Previous Page Table of Contents Next Page

CHAPTER 10 - LANDSLIDE HAZARD ASSESSMENT

A. OVERVIEW OF LANDSLIDE HAZARD MAPPING AND THE DEVELOPMENT PLANNING PROCESS
B. LANDSLIDES, LANDSLIDE HAZARD ASSESSMENT, AND AREAS OF CONCERN
C. MAPPING PHYSICAL FACTORS AND PREPARATION OF A LANDSLIDE HAZARD MAP
CONCLUSION
REFERENCES
APPENDIX

SUMMARY

This chapter presents (1) key terms, concepts, and important considerations related to landslide susceptibility; (2) a technique-hazard zonation mapping-for examining landslide risks; and (3) the critical issues that need to be addressed in integrating landslide hazards into the development planning process.

In 1974, one of the largest landslides in recorded history occurred in the Mantaro River valley in the Andes Mountains of Peru (Hutchinson and Kogan, 1975). A temporary lake was formed when the slide dammed the Mantaro River, resulting in the flooding of farms, three bridges, and twelve miles of roadway. Almost five hundred people in and around the village of Mayunmarca lost their lives. This disaster is one example of the destructive potential of landslides and why they are considered hazards. While not every landslide results in catastrophe, the damage from many small ones may equal or exceed the impact of a single major failure. Thus, both large and small landslides are capable of causing significant damage and loss of life.

Assessing relative landslide hazard is the objective of the method described in this chapter. Its primary product, a landslide hazard map, provides planners with a practical and cost-effective way to zone areas susceptible to landsliding.

This method can be used both by planners and by landslide technical specialists. The planner will gain a working knowledge of concepts and considerations for incorporating landslide hazard evaluation into the planning process, using a level of evaluation suitable for each stage of the process, and thus should be able to ask the appropriate questions of the technical specialist and prepare terms of reference to ensure that the needed information is obtained. The technical specialist will find a review of landslide hazard issues and guidelines for conducting a landslide zonation. As is often the case in natural hazards management, planning studies are often the link between scientific information and the general development planning process.

The method presented, one of several that are available, has the following characteristics:

- Various thematic maps and remote sensing information usually available to a development study are used.

- It is designed to provide landslide hazard information appropriate for each of the stages of the planning process.

- Relative susceptibility to landsliding is used as a measure of the potential hazard within an area.

- It is applicable to regions with different geomorphologic and vegetation characteristics.

- It can usually be used within the time and budget constraints of a planning study.

A. OVERVIEW OF LANDSLIDE HAZARD MAPPING AND THE DEVELOPMENT PLANNING PROCESS

1. Determining Acceptable Risk
2. Landslide Hazard Mapping
3. Integrating Landslide Hazard Zonation Maps into the Development Planning Process

The susceptibility of a given area to landslides can be determined and depicted using hazard zonation. A landslide hazard map can be prepared early in the planning study and developed in more detail as the study progresses. It can be used as a tool to help identify land areas best suited for development by examining the potential risk of landsliding. Furthermore, once landslide susceptibility is identified, investment projects can be developed which avoid, prevent, or substantially mitigate the hazard.

Determining the extent of landslide hazard requires identifying those areas which could be affected by a damaging landslide and assessing the probability of the landslide occurring within some time period. In general, however, specifying a time frame for the occurrence of a landslide is difficult to determine even under ideal conditions. As a result, landslide hazard is often represented by landslide susceptibility (Brabb, 1985). Similar to the concept of flood-prone areas (see Chapter 8), landslide susceptibility only identifies areas potentially affected and does not imply a time frame when a landslide might occur. To simplify these concepts, landslide susceptibility will be referred to as landslide hazard in this chapter. Comparing the location of an area of proposed development to the degree of landslide hazard present enables the planner to estimate the landslide risk. This can be used to define land use capability and identify appropriate mitigation measures.

DEFINITIONS

- Landslide Hazard: as represented by susceptibility, which is the likelihood of a potentially damaging landslide occurring within a given area.

- Vulnerability: the level of population, property, economic activity, including public services, etc., at risk in a given area resulting from the occurrence of a landslide of a given type.

- Risk (specific): the expected degree of loss due to a particular landslide phenomenon.

A landslide hazard map which identifies areas of differing landslide potential may be generated. The need for such landslide hazard information may vary according to the future land use. The degree of landslide hazard present is considered relative since it represents the expectation of future landslide occurrence based on the conditions of that particular area. Another area may appear similar but, in fact, may have a differing landslide hazard due to a slightly different combination of landslide conditions. Thus, landslide susceptibility is relative to the conditions of each specific area, and it cannot be assumed to be identical for a similar appearing area.

Even with detailed investigation and monitoring, it is extremely difficult to predict landslide hazards in absolute terms. Sufficient understanding of landslide processes does exist, however, to be able to make an estimation of landslide hazard potential. The planner can use this estimation to make certain decisions regarding site suitability, type of development, and appropriate mitigation measures. Thus, the planner is determining acceptable risk.

1. Determining Acceptable Risk

Determining whether there is a need for landslide hazard information is the first step in ensuring that landslide risk does not exceed an acceptable level in planning future land use. The objective of landslide information is to identify which relatively landslide-susceptible areas are best suited for what types of development activities. For example, assessing landslide hazard would have a low priority in planning areas to be set aside for national parks or game preserves. Conversely, landslides can be an important factor in the development of newly cleared forest areas or in building infrastructure in mountainous or hilly terrain. Clearly, the amount of landslide hazard information needed is based on the level and type of anticipated development for an area. Failure to understand the potential effects landsliding can have on a project or how the project might affect landslide potential can bring increased risk.

Natural changes as well as human-induced changes can affect the susceptibility to landslides and must be understood when assessing the landslide potential of an area. It is critical for a planner to appreciate these issues early in the planning process. A decision is ultimately made regarding the degree of risk that is acceptable or unacceptable to a project. Mitigation strategies are then designed to reduce risk. These concepts are discussed at more length later in this chapter.

Early consultation with landslide technical specialists is recommended so that they can assess the risk of proposed activities in a landslide hazard area. The planner, while not expected to be a technical expert, must know the questions to ask of a landslide specialist. By asking the appropriate questions, the planner will be able to identify and evaluate measures to minimize or avoid landslide vulnerability.

2. Landslide Hazard Mapping

Interpretation of future landslide occurrence requires an understanding of conditions and processes controlling landslides in the study area. Three physical factors-past history, slope steepness, and bedrock-are the minimum components necessary to assess landslide hazards. It is also desirable to add a hydrologic factor to reflect the important role which ground water often plays in the occurrence of landslides. An indication of this factor is usually obtained indirectly by looking at vegetation, slope orientation, or precipitation zones. All of these factors are capable of being mapped. Specific combinations of these factors are associated with differing degrees of landslide hazards. The identification of the extension of these combinations over the area being assessed results in a landslide hazard map. The technique used to prepare hazard maps is called a combined factor analysis and is described in detail in Section C of this chapter.

