most often occurs on irrigated land as the result of poor water control. Salts accumulate because of flooding of low-lying lands, evaporation from depressions having no outlets, and the rising of groundwater close to the soil surface.

Many of the problems associated with desertification can be circumvented by sound planning. This requires information about physical conditions and the social-cultural context of the planning area. If the area has the potential for desertification (i.e., if it is in one of the areas shown in Figure 10), a desertification hazard assessment should be undertaken at the very outset of a development planning study.

The OAS has developed a simple, quick method for conducting such an assessment that can be applied in the earliest stages of development planning. This method uses only four variables-precipitation, soil texture, slope, and the ratio of precipitation to evapotranspiration. Wind and other descriptors can also be important in some regions, but the four variables selected are those for which data are most readily available. The method defines a maximum of 16 mappable units, as shown in Figure 11. Each unit has a set of characteristics that indicate suitable and unsuitable land uses or management practices and the kinds of problems that misuse can cause. The resulting desertification hazard map can be used to design and evaluate development projects according to the conditions of water scarcity and the potential for desertification. Relying on data that are almost universally available, the approach can be used in the preliminary mission to make a first approximation of the hazard, which can be refined in Phase I.

Other approaches are available. For example, an OAS study of the Paraguayan Chaco delineated four degrees of severity of desertification risk based on characteristics of bioclimate, terrain, and human pressure, and used these units to prescribe appropriate soil management methods and precautions to be incorporated into the proposed irrigation and animal husbandry projects.2/

2/ Gobierno de la República del Paraguay y Departamento de Desarrollo Regional de la Secretaria General de la Organización de los Estados Americanos. Desarrollo Regional Integrado del Chaco Paraguayo (Asunción, Paraguay: Marzo, 1985).

The following are some mitigation measures for overgrazing, dry-land farming, and salinization. Addressing the problem of overgrazing starts with recognizing the needs of the pastoralists. Reducing the number of stock and introducing improvements such as fencing and watering points can help. Ameliorative range management techniques must meet the particular requirements of the area, taking into consideration the most appropriate treatment for flood-plains and sandy hills, the kinds of animals that are most suitable to the area, and the social structure and cultural context. Pastoralists are more willing to accept management alternatives that are relatively less capital-intensive even though they may take longer to be beneficial.

The difficulties of dry-land agriculture include low and unreliable rainfall, hot and dry winds, dependence on extensive rather than intensive farming, a restricted choice of crops, soils that are highly susceptible to wind erosion, and crop yields that are seldom sufficient to justify major investment in chemicals or erosion control measures. Thus the prospects for mitigating problems of dry-land farming are not as favorable as those for range lands. Major dry-land farming problems are erosion by wind and water and loss of fertility owing to the removal of nutrients by crops.

Fertility can be restored with fertilizer-expensive over the short term, but the long-term alternative is severe loss of production. Coarse, sandy soils and soils on steeply sloping land are the most difficult to improve.

Mitigation of soil erosion problems can draw on an array of well-known soil and water conservation practices such as the use of drought-resistant plants, fallow periods and mulches, the installation of water-retaining terraces, wide spacing of plants in and between rows, special practices such as minimum tillage and zero tillage, and leaving crop residues in place after harvesting. With some experimentation it is usually possible to find a set of management practices that the farmer will accept and that will result in greater profits for him within a few years.