3. Integrating Landslide Hazard Zonation Maps into the Development Planning Process

a. Preliminary Mission
b. Phase I - Development Diagnosis
c. Phase II - Development Strategy and Project Formulation
d. Project Implementation

Landslide hazard information serves as one of the many components in an integrated development planning study. Since landslide activity can adversely affect or interfere with human activity, landslide hazards constrain or limit land-use capability. For this reason, it is important to identify relative landslide hazard levels early in the planning process. This permits planners to determine the degree of landslide risk that is acceptable or unacceptable to a development program. Decisions can then be made regarding which of these measures will be taken: avoidance, prevention, or mitigation of existing and future landslide hazards in the development program. The method described in this chapter places emphasis on landslide hazard identification and its use in an integrated planning study as natural resources are evaluated, a development strategy is formulated, and investment projects are identified at the profile level.

a. Preliminary Mission

During the Preliminary Mission of an integrated development planning study, an initial review is made of the type and content of available information, including natural hazard information (see Appendix A). The availability of geologic, topographic, hydrologic, and vegetation maps and aerial photographs is usually ascertained. This information is essential for executing a landslide hazard zonation (see Figure 10-1). Also during this stage of the study, available information should be collected and reviewed concerning assessments of natural hazards, including landslides and disasters, which are known to have affected the study area. See Chapter 1 for a more detailed discussion of the integrated development planning process.

b. Phase I - Development Diagnosis

In the context of planning the development of a river basin, province, or other planning unit, a development diagnosis assists in identifying areas with highest development potential. These are designated "target areas," in which subsequent, more detailed studies are concentrated. Part of the development diagnosis process involves identifying and delineating natural resource factors that favor or constrain development in a particular area. Landslide hazard is an undesirable factor, and the greater the hazard, the more it may shape the development potential.

When a potential hazard is present in the study area, the first step is to undertake a brief survey to establish whether landslides have occurred in recent times. Roads, railroads, and river banks are good sites for seeking signs of past landslide occurrence. Discussions with local authorities responsible for public works, forestry, and agricultural activities can prove to be a valuable source of information since they may be familiar with past landslides in an area. However, it is important to bear in mind that new development activities may increase landslide hazards, and the absence of evidence from past landslides does not guarantee that landslides will not pose any problems in the future.

PRELIMINARY MISSION (STUDY DESIGN) QUESTIONS PLANNERS NEED TO ASK:

- Are geologic, topographic, hydrologic, and vegetation maps available? At what scale?

- Are aerial photographs available? At what scale?

- Does the study area have a history of landslides and/or disasters caused by landslide events?

- Is landslide hazard assessment information available?

KEY DECISIONS TO BE MADE AT THIS STAGE:

- Is it likely that landslides will affect major and/or significant portions of the study area?

- Will the study have access to landslide hazard assessment information other than what it may produce?

PHASE I: DEVELOPMENT DIAGNOSIS (STRATEGY AND PROJECT IDENTIFICATION)

QUESTIONS PLANNERS NEED TO ASK:

- Is sufficient information to prepare a landslide inventory map, an isopleth map of existing landslides, and/or a landslide hazard map using factor analysis available?

- How will the assessment be carried out? During what time period? How will the assessment information be integrated into the overall study development strategy and project identification activities?

KEY DECISIONS TO BE MADE AT THIS STAGE:

- Is a landslide hazard map necessary?

- At what scale should the map be prepared?

- Who will execute the assessment?

- Who will be responsible for incorporating the assessment information into the overall study activities?

Figure 10-1 DEVELOPMENT PLANNING STUDY AND LANDSLIDE HAZARD ASSESSMENT ACTIVITIES

The areal extent and variety of development activities being considered make determining the landslide susceptibility based on all existing landslides, regardless of type, an appropriate approach (DeGraff, 1982). A simple inventory of past landslides, along with data regarding bedrock, slope steepness, and - when available-the hydrologic factor, produces a landslide hazard map that will satisfy the needs of the development diagnosis (see Figure 10-1). Suitable scales for the landslide hazard map range from 1:250,000 to 1:50,000. (See Figure 10-2 for a description of hazard identification needs and appropriate map scales for the different planning stages).

Having limited or insufficient data for preparing the combined factor analysis is most likely to be a problem encountered at the development diagnosis level. When this situation arises, there are two options: (1) invest the money and human resources to obtain the data needed to produce a landslide hazard map, or (2) prepare an isopleth map of existing landslides (described in Section C of this chapter). The isopleth map shows areas of frequent or infrequent landslide occurrence. While this type of map provides some idea of where landslides can be a major influence on development, it is only a rough approximation of where a problem can be encountered during development. Isopleth maps are an acceptable option at this stage of development but are wholly unsuitable for use in the more detailed planning stages.

The degree of landslide hazard in an area is a limiting factor only for those activities that may alter the existing balance between forces driving and resisting movement on an unfailed slope. Planners need to understand what effects development activities may have on this balance of forces. For example, placing a fence around a field is not going to produce a landslide, nor will it prevent one. Removing forest cover to create a field for cultivating crops is much more likely to lead to landslide occurrence, since it alters the balance of forces and may increase the susceptibility to slope failure by some triggering event, such as prolonged rainfall, which would not have produced the landslide under the original conditions. This increased susceptibility may not be immediately apparent since there may be a lag time before this is evident.

Landslide hazard zonation can be represented as an individual factor limiting land capability or it can be combined with hazard zonation for other natural hazards as an aggregate hazard. There are at least 10 different approaches that have been used to generate land capability maps (Hopkins, 1977).

Figure 10-2 - LANDSLIDE HAZARD CONSIDERATIONS AT DIFFERENT PLANNING STAGES

Planning Stage

Hazard Identification Need

Landslide Inventory Level

Suitable Scales For Hazard Maps

Preliminary Mission

Identify hazard issue

As available

As available

Phase I- Development Diagnosis

Degree of hazard from all types of landslides

Simple

1:250,000 to 1:62,500

Phase II - Action Plan and Project Formulation

Degree of hazard from all types of landslides supplemented by hazard from some specific types

Intermediate

1:62,500 to 1:10,000

Project Implementation

Site-specific hazard based on geotechnical models

Detailed

1:12,500 to 1:500

Chapter 3 discusses land capability in more detail. The method for landslide hazard assessment presented in this chapter results in the production of a map. Thus, it can be considered in the application of land-use capability approaches.