Figure 10 - AREAS OF POTENTIAL DESERTIFICATION IN SOUTH AMERICA AND MEXICOa/

COUNTRY

Hyperarid region

STATUS OF DESERTIFICATION

Slight

Moderate

Severe

Very severe

ARGENTINA


Catamarca

Chubut

Catamarca

La Pampa


Chaco

La Pampa

Córdoba



Chubut

Mendoza

Jujuy



Formosa

Neuquén

La Pampa



Jujuy

Río Negro

La Rioja



La Rioja


Mendoza



Mendoza


Salta



Neuquén


San Juan



Río Negro


San Luis



Salta


Santiago del Estero



San Juan





Santa Cruz





Santiago del Estero




BOLIVIA


Cochabamba


Cochabamba



Chuquisaca


Chuquisaca



La Paz


La Paz



Oruro


Potosí



Potosí


Tarija



Santa Cruz





Tarija




BRAZIL



Alagoas





Bahía





Ceará





Paraíba





Pernambuco





Piauí





Río Grande do Norte





Sergipe



COLOMBIA



Atlántico





Guajira


Magdalena

CHILE

Antofagasta

Antofagasta

Aconcagua

Antofagasta


Atacama

Atacama

Coquimbo

Atacama


Tarapacá

Tarapacá

Valparaíso



ECUADOR


Esmeraldas





Guayas





Manabí




MEXICO

Sonora

Baja California Norte

Baja California Norte

Aguascalientes

Chihuahua


Baja California Sur

Nuevo León

Baja California Norte



Sonora

Sinaloa

Chihuahua




Sonora

Coahuila





Durango





Guanajuato





Guerrero





Hidalgo





Michoacán





Nuevo León





Oaxaca





Puebla





Querétaro





San Luis Potosí





Sinaloa





Sonora





Tamaulipas





Zacatecas


PARAGUAY


Boquerón





Chaco





Nueva Asunción




PERU

Ancash

Ancash


Arequipa


Arequipa

Arequipa


Ayacucho


Ica

Ayacucho


Moquegua


La Libertad

Cajamarca


Puno


Lima

Huancavelica


Tacna


Moquegua

Ica




Tacna

La Libertad





Lambayeque





Lima





Moquegua





Piura





Puno





Tacna





Tumbes




VENEZUELA



Falcón





Zulia



a/ Area is defined as the largest political subdivision of the country: province in Argentina, Chile, and Ecuador; department in Bolivia, Colombia, Paraguay, and Peru; and state in Brazil, Mexico, and Venezuela. The fact that an area appears with a specific status of desertification does not necessarily imply that the entire area is affected. Moreover, an area can have more than one status when different portions are affected to different degrees.

Source: Adapted from: Dregne, H.E. Desertification of Arid Lands (New York, New York: Harwood Academic Publishers GmbH, 1983).

Figure 11 - DESERTIFICATION TYPES

Precipitation Class

P/PET classa/

Texture class

Slope class

Desertification Unit

(mm)

(% Sand)

(% Slope)

More than 1500

-

More than 50

More than 10

1

Less than 10

2

Less than 50


More than 10

3

Less than 10

4

Less than 1500

1.0 or greater

More than 50


More than 10

5

Less than 10

6

Less than 50


More than 10

7

Less than 10

8

0.76-0.99

More than 50


More than 10

9

Less than 10

10

Less than 50


More than 10

11

Less than 10

12

0.01-0.75

More than 50


More than 10

13

Less than 10

14

Less than 50


More than 10

15

Less than 10

16

a/ Ratio of precipitation to potential evapotranspiration.

Source: Organization of American States. Primer on Natural Hazard Management in Integrated Development Planning (Washington, D.C.: In Press).

Salinization can be mitigated with currently available technology. Leaching is a practical way to remove excess salts from soils, but requires good drainage. Essentially, what is needed is properly designed and managed irrigation systems. This involves, at a minimum, consideration of the peculiarities of the natural soil situation (e.g., chemical composition of ground water, salinity of soils up to the water table, conditions of natural drainage), deep drainage installations to carry off excess water, and avoidance of over-watering. Over-watering is a common consequence of the propensity to undercharge for water; it can result in such heavy use of it as to result in waterlogging and salinization.

3. Geologic hazards

The most damaging geologic hazards are earthquakes, volcanic eruptions, and tsunamis (large sea waves, erroneously called tidal waves, which are usually caused by earthquakes). Landslides, which can be triggered by earthquakes and other mechanisms, are discussed in Section 5 of this chapter.

Geologic hazards are characterized by (1) very rapid onset; (2) geographically limited impact (the phenomena occur in limited and clearly defined zones in Latin America and the Caribbean); (3) lack of predictability except in the most general sense; and (4) extreme destructiveness (in spite of their relative rarity, earthquakes in urban areas, pyroclastic flows and mudflows caused by volcanic eruptions, and flooding due to tsunamis are some of the most damaging and feared natural hazards).

This combination of characteristics makes the non-structural strategy of avoidance the best way to cope with geologic hazards. As has been emphasized, the avoidance strategy requires information about the threat of the hazards as early as possible in the development planning process. The information requirements are very general early in the process, becoming more explicit with each successive stage so as to provide answers to the following questions in order:

- Does the hazard pose a threat in the study area?

- Is the danger great enough to merit mitigation?

- What kind of mitigation mechanisms are appropriate?

- What are the costs and benefits of a particular mitigation measure, in terms of both economics and quality of life?

Scientific data to answer the first question exist for the principal geologic hazards in most of Latin America and the Caribbean, but up to now they have not been readily accessible. One of the services of the OAS has been to compile this information in a form suitable for use by planners. This section summarizes that information for earthquakes, volcanic eruptions, and tsunamis.

Earthquakes

Two kinds of information are needed to evaluate earthquake threat: the potential severity of an earthquake and the likelihood that a damaging earthquake will occur during a specific time frame. When either type of information is not available, a partial evaluation can be made with the information that does exist.

Potential severity is usually defined historically; that is, the largest earthquake determined to have occurred in an area is taken as the maximum that is likely to occur there again. A map of Earthquake Intensities of South America has been prepared that delineates zones according to the Modified Mercalli Intensity (MMI) scale, a 12-unit scale of increasing shaking intensity. MMI VI, for example, is defined as follows: "Felt by all; many frightened and run outdoors; falling plaster and chimneys; damage small."3/ At MMI X, roughly equivalent to magnitude 7 on the Richter scale, "Most masonry and frame structures destroyed; ground cracked; rails bent; landslides." Taking MMI VI as a cut-off, on the assumption that mitigation measures would be difficult to justify economically at or below this level, mitigation measures should be considered for areas of MMI VII and above.

3/ Centro Regional de Sismología para América del Sur (CERESIS). Mapa de Intensidades Máximas de América del Sur (Santiago, Chile: CERESIS, 1985).

The threat is serious. Figure 12 shows twelve locations in Central and South America where there is a probability of 50 percent or greater that an earthquake with magnitude of 7+ will occur in the next 20 years. Damaging quakes in Costa Rica and Ecuador are almost a certainty during that time (probabilities of over 90 percent). Table A-1 (see Appendix A for Tables A-1 to A-6) gives the MMI rating of all departments (provinces, states) in South America having an MMI of VI or greater. Table A-2 gives the conditional probability of a large or great earthquake occurring on the west coast of South America in the next five, ten, or 20 years, again at the departmental level, as well as the maximum likely seismic intensity of such a quake. Tables A-3 and A-4 give comparable information for Central America. The Primer on Natural Hazard Management in Integrated Development Planning gives similar data for Mexico and the Caribbean.

This is, of course, very preliminary information, but if, for example, a study area has an 80 percent probability of being struck by an MMI X earthquake in the next 20 years, the planner recognizes that reality as something that cannot be ignored.

OAS work with geologic hazards has been confined largely to pre-event planning and non-structural mitigation measures. The study for the integrated development of the San Miguel-Putumayo river basin on the Colombia-Ecuador border, for example, included a comprehensive examination of all natural hazards that could affect the projects identified. Active fault zones-the locus of potential earthquakes and unstable ground unsuitable for locating infrastructure-were one of the elements studied.

The principal earthquake hazards are ground shaking, fault rupture, and propensity to liquefaction (see the section below on landslides). Once it is recognized that an area is prone to earthquakes, it is important to prepare maps of high-risk areas delineating zones subject to the particular hazards. Some hazard and risk mapping has been completed in Latin America and Caribbean countries, but in general it is not very reliable or useful to engineers, government officials, or planners for site-specific engineering design work. Some national and regional projects have begun to incorporate recent scientific and technological advances into seismic hazard and risk mapping, and are producing work of much higher quality. The availability of existing information, and more particularly the quality of that information, must be determined for areas under seismic threat and supplemented as necessary.