There are two main applications of a landslide hazard assessment in land-use capability that includes relative surveys. First, it is used in overall development planning to emphasize the subjective nature of assigning land-use capability. For example, at the development diagnosis stage, the relative classification of "highest" capability can be assessed in relation to the constraints that possible increased landslide hazards may pose to proposed development activities. Second, it can show where existing development may face some risk previously unidentified. This enables prioritizing mitigation activities to be assigned to different development activities.

c. Phase II - Development Strategy and Project Formulation

An action plan is defined with the objective of facilitating development of target areas identified in Phase I. Development projects considered for the target area are formulated at this stage. Also at this time, landslide hazard evaluation within the study area is refined. The general landslide hazard assessment must be supplemented with an intermediate inventory to show the degree of hazard for specific types of landslides that may impact on proposed development activities. For example, introducing widespread agricultural activities into a forested environment requires a greater understanding of the hazard from shallow landsliding rather than from deep-seated rockslides.

In developing areas with landslide hazards, mitigation measures should be selected if they are not already part of the project identification information. It is possible to reduce the possible impact of natural landslide activity and limit landslides which occur as a result of human activity (Kockelman, 1985). There are two basic approaches: first, to avoid landslide-susceptible areas, and second, to design measures to compensate for the inducement of landslides (see the box below). For example, make location decisions so as to avoid building in certain areas, such as placing dwellings and critical infrastructure outside areas with a high likelihood of natural landslide activity. In some instances, the potential effects of a landslide can be mitigated. Landslide hazards resulting from development can be reduced by designing changes to counteract the impact that development may have on slope integrity. This might take the form of permitting only warehouses and storage facilities in higher hazard areas, to reduce the vulnerability to the population should a landslide occur.

LANDSLIDE TRIGGERING ACTIONS, MITIGATION APPROACHES, AND DEVELOPMENT VARIABLES

- Triggering Actions:

From other hazards:

Earthquakes
Floods
Fire (and resulting loss of vegetation)
Volcanoes

Development related:

Changes in vegetation cover
Earthen dams
Excavation and mining
Irrigation
Infrastructure and structure construction
Liquid disposal (sanitary, sewers, latrines, etc.)
Pilings
Soils deposits

- Mitigation Approaches:

Avoidance
Insurance and taxation
Land-use zoning
Stabilization
Structural design

- Development Variables:

Available information
Economic, social, and political concerns
Existing development
Proposed development

In formulating investment projects, a more detailed hazard zonation map is needed. An intermediate landslide inventory is needed which provides greater detail to distinguish different landslide types. This data can be used for a reanalysis of the combined factor analysis. This reanalysis yields an improved landslide hazard map. If the hydrologic factor was not part of the earlier landslide hazard analysis, its inclusion at this stage would greatly improve the resulting hazard map.

At this stage, the value of a landslide hazard map to planners can be enhanced by representing areas where specific landslide types are prevalent. This is accomplished by preparing an isopleth map, as mentioned in Phase 1. Preparation, however, should be altered to meet the specific needs of this planning stage. Alteration of the isopleth map is described in detail in Section C's discussion of "Compensating for Insufficient Data-The Isopleth Map". This produces a map representing the intensity of past landslide occurrence in a form resembling a topographic map. The isopleth lines appear similar to the contour lines showing elevation. The final isopleth map is used as an overlay on the landslide hazard map.

It should be noted that the isopleth map does not alter the basic hazard zones determined previously. It is still an analytic map, which in this instance shows the varying prevalence of a specific landslide type in an area. It provides an additional criterion for the planner to make use of in deciding which area may be best suited for certain development activities. This is especially helpful in evaluating moderate hazard zones.

PHASE II: PROJECT FORMULATION AND ACTION PLAN DEFINITION QUESTIONS PLANNERS NEED TO ASK:

- Does the initial combined factor analysis landslide hazard map provide sufficient information to proceed with investment project formulation?

- If not already included, is there a hydrologic factor which could be added for further detail on hazard zones?

- Should an isopleth overlay be added to the hazard zonation map?

- Are there certain proposed land uses for which recommendations for mitigation should be included in the formulation of investment projects?

KEY DECISIONS TO BE MADE AT THIS STAGE:

- Who will execute the intermediate landslide hazard assessment?

- Which areas should be included in the additional assessment?

- Who will be responsible for incorporating additional information in investment project formulation activities

Where proposed land use is recognized as susceptible to a specific landslide type, the proposed activity is best located in a low hazard zone or moderate hazard zone with the lowest occurrence, i.e., smallest isopleth value, of that landslide type. The improved landslide hazard map and isopleth overlay require that an intermediate landslide inventory be prepared at this planning level. The landslide hazard map suitable for formulating development projects should be at a scale of 1:62,500 to 1:12,500 (see Figure 10-2).

d. Project Implementation

The landslide hazard map can contribute to planning for a project's implementation. There are two situations when this map may prove beneficial, both of which are related to mitigating the potential effects of landslides. In one case, it is conceivable that if areas identified with a moderate landslide hazard are targeted for development, greater detail of those areas is needed to ensure the project design compensates for this greater hazard potential. For example, moderate or higher hazard areas may not be entirely avoidable along a proposed road. Detailed investigation can provide information on groundwater conditions and on the stability characteristics of soil and rock to ensure a stable design (Morgenstern and Sangrey, 1978).

In another case, existing infrastructure or communities may be located in previously unidentified high hazard zones. These areas should be given priority for introducing some measure of mitigation.

For example, the effect of landslides issuing into an inhabited area from nearby mountain canyons might be mitigated by constructing debris basins to trap most of the material. Where such mitigation is impossible and the risk is identified as being extremely high, relocation to a safer area may be considered.

A detailed hazard map for the specific site in question is necessary at this stage of project design. Preparation of a detailed landslide inventory is now necessary. The large-scale features represented on landslides mapped in this detailed inventory are valuable for test drilling of a site and other sampling activities of engineering design work. Detailed landslide inventories and related interpretation of test results require map scales from 1:12,500 to 1:500 (see Figure 10-2).

The next section provides a detailed discussion of the types and nature of landslides, the basis for assessing landslide hazards, and the factors associated with landslide activity.

B. LANDSLIDES, LANDSLIDE HAZARD ASSESSMENT, AND AREAS OF CONCERN

1. Landslides and Landslide Susceptibility
2. Hazard Assessment of Landslides
3. Factors Associated with Landslide Activity

1. Landslides and Landslide Susceptibility

Landslides are caused when the force of gravity pulls rock, debris or soil down a slope. They are one of the forms of erosion called mass wasting, which is broadly defined as erosion involving gravity as the agent causing movement. Because gravity constantly acts on a slope, landslides only occur when the stress produced by the force of the gravity exceeds the resistance of the material. This is distinct from some other forms of erosion caused by running water, for instance, which occurs when precipitation falls on a slope or within a channel carrying a stream or river. Figure 10-3 depicts a list and diagram with terminology commonly used for describing landslides.