The science of earthquake engineering has devised building techniques and materials that resist all but the strongest earth shaking. Building codes stipulate the application of these structures. Retrofitting may provide important economic benefits for large buildings and public infrastructure. It is also of great importance in saving lives of millions who live in non-engineered mud constructions. Basic do-it-yourself techniques can prolong resistance to shaking of these structures long enough to allow people to escape before collapse. With regard to fault rupture, the best way to cope with this hazard is to avoid the narrow zones prone to movement along faults.

Volcanic Eruptions

The principal volcanic hazards are pyroclastic flows, mudflows (or lahars), ash falls, projectiles, and lava flows. These hazards usually do not constitute a serious problem more than 30 km from the source, although in exceptional cases, a lahar or an ash fall can cause serious damage as much as 60 km away. Table A-5 characterizes all "active" volcanoes in Latin America and the Caribbean. Because some of the most terrible eruptions have come from volcanoes that had been regarded as dormant, an active volcano is defined as one that has erupted in the past 10,000 years (the Holocene Epoch of geologic time). The degree of threat is gauged by periodicity, with short-term periodicity (interval between eruptions of less than 100 years) posing a greater threat than long-term periodicity. The information given for each volcano includes location, periodicity, date of last eruption, a measure of the size or "bigness" of the largest historic eruption, and the hazards associated with its eruptions.

Figure 12 - TOP SEISMIC HAZARD SITES: AREAS IN LATIN AMERICA WITH GREATER THAN 50 PERCENT PROBABILITY OF AN EARTHQUAKE OF MAGNITUDE 7+ DURING 1989-2009

Location

Magnitude
(Richter)

Probability
(Percent)

Ometepec, Mexico

7.3

74

Central Oaxaca, Mexico

7.8

(72)a/

Eastern Oaxaca, Mexico

7.8

70

Western Oaxaca, Mexico

7.4

64

Colima, Mexico

7.5

66

Central Guerrero, Mexico

7.8

(52)a/

Southeastern Guatemala

7.5

79

Central Guatemala

7.9

50

Nicoya, Costa Rica

7.4

93

Papagayo, Costa Rica

7.5

55

Jama, Ecuador

7.7

90

Southern Valparaíso, Chile

7.5

61

a/Probability values in parentheses reflect less reliable estimates.

Source: Nishenko, S. P. Circum-Pacific Seismic Potential 1989-1999. National Earthquake Information Center, U.S. Geological Survey, Open File Report 89-86 (Reston, Virginia: U.S. Geological Survey, 1989).

If a study area lies within 30 km of a volcano with short-term periodicity, a volcanic hazard zonation map showing the likelihood of occurrence and severity of each hazard in the vicinity of the volcano should be prepared as part of the planning process. Few mitigation measures other than avoidance are effective in resisting volcanic hazards such as lava flows or pyroclastic flows. Steeply sloping roofs help to reduce damage from heavy ash falls.

Tsunamis

These awesome seismic sea waves are caused by large-scale sudden movement of the sea floor, usually due to earthquakes. In Latin America they are a significant threat only on the west coast of South America, where every off-shore earthquake over a magnitude 7.5 is potentially tsunamigenic. While they occur in the Caribbean, they are so infrequent and cause so little damage, mitigation is difficult to justify economically. Even where tsunamis pose a significant threat, mitigation is feasible only for urban concentrations. The construction of sea walls along low-lying stretches of coast, planting tree belts between the shoreline and built-up areas, and zoning restrictions provide some measure of help, but effective warning and evacuation systems are the most reliable defense against this intractable hazard.

Table A-6 gives estimates of the potential tsunami threat for the west coast of South America, showing the potential wave height for population centers from Colombia to Chile that face tsunami hazards. Comparable information is available for Mexico and Central America.

4. Floods

Floods are usually described in terms of their statistical frequency. For example, the flat land bordering a stream that is inundated at a time of high water is called a "100-year floodplain" if it is subject to a 1 percent probability of being flooded in a given year. Commonly, any risk this great or greater is considered significant.