PROJECT IMPLEMENTATION

QUESTIONS PLANNERS NEED TO ASK:

- What type of landslide problems exist?

- What site-specific conditions need to be known for final design of an investment project with low landslide vulnerability?

KEY DECISIONS TO BE MADE AT THIS STAGE:

- Who will execute the detailed landslide hazard assessment?

- Which mitigation measures should be considered to reduce the risk to an acceptable level?

- Who will be responsible for incorporating the additional information in project implementation activities?

Landslide movement is perceptible and may take the form of falls, topples, slides, or flows. It can consist of free-falling material from cliffs, broken or unbroken masses sliding down mountains or hillsides, or fluid flows. Materials can move up to 120mph or more, and slides can last a few seconds or a few minutes, or can be gradual, slower movements over several hours or days. Accordingly, landslides are recognized on the basis of type of movement.

The most widely used classification scheme divides landslides into different types according to the material being moved and type of movement (Varnes, 1978). Speed of movement and amount of water mixed with the material are secondary parameters defining some landslide types. Recognizing the types of landslides presents in an area helps explain how and where factors have contributed to natural slope instability in the past.

Factors influencing where landslides occur can be divided into two sets, permanent and variable (Sharpe, 1938). Permanent factors are characteristics of the landscape which remain unchanged or vary little from a human perspective. The steepness of a slope or the type of rock, for example, presents changes only with the passage of long periods of time. Permanent factors such as rock type and slope steepness can be recognized and identified for specific landslides long after their occurrence (DeGraff, 1978). By examining existing landslides in an area, it is possible to recognize how permanent factors contributed to these slope failures. Identifying conditions and processes promoting past instability makes it possible to use these factors to estimate future landslides (Varnes, 1985).

Variable factors are landscape characteristics that change quickly as a result of some triggering event. Ground vibration due to earthquakes, a rapid rise in groundwater levels, and increased soil moisture due to intense precipitation are examples of variable factors. It is often necessary to be present at the time a landslide occurs or shortly thereafter to assess these factors. Due to the lack of long-term records relating landslide activity to historic earthquakes, storms, or other initiating factors, permanent factors are usually used to estimate landslide hazard. As such, identifying landslide areas is not an accurate science and leads, in general, to depicting hazard-prone areas based on an estimation. At best, landslide and landslide susceptible areas can be identified along with expected triggering events. At worst, some areas may not be detected at all.

2. Hazard Assessment of Landslides

Landslides are not currently amenable to risk assessment since there is no basis to determine the probability of landslides occurring within a given time period. Hazard assessments are possible and can be used in place of risk assessments. Hazard assessments are estimations of an area's susceptibility to landslides based on a few key factors. These are each capable of being mapped and allow land areas to be evaluated on their relative susceptibility to landslides.

Figure 10-3 DEFINITION OF BASIC LANDSLIDE TERMS

NOMENCLATURE

Main Scarp: A steep surface on the undisturbed ground around the periphery of the slide, caused by the movement of slide material away from undisturbed ground. The projection of the scarp surface under the displaced material becomes the surface of rupture.

Minor Scarp: A steep surface on the displaced material produced by differential movements within the sliding mass.

Head: The upper parts of the slide material along the contact between the displaced material and the main scarp.

Top: The highest point of contact between the displaced material and the main scarp.

Toe of Surface of Rupture: The intersection (sometimes buried) between the lower part of the surface of rupture and the original ground surface.

Toe: The margin of displaced material most distant from the main scarp.

Tip: The point on the toe most distant from the top of the slide.

Foot: The portion of the displaced material that lies downslope from the toe of the surface of rupture.

Main Body: That part of the displaced material that overlies the surface of rupture between the main scarp and toe of the surface of rupture.

Flank: The side of the Landslide.

Crown: The material that is still in place, practically undisplaced and adjacent to the highest parts of the main scarp.

Original Ground Surface: The slope that existed before the movement which is being considered took place. If this is the surface of an older landslide, that fact should be stated.

Left and Right: Compass directions are preferable in describing a slide, but if right and left are used they refer to the slide as viewed from the crown.

Surface of Separation: The surface separating displaced material from stable material but not known to have been a surface of which failure occurred.

Displaced Material: The material that has moved away from its original position on the slope. It may be in a deformed or unreformed state.

Zone of Depletion: The area within which the displaced material lies below the original ground surface.

Zone of Accumulation: The area within which the displaced material lies above the original ground surface.

Source: Adapted from Varnes, D. "Slope Movement and Processes" in Landslides: Analysis and Control, Special Report 176, Chapter 2 (Washington, D.C.: National Academy of Sciences, 1978).

TYPES OF LANDSLIDES

- Falls: a mass detaches from a steep slope or cliff and descends by free-fall, bounding, or rolling.

- Topples: a mass tilts or rotates forward as a unit.

- Slides: a mass displaces on one or more recognizable surfaces, which may be curved or planar.

- Flows: a mass moves downslope with a fluid motion. A significant amount of water may or may not be part of the mass.

Three principles guide landslide hazard assessment. First, landslides in the future will most likely occur under geomorphic, geologic, and topographic conditions that have produced past and present landslides. Second, the underlying conditions and processes which cause landslides are understood. Third, the relative importance of conditions and processes contributing to landslide occurrence can be determined and each assigned some measure reflecting its contribution (Varnes, 1985). The number of conditions present in an area can then be factored together to represent the degree of potential hazard present.

Landslide hazard has been determined with a high degree of reliability for only a few locations. These have required careful, detailed study of the Interaction of pertinent permanent and variable conditions in the target area. This can be a very expensive and time-consuming process that is unjustified for the purpose of broad-scale development planning. Landslide hazard zonation is one technique that can be used in the early stage of a planning study.

Most assessment procedures for landslide hazard zonation employ a few key or significant physical factors to estimate relative landslide hazard. The method described here requires a minimum of three factors mentioned earlier: distribution of past landslides, type of bedrock, and slope steepness, and a fourth, hydrologic factor may be added to reflect the important role which groundwater often plays in landslide occurrences (Varnes, 1985, and USGS, 1982).

Each factor is represented in a quantitative or semi-quantitative manner to facilitate the identification of different degrees of landslide hazard in an area.

Since all of these are permanent features, it is usually possible to map each factor. Specific combinations of these factors can be associated with differing degrees of landslide hazard. By extending these combinations over an entire area, a landslide hazard map is produced.

3. Factors Associated with Landslide Activity

a. Past Landslides and Their Distribution
b. Bedrock
c. Slope Steepness or Inclination
d. Hydrologic Factor
e. Human-Initiated Effects

The distribution of past landslides within the area, type of bedrock, and slope steepness represent, respectively, geomorphic, geologic, and topographic factors (Varnes, 1985, and USGS, 1982). Each of these factors is described in more detail below to give the planner a better understanding of their contribution to landsliding. The final section, "C. Mapping Physical Factors and Preparation of a Landslide Hazard Map," provides information on mapping them.

a. Past Landslides and Their Distribution

Interpreting the likelihood of future landslide occurrences requires an understanding of conditions and processes controlling past landslides in the area of interest. This can be achieved by examining and mapping past landslide activity in the area. Geologic, topographic, and hydrologic circumstances associated with past landslides indicate which natural or artificially created circumstances are likely to produce landslides in the future.

A primary consideration of the planner is the effect of existing land use on landslide activity. Certain types of landslides may be associated with specific land uses. For example, certain slides may only occur in road cuts or excavations. There may even be a critical height-to-inclination relationship for cutslopes below which these landslides will not occur. Field studies can provide insight into how different factors have contributed to failures. In some investigations special forms have been employed to ensure consistent collection of this ancillary information (see Figure 10-4). A summary of observations about landslide conditions and processes are incorporated into each landslide inventory, e.g., as in Pomeroy (1979), and mapped.

IN ORDER TO PLAN FOR DEVELOPMENT IN LANDSLIDE AREAS, A PLANNER NEEDS TO KNOW:

- Past distribution of landslides
- Types of bed rock
- How steep the slopes are
- Indirect measures of the area's hydrologic characteristics that are available
- Effects development activity can have on landslide susceptibility

b. Bedrock

Bedrock influences landslide occurrence in several ways. Weak, incompetent rock is more likely to fail than strong, competent rock. (See Figure 10-5 for an example of this.) On slopes where weak rock overlain by strong rock is exposed, the difference in strength increases the potential for landsliding in the stronger rock as well since the weak rock tends to erode and undermine the stronger rock.

The strength of a rock mass depends on the type of rock and the presence and nature of discontinuities such as joints or other fractures. The more discontinuities present in bedrock, the greater the likelihood of rock instability. Rock type may exert control on landsliding by influencing the strength of surface material in the area. For example, soils (in the engineering rather than agricultural sense of the term) derived from schists or shales will contain high percentages of clay. These soils will have different strength characteristics than coarser-grained soils such as those derived from granitic bedrock. There are many ways, then, that rock type or structure contributes to the instability, which can be represented on a map.

c. Slope Steepness or Inclination

The influence of slope steepness on landslide occurrence is the easiest factor to understand. Generally, steeper slopes have a greater chance of landsliding (see Figure 10-6). This does not prevent failures from occurring on gentler slopes. Other factors may make a gentle slope especially sensitive to failure, and thus in this situation could be determined to have a relatively high hazard potential.

For example, high ground water conditions occurring in sandy soils may liquefy during an earthquake. This can cause a landslide on a slope as gentle as 5 to 10 percent. Conversely, the steepest slopes may not always be the most hazardous. Steep slopes are less likely to develop a thick cover of superficial material conducive to certain types of landslides. Slope steepness can be mapped using generally available topographic maps.

d. Hydrologic Factor

Water is recognized as an important factor in slope stability-almost as important as gravity. Information on water table levels and fluctuations is rarely available. To represent the hydrologic factor in landslide hazard assessments, indirect measures can be used which can be mapped to show the influence of the area's hydrology, such as vegetation, slope orientation (aspect), or precipitation zones.

The type of vegetation and its density over an area will often reflect the variation in subsurface water. Certain species are water-loving or phreatophytes. Presence of these species shows near-surface water table conditions and springs. In mountainous regions, microclimatic differences produce different hydrologic conditions which in turn result in plant communities that vary according to the moisture available to the slope and its distribution throughout the year.

Slope orientation (aspect) refers to the direction a slope faces. It can be an indirect measure of climatic influence on the hydrologic characteristics of the landscape. Important characteristics associated with landslides are related to such factors as subsurface recharge resulting from prevailing winds and their influence on local frontal storms or accumulated snow. In other cases, a slope may experience more freeze/thaw cycles or wet/dry cycles which can reduce the strength of the soil and make the area more susceptible to landslides. In general, due to the complexity of these factors and existing development activities, there is usually no direct observable correlation between slope orientation and landslide hazard.

Figure 10-4 INVENTORY AND STATISTICAL ANALYSIS INFORMATION FORM FOR LANDSLIDES

Source: Carrara, A., and Merenda, L. "Landslide Inventory in Northern Calabria, Southern Italy" in Geological Society of American Bulletin, vol. 87 (1976), pp. 1153-1162.

Figure 10-5 - BEDROCK AS A FACTOR IN LANDSLIDE OCCURRENCE

Figure 10-6 - SLOPE STEEPNESS ASSOCIATED WITH LANDSLIDE ACTIVITY

e. Human-Initiated Effects

In addition to natural phenomena, human activities may increase the natural tendency for a landslide to occur. Landslides which result from development activities are usually the result of increasing moisture in the soil or changing the form of a slope. Development activities such as cutting and filling along roads and the removing of forest vegetation are capable of greatly altering slope form and ground water conditions (Swanson and Dyrness, 1975). These altered conditions may significantly increase the degree of landslide hazard present (Varnes, 1985, and Sidle, Pearce, and O'Loughlin, 1985).

For example, converting a forested area to grassland or one where crops are cultivated can increase the moisture in the soil enough to cause landslide problems (DeGraff, 1979). Or building a road which cuts off the toe of a steep slope can increase landslide susceptibility. It is possible to reduce the potential impact of natural landslide activity and limit development-initiated landslide occurrence by early consideration of these effects (Kockelman, 1985).

Now that the general points with regard to mapping the various land characteristics have been covered, the final section provides details on the techniques to do so in addition to presenting a step-by-step approach for preparing a landslide hazard map.

C. MAPPING PHYSICAL FACTORS AND PREPARATION OF A LANDSLIDE HAZARD MAP

1. Mapping the Physical Factors Associated with Landslides
2. Interpreting Landslide Hazards: The Landslide Hazard Map
3. Factor Analysis: The Technique to Prepare a Hazard Map
4. Compensating for Insufficient Data: The Isopleth Map
5. Computer-Generated Mapping

A landslide inventory produces a descriptive or data map (Cotecchia, 1978). By overlaying the landslide inventory map on the maps of the type of bedrock, slope steepness, and indirect hydrologic measures, the association of past landslides with the factors controlling landslide occurrence can be recognized. The method described below employs these associations in synthesizing a landslide hazard map. Extrapolating the data to areas with characteristics similar to those found associated with past landslides is an effective tool for forecasting where, but not when, landslides are more likely to occur in the future.