Development practices may unwittingly increase the risk of flooding by increasing the amount of water that must be carried off, or decreasing the area available to absorb it. Drainage and irrigation ditches, as well as water diversions, can alter the discharge into floodplains and a channel's capacity to carry the discharge. Deforestation or logging practices will reduce a forest's water absorption capacity, thus increasing runoff. Large dams will affect the river channel both upstream and downstream: the reservoir acts as a sediment trap, and the sediment-free stream below the dam scours the channel. Urbanization of a floodplain or adjacent areas increases runoff because it reduces the amount of surface area available to absorb rainfall. In short, integrated development planning must examine the potential effect on flooding of any proposed change and must identify mitigation measures that would avoid or minimize flooding for inclusion in investment projects.

First, however, the planning study must establish river flow patterns and propensity to flood. This has commonly been accomplished by gauging rivers and streams, thus directly measuring flood levels and recurrence intervals over a period of many years to determine the statistical probability of given flood events. Without a record of at least twenty years such assessments are difficult, but in many countries stream-gauging records are insufficient or absent. In this situation hazard assessments based on remote sensing data, damage reports, and field observations can be used to map flood-prone areas that are likely to be inundated by a flood of a specified interval.

Remote Sensing Techniques for Floodplain Mapping

Integrated regional development planning studies do not traditionally include original flood hazard assessments but rather depend on available information. If such information is needed but not available, an assessment should be undertaken as part of the study. If time and budget constraints preclude a detailed, large-scale assessment, a floodplain map and a flood hazard assessment can be prepared through the photo-optical method, using Landsat data and whatever information can be found.

Floods and engineering structures on a floodplain cause changes in the river channel, sedimentation patterns, and flood boundaries. A flood often leaves its imprint, or "signature," on the surface in the form of soil moisture anomalies, ponded areas, soil scours, stressed vegetation, and debris lines for days or even weeks after the flood waters have receded. Because satellite imagery can provide a record of these changes and imprints, up-to-date imagery can be compared with previously collected data to determine alterations during specific time periods. Similarly, the inundated area can be compared with a map of the area under pre-flood conditions.

It should be noted that the delineation of flood-plains through remote-sensing data cannot, by itself, be directly related to probabilities of recurrence. However, when these data are used in conjunction with other information such as precipitation records and history of flooding, the delineated floodplain can be related to an event's likelihood of occurrence. This method can reveal the degree to which an area is flood prone and yield information useful for a flood hazard assessment.

As an example, the Government of Paraguay requested the OAS to delineate the floodplain and flood hazards along the Pilcomayo River because of its recurrent flooding. The study team found Landsat data showing the river in normal and flooded conditions, which, after processing and interpretation, made it possible to plot the floodplain boundaries and hazardous zones rapidly and confidently. From imagery taken at three points in time, the analysts were able to identify areas of sediment deposit (Figure 13) and changes in the course of the river (Figure 14). Since this was a preliminary analysis, the map was produced without field checking, thus lowering the cost. The 1:500,000-scale map of about 60,000 sq km was produced in one month at a cost of US$3,800.

Dynamic features of flooding that can cause changes, e.g., changes in the channel of the river itself or floodplain boundaries, can be monitored through repetitive coverage of any area by earth observation satellites. Further, the spatial distribution of the features that have changed can be readily mapped by techniques of temporal analysis developed since the launch of Landsat 1 in 1972.4/ Slides of full scenes and subscenes can be projected at any scale. The slides can be projected onto a base map, thematic maps, and enlarged satellite single-band prints to define hazard-prone areas for further analysis.

4/ Deutsch, M. "Optical Processing of ERTS Data for Determining Extent of the 1973 Mississippi River Flood" in. R.C. Williams and W.D. Carter (eds.). ERTS 1 - A New Window on Our Planet, U.S. Geological Survey Professional Paper 929 (Reston, Virginia: U.S. Geological Survey, 1976): pp. 209-213.