This section presents the techniques used to map each of the key factors associated with landslides. With these maps, a landslide hazard map can be prepared. Hazard zonation is a means of identifying areas with differing landslide hazards. The step-by-step approach, or factor analysis, used to prepare a landslide hazard map is described.

1. Mapping the Physical Factors Associated with Landslides

a. Mapping the Inventory of Existing Landslides
b. Mapping the Types of Bedrock Contributing to Instability
c. Mapping Slope Steepness or Inclination
d. The Optional Hydrologic Factor-Mapping Indirect Measures

Each factor is mapped separately by a different technique.

a. Mapping the Inventory of Existing Landslides

A map of existing landslides serves as the basic data source for understanding conditions contributing to landslide occurrence. Normally such a map is prepared by the interpretation of aerial photography and field examination of selected locations. While this map could also be compiled by field methods alone, the time and expense involved would only be justified by the unavailability of photo coverage. Either means of map preparation requires the skills of a geologist with experience in landslide or landform interpretation.

Aerial photography can serve as the source for data on existing landslides, type of bedrock, and vegetation cover. Typically, large-scale photography is necessary to be useful for existing landslides. The photo scale depends on the size of landslides common to the study area. Small-scale photography is less of a concern where bedrock and vegetation exist, since delineating areas with similar texture and appearance is easier than recognizing discrete features. Satellite imagery is generally unsuitable for landslide mapping except where data products can be enlarged to at least 1:50,000 scale. Photographic and satellite information is valuable in mapping other spatial information and for use in conjunction with computer mapping techniques as part of the development planning study (see Chapters 4 and 5 for a more detailed discussion).

Depending on vegetative cover, photo quality, and the skill of the interpreter, overall identification accuracy of 80 to 85 percent is realistic using aerial photography (Rib and Liang, 1978). The range of useful scales of aerial photography for landslide inventory work is limited to about 1:40,000 or larger. The selected scale will depend on the size of landslides common to the study area and, to some extent, the relief of the area. Large failures of four or more square kilometers are extremely difficult to detect on aerial photography smaller than 1:40,000. Where the majority of landslides are one hectare or smaller in size, large-scale photography on the order of 1:4,800 is necessary. The usefulness of black and white, color, or color-infrared photography for landslide inventory work will vary with local conditions and the individual making the interpretation. Each type of photography has advantages and disadvantages that will vary in their importance according to the characteristics of the area being mapped.

The map may be prepared at different levels of detail concerning existing landslides (USGS, 1982). A simple inventory identifies the definite and probable areas of existing landslides and is the minimum level required for a landslide hazard assessment. A map is produced in which each landslide is outlined and an arrow is drawn to denote the direction it moved. (See Figure 10-7 for a simple inventory map.)

More information can be provided by producing an intermediate inventory. The map produced at this level would show the outlined landslide types and distinguish between areas of landslide origin and deposit. The former is the area where material once existed as the source of the landslide and appears as a scar. The latter is deposited material from the landslide. (See Figure 10-8 for a sample intermediate inventory map.) The most information is obtained by producing a detailed inventory (Wieczorek, 1984). Large-scale features such as secondary scarps, sag ponds, and ground-crack patterns may be represented on individual landslides. (See Figure 10-9 for a detailed landslide map.)

These three types of inventories can be prepared as the development study progresses. To reiterate what was presented in Section A of this chapter: the simple inventory is adequate for Phase I development diagnosis activities; the intermediate inventory provides greater detail for an improved hazard map of a target area in Phase II; and the large-scale features of the detailed inventory are necessary for final project design in the implementation stage. Refer to Figure 10-2 for the appropriate map scales.

There are several considerations to keep in mind when gathering data on existing landslides. First, the time and effort required to conduct an inventory varies with (1) geologic and topographic complexity; (2) size of an area; and (3) desired level of inventory detail (Varnes, 1985). Figure 10-10 characterizes the relationship between the amount of time and level of effort for these three variables. Second, more detailed inventories will require larger map scales to reveal the small features of this added detail. Third, additional data gathering can add detail to an existing inventory. This enables a previously completed simple inventory to be transformed into an intermediate inventory with less time and effort than producing the intermediate inventory solely from field work and aerial photography.

b. Mapping the Types of Bedrock Contributing to Instability

By using bedrock as one factor in the landslide hazard assessment, the many different ways rock type or structure contribute to instability are represented. Comparing a bedrock map with the landslide map, one can discriminate between rock units associated with existing landslides and those devoid or largely free of landslide activity.

To produce a usable bedrock map for a hazard assessment, bedrock unit boundaries should be traced to produce new, more suitable units. Existing standard geologic maps define units according to factors such as age, composition or lithology (rock type), and structure (faulting, folding, etc.). For example, a standard geologic map may show a series of volcanic ash deposits of similar mineral compositions which vary only slightly in age. In most instances, these different units will affect landslide occurrence in a similar way and should be delineated as a single bedrock unit in a revised map for hazard assessment work. The geologist must use professional judgement to ensure that the number of bedrock units is sufficient to distinguish differences in their effect on landslide occurrence.

When a geologic map does not exist, a bedrock map based on aerial photography with limited field verification is needed. This map may be no more detailed than a delineation of sedimentary, igneous, and metamorphic rock types. Obviously, a bedrock map generalized from a more detailed map is preferable, but this is an acceptable alternative under such circumstances. Delineating areas with similar texture and appearance is easier than recognizing discrete features. Scales as small as 1:62,500 are useful for this work. Photos at scales of 1:20,000 or larger are difficult to use because the limited area shown restricts comparison with adjacent contrasting areas. It also significantly increases the number of photographs to examine in mapping the area. Black and white, color, and color infrared photography are all suitable for bedrock mapping. Satellite imagery is generally unsuitable for this mapping except when the imagery is enlarged to usable scales. For example, imagery at a 1:50,000 scale produced from satellite imagery is acceptable for this mapping (see Chapter 4).

A soils map is an inadequate substitute for a bedrock map. Soils maps are based on factors concentrated in the upper meter or less of superficial material that affects agricultural activities. Generally, there is little or no correlation between "agricultural" soil characteristics and the likelihood of failures originating along surfaces a few meters to tens of meters deep in superficial material.