Deutsch, M. and Ruggles, F.H. "Optical Data Processing and Projected Applications of the ERTS-1 Imagery Covering the 1973 Mississippi River Valley Floods" in Water Resources Bulletin, Vol. 10, No. 5 (1974): pp. 1023-1039.

Deutsch, M., Kruus, J., Hansen, P., and Ferguson, H. "Flood Applications of Satellite Imagery" in M. Deutsch, D. Wiesnet, and R. Rango (eds.). Satellite Hydrology, American Water Resources Association Proceedings from the Fifth Annual W.T. Pecora Memorial Symposium on Remote-Sensing (Sioux Falls, South Dakota: 10-15 June 1979): pp. 292-301.

Figure 13 - USE OF SATELLITE IMAGERY TO DETECT SEDIMENT DEPOSITION

Legend:

A. LANDSAT-1 MSS band-5 negative.
Subscene (path 245/row 76) covering a portion of the Pilcomayo River basin.

Collected September 1, 1972.

B. LANDSAT-2 MSS band-5 negative.

Subscene (path 245/row 76) covering the same portion of the Pilcomayo River basin as in "A" above.

Collected March 29, 1976.

C. Temporal composite of subscenes A and B.
The arrows show the areas of sediment deposition in the interval between 1972 and 1976.

Source: OAS. Primer on Natural Hazard Management in Integrated Development Planning. (Washington, D.C.: In Press)

Figure 14 - USE OF SATELLITE IMAGERY TO DETECT RIVER COURSE CHANGE

Legend:

A. LANDSAT-2 MSS band-7 subscence showing a reach of the Pilcomayo River.

Collected March 30, 1976.

B. LANDSAT-4 MSS band-7 subscence showing the same reach of the Pilcomayo River as in "A" above.

Collected October 12, 1982.

C. Temporal composite of subscenes A and B.
The arrows show the change in the course of the Pilcomayo River between 1976 and 1982.

Source: OAS. Primer on Natural Hazard Management in Integrated Development Planning. (Washington, D.C.: In Press)

While data prices vary from source to source and country to country, the cost of data acquisition, analysis, and preparation of analog products usually ranges from about four to 20 U.S. cents per square kilometer. A remote-sensing specialist familiar with photo-optical or computer-enhanced multispectral analysis systems, in collaboration with other planning studies and with regional complementary information and logistical support, would be able to carry out a flood hazard assessment and prepare a flood plain map covering 30,000 to 90,000 square kilometers, at a scale of up to 1:50,000, in approximately one month.

5. Landslides

The term "landslide" conjures up an image of a great mass of rock and dirt roaring down a mountainside, uprooting huge trees, burying whole villages, pushing before it a howling wind that flattens every structure. This is a fair description of an avalanche, one type of the earth mass movements grouped popularly as landslides, but there are several others. Many are less dramatic but nevertheless cause much damage.

For the purpose of hazard management, three general types of earth mass movement merit consideration: (1) slides and avalanches, (2) flows and lateral spreads (liquefaction phenomena), and (3) rockfalls. Slides and avalanches are very rapid movement of colluvial material on over-steepened slopes under conditions of high moisture. They occur commonly to frequently, and each event can cause moderate to great damage, so that collectively they cause very great damage. Liquefaction refers to rapid, fluid movement of unconsolidated material on gently sloping to nearly flat land. These earth movements occur commonly and can cause great to very great damage. Rockfalls are free-falling or tumbling rocks from cliffs and steep slopes. Each event may cause limited damage, but because they are so frequent, cumulatively they cause great damage and loss of life.

Landslides are often triggered by earthquakes but can also be set off by volcanic eruptions, heavy rains, groundwater rise, undercutting by streams, and other mechanisms; consequently, they occur more widely than earthquakes.

The best strategies for mitigating landslide hazards are to avoid construction in hazardous areas and to avoid land uses that provoke mass movement. To build these strategies into development planning requires information on the likely occurrence of landslides. Such information should be compiled only for areas of intensive present or planned land use, since mitigation is not needed in areas of non-intensive use such as extensive grazing or park land.

A map showing landslide potential is suitable for making recommendations on land use intensity, but the more explicit information afforded by a landslide zonation map is required for land use management. Methods of preparing both these types of landslide hazard assessments are discussed briefly below.

The best indicator of landslide potential is the evidence of past landslides. The location, size, and structure of past landslides can be interpreted from remotely-sensed imagery (aerial photography and satellite imagery). A map showing the aerial distribution of landslides can be compiled, and zones of differing landslide potential can be interpreted. Since the map is based simply on frequency of occurrence and not on causal factors, it has limited predictive power.

Slides and avalanches are associated with steep slopes, certain types and structures of bedrock, and particular hydrological conditions. Maps of these characteristics can be prepared, and a landslide zonation map can be compiled by overlaying these causal factors. Much of the required data such as bedrock geology and topography may already be available. The rest can be compiled-again using remote-sensing imagery. The geology, slope, and hydrology data can be overlaid to compile a map on which each unit is a combination of the three natural characteristics. Development activities (e.g., the conversion of forest to grasslands or crops, which increases soil moisture) can increase susceptibility to landslides, and the map units of natural characteristics can be adjusted to show the effects of these human activities. Each of the resulting units can then be characterized as to landslide potential, to provide the basis for preparing a landslide hazard zoning map.

The same process can be followed in evaluating the potential liquefaction, except that for this type of mass earth movement the critical factors are the presence of unconsolidated Holocene sediments (sands and silts less than 10,000 years old) and less than 30-foot depth to the water table.

An example is the landslide hazard assessment prepared by the OAS at the request of Dominica.5/ The study found that the volcanic origin of the country, resulting in steep slopes and unstable bedrock, and the abundant rainfall together create conditions which readily generate landslides. A full 2 percent of the land area of the country is disturbed by existing landslides, of which the most abundant type is debris flows. The landslide analysis team first delineated all past landslides on black-and-white 1:20,000 aerial photographs and prepared a landslide map at a scale of 1:50,000. Next a map of surface geology was compiled from existing information and overlaid on the landslide map to determine which bedrock units were associated with existing landslides. Six of the eight bedrock units were found to be so associated. Next a map of slope classes was compiled, again from existing information. Four classes were defined that corresponded to present land uses. Hydrologic factors were examined, but no correlation between rainfall distribution or vegetation zones with landslides could be established. Finally, the bedrock and slope units were combined, the composite units were compared with the landslide map, and the proportion of each bedrock-slope unit subject to landslide disturbance was determined.

5/ Organization of American States. Landslide Hazard on Dominica, West Indies (Washington, D.C.: OAS, February 1987).

The landslide hazard map was used to locate areas unsuitable for development. Surprisingly, it also showed that an active landslide area could dam a tributary of the Trois Pitons river, threatening the lives of the downstream population. The map of the 290 sq mi country was compiled in six weeks at a total cost of US$13,000.

The important message here is that by using modern remote-sensing techniques a landslide hazard zoning map-which greatly enhances the ability of planners to make intelligent choices about future land use-can be compiled in one or two months at only the costs of technician time and the acquisition of the imagery.

As wise as the strategy of avoiding hazardous areas may be, it is not always possible to follow it. The poor commonly establish squatter settlements on the slide-prone steeply sloping areas surrounding many Latin American urban centers. Landslide mitigation mechanisms tend to be very expensive under these circumstances. At a minimum, squatters should be helped to avoid settling on previous slides, and care should be taken to avoid cutting off the toe of a steep slope to increase the area of a settlement. These areas are most susceptible to sliding in heavy rains, at which time preparatory measures should be taken to deal with large slides that may occur.

Liquefaction can be prevented by ground stabilization techniques or accommodated through appropriate engineering design, but both are expensive.

As with all mitigation measures, these proposals hold only within the constraints of cost-benefit analysis. Avoidance mechanisms will almost invariably yield high cost-benefit ratios. The results for other mechanisms are not as predictable.

Previous Page Top of Page Next Page