Figure 10-7 SIMPLE LANDSLIDE INVENTORY MAP

Figure 10-8 INTERMEDIATE LANDSLIDE INVENTORY MAP

Figure 10-9 - DETAILED LANDSLIDE INVENTORY MAP

c. Mapping Slope Steepness or Inclination

Slope steepness is a factor which associates the effectiveness of gravity acting on a slope to landslide susceptibility. A topographic map is the source for preparing a slope steepness map. The slope steepness map displays the steepness values associated with the majority of existing landslides and is derived from an existing topographic map. Steepness for landslide hazard assessments is commonly expressed as a percentage rather than in degrees. The categories or grouping of steepness values for use in analyzing landslide hazards should approximate those of the slopes present in the study area. Too many classes will make it difficult to identify slopes critical to landslide occurrence and too few will be equally useless.

d. The Optional Hydrologic Factor-Mapping Indirect Measures

Since information on water table levels and fluctuations is rarely available, mapping indirect measures such as vegetation and slope orientation can reveal the influence of hydrology on an area. Any vegetation map used to represent the hydrologic factor in the landslide hazard assessment must employ units that are dependent on water. This may be as simple as representing phyreatic and non-phyreatic plant communities or as complex as distinguishing different forest types. Selection of the appropriate vegetation map units to indicate the effects of water in causing landslides requires careful field observation by the geologist.

Aerial photography is an appropriate source of data for preparing vegetation maps. In preparing vegetation maps, as in mapping bedrock, scale is less of a concern. Here, too, delineating areas with similar features is easier than recognizing discrete features. Scales of 1:62,500 are useful in identifying vegetation since 1:20,000 and larger scales do not reveal the contrasting characteristics of adjacent areas. Black and white, color, and color-infrared photography are all suitable for this mapping. Satellite imagery is acceptable only when enlarged to usable scale.

The direction in which a slope faces can also be -mapped and used as an indirect indicator of the hydrologic factor. Slope orientation or aspect is described in terms of the eight cardinal directions, i.e., north, northeast, etc. For convenience in establishing a data base, slope orientation is measured in degrees of azimuth from 0 to 360 degrees. Each cardinal direction is defined by a set of azimuth values. For example, slopes facing the northeast can have an azimuth reading ranging from 22.5 degrees to 67.5 degrees.

2. Interpreting Landslide Hazards: The Landslide Hazard Map

A landslide hazard map is generated to identify areas with differing landslide hazards. A hazard map is produced for each stage of the planning process, from the more generalized map in the initial stage to a detailed zonation map for specific site use. As the name suggests, this map divides the entire study area into sub-areas based on the degree of a potential hazard from landslides. The landslide hazard map is produced by interpreting the data represented by the maps of the inventoried landslides and the permanent factors found to influence the occurrence of landslides.

As with any map, scale is an important consideration. There are two points to keep in mind concerning the scale of the landslide hazard map. First, such a map should be produced at a scale capable of representing the information needed at a particular planning level. Compatibility of scale would be important when the hazard map is to be combined with other maps to yield a land capability map (see Chapter 3). Second, the landslide hazard map should be at a scale not markedly different from the data maps used to produce it. In other words, reliability may be questionable when a landslide hazard map produced at a scale of 1:50,000 has been based on a 1:250,000 slope steepness map.

Four levels of relative hazard are identified on a landslide hazard map: (1) low; (2) moderate; (3) high; and (4) extreme hazard. The level of landslide hazard is measured on the ordinal scale with this method and is a quantitative representation of differing hazard levels that shows only the order of relative hazard at a particular site and not absolute hazard. Predicting absolute hazard is impractical with current capabilities.

As a consequence, there is no way to compare hazard zones at different sites or to determine the likelihood that a high hazard area, for example, is two or ten times more likely to fail in the future than low hazard areas. It should be stressed that these relative hazard zones are based on the existing landslides and conditions influencing their occurrence in a specific area. The hazard zones which are determined for an area hold true only for the area for which they were prepared. Similar conditions found outside the assessed area may not produce the same degree of hazard because of some seemingly minor difference in one of the factors.

3. Factor Analysis: The Technique to Prepare a Hazard Map

a. Step One: Combined Map of Permanent Factors
b. Step Two: Overlay of Landslide Inventory
c. Step Three: Group Combinations Using a Factor Analysis
d. Step Four: Producing Landslide Hazard Zones

A factor analysis is a step-by-step approach used to prepare a landslide hazard zonation map of an area.

There are four steps to complete the factor analysis and produce a hazard map: (1) map the existing landslides and prepare a map combining the permanent factors (bedrock, slope steepness, and, when available, the hydrologic factors) into individual map units; (2) overlay the landslide inventory on the combined factor map; (3) prepare a combined factor analysis for all combinations of the factors and group combinations of these factors in a way that defines the four levels of landslide hazard; and (4) produce a map with four landslide hazard zones from the grouped combinations.

a. Step One: Combined Map of Permanent Factors

The first step is to prepare a map of the inventoried existing landslides. Also, compile a map which combines the bedrock, slope steepness, and, when included, hydrologic factor units or categories into individual cartographic units. As an example, assume that only bedrock and slope steepness are being used. The compiled map will be composed of cartographic units delineating certain bedrock type and slope values, e.g., Bedrock B3 on slopes between 25-50 percent (see Figure 10-10).

Figure 10-10 STUDY AREA MAP

Legend

A representation of how the proportion of bedrock slope combinations subject to past landslide activity is determined. Note that while combination B3 obviously has more landslides than combination C4, the smaller size of C4 area will likely result in its having a higher proportion than B3.

b. Step Two: Overlay of Landslide Inventory

The second step is to overlay the landslide inventory map with this combined factor map. This will identify which combinations are associated with past landslides and which are not. A landslide inventory table is developed indicating total area of landslides occurring on each specific bedrock unit and slope steepness combination (and other factors, if considered) (see Figure 10-12). When a hydrologic factor such as vegetative zone or slope orientation is used, the table will include the area of landslides for each specific combinations of bedrock, slope steepness, and the hydrologic factor. Summing the areas from all combinations found in the table will yield the total area of landslides in the study area. This is a way to check that all combinations are included in the analysis. Figure 10-11 shows the extent to which each combination is present in the study area. For example, Bedrock B on slopes between 25 and 50 percent has 784 hectares of landslides.

c. Step Three: Group Combinations Using a Factor Analysis

The third step is to group combinations of these factors in a way that will define four levels of landslide hazard. This grouping is achieved by performing a combined factor analysis or matrix assessment (DeGraff and Romesburg, 1980). This analysis permits incorporating the interaction among factors affecting landslide occurrence without explicitly understanding those interactions.

To start, measure the total area for every combination of bedrock, slope steepness, and hydrologic factors in the study area represented in the table prepared in Step 2. The total area with these combinations is to be calculated, not just those associated with landslide activity. Continuing with the example, assume a total area of 2,327 hectares of bedrock B on slopes greater than 25 percent but less than 50 percent was found. The landslide inventory table prepared in Step 2 shows only the area of past landslides present for each combination. Then the total area for every combination associated with landslides found in the landslide inventory table is divided by the area for the same combination of factors found in the study area (see Figure 10-12). In the example, this would be 784 divided by 2,327. This yields a proportion of each combination which is subject to past landslide occurrence, e.g., 0.34. This represents the proportion of the combination disturbed by past landslides in that area (see Figure 10-11).

The combination of bedrock, slope steepness, and hydrologic factors associated with the largest area disturbed by landslides may not, in fact, be the most hazardous: it may simply be the combination which is most common to the study area. Since such an area is the most prevalent combination, it has the greatest chance for being associated with past landslides rather than being most hazardous. The process described above ensures that comparison of landslide hazard among different combinations takes place on an equal basis.

Figure 10-11 - COMBINED PERMANENT FACTOR (SAMPLE BEDROCK AND SLOPE CLASS) AND LAND AREA COVERAGE (IN HECTARES)

SLOPE CLASS

BEDROCK GROUP

0<12%
(1)

12<25%
(2)

25<50%
(3)

>50X
(4)

TOTAL AREA (HA)

A


-

52

78

-

130 Landslide Area

1,570

722

512

237

3,041 Combined Area a/

B


-

301

784

-

1,085 Landslide Area

-

1,776

2,327

-

4,103 Combined Area a/

C


78


351

180

609 Landslide Area

673

2,450

1,790

793

5,706 Combined Area a/

a/ Combined Area = Combined Permanent Factor Area

Figure 10-12 LANDSLIDE HAZARD ZONES

There will be a proportional value for each combination of bedrock, slope steepness, and other factors associated with existing landslides ranging from 0.1. to 1.0. The proportions are sorted from the smallest to the largest. This range of values is broken into three groups to represent the relative landslide hazard in the study area. To ensure that the points used to define the three groups are determined objectively, a non-hierarchical cluster analysis is used. (See the Appendix of this chapter for a sample computation.)

An initial division into three groups is achieved by breaking equally the range of proportional values present. The upper and lower boundaries of each group are retained or adjusted to ensure that the final division represents the minimum sum of the squared deviations around the three group means. This is based on the W function (Anderberg, 1973).

d. Step Four: Producing Landslide Hazard Zones

The fourth, and final, step uses the grouped combinations to produce landslide hazard zones-extreme, high, moderate, and low. Once the proportions are divided into three groups, bedrock, slope steepness, and hydrologic factor combinations representing different degrees of relative landslide hazard are identified. The group of proportions with the larger values, i.e., toward the 1.0 end of the range, represent combinations defining extreme landslide hazard. The group of proportions with the next smaller values represents combinations defining high landslide hazard. The group of proportions with the smallest values, i.e., toward the 0.1 end of the range, represents combinations defining moderate landslide hazard. All bedrock, slope, and hydrologic factors not found to be associated at all with existing landslides define low landslide hazard.

THE W FUNCTION TO DEFINE FOUR LEVELS OF LANDSLIDE HAZARD

The map overlays used to determine areas of bedrock, slope steepness, and hydrologic factor present in the entire study area can now be revised to make the hazard zonation map. Figure 10-12 shows the original maps redrawn into hazard zones. Combinations with extreme hazard are redrawn and relabeled as extreme hazard zones. Redrawing and relabeling for combinations representing other hazard zones produces a completed hazard zonation map displaying four levels of relative hazard. The empirical relationship of the physical factors, as defined by the factor analysis, is valid for only the area evaluated, and cannot be extrapolated to cover additional areas.

Once these hazard areas are identified, a decision can be made regarding the appropriate development activities, type of mitigation measures to be included in the process, or the areas which should be avoided. It is important to note that the essential bedrock and slope steepness maps are not always available. Without these maps, an isopleth map can be produced which is an acceptable substitute.

4. Compensating for Insufficient Data: The Isopleth Map

In the absence of bedrock and slope steepness maps, the landslide inventory map can be used to produce an analytical map suitable for representing landslide activity in an area. An isopleth map of landslide frequency is recommended for this purpose. An isopleth or any other analytic map can only serve as an initial assessment of landslide activity and not as a substitute for a landslide hazard map. The underlying conditions producing landslides will remain unknown and prevent making the distinction between the relative degrees of landslide hazard.

It is reasonable to assume that areas with a high frequency of landslide activity represent areas with a greater chance of future landslides than those areas with a low frequency. An isopleth map can be made based on this assumption. Preparing an isopleth map begins with the map of inventoried landslides (Wright et al, 1974). A transparent overlay with a 2cm x 2cm grid is placed over the landslide inventory map. (See Figure 10-13 for a graphic depiction of each step.) At each grid intersection a transparent gridded circle 2.5cm in diameter is centered on each grid intersection on the transparent overlay. The number of grid squares in the circle through which landslide deposits are visible is counted. Divide this number by the total number of grid squares within the inscribed circle. This yields the proportion of the unit area within the circle that is underlain by landslide deposits. This proportion is multiplied by 100 and rounded off to the nearest whole number to compute the percentage of landslide-disturbed terrain. The percent value is written on the gridded overlay next to the grid intersection.

Once all grid intersections are marked with percent values, the isopleth lines can be drawn. Isopleth lines connect the points of equal value. These show the generalized frequency of landslide activity as represented by the percent of landslide-disturbed area. The interval between isopleths drawn to produce the map will depend on the proposed use. A single value representing a boundary between areas of frequent landsliding and infrequent landsliding shows areas where this phenomenon is a major factor in shaping the landscape and areas where it is not. This serves as an initial assessment of areas subject to landslide problems when additional factors are not available for a study area. It is important to remember that this is an analytic technique producing a limited assessment of an area rather than a technique developed by an interpretative process.

During Phase II of the planning process, in addition to the intermediate landslide inventory, the preparation of an isopleth map which would enhance the information available to planners is recommended. Using the technique described above, preparation is altered in two ways: (1) only the specific landslide types identified in the intermediate inventory which are likely to be initiated by the land use proposed would be used in compiling this isopleth map; the choice of landslide types should be governed by the information on landslide activity developed by the geologist completing the intermediate inventory of existing landslides and by existing and proposed land use; and (2) isopleths are drawn at regular intervals similar to the way elevation is represented by a contour interval instead of the single value used in the isopleth map. For example, an interval of 10 percent has been used with some isopleth maps applied to land-use planning (Campbell, 1980, and Pomeroy, 1978). This produces a map representing the intensity of past landslide occurrence in a form resembling a topographic map. The isopleth lines would appear like the contour line showing elevation. The final isopleth map is used as an overlay on the landslide hazard map.

Figure 10-13 STEPS TO PREPARATION OF AN ISOPLETH MAP

5. Computer-Generated Mapping

The method described in this chapter can be readily adapted to computer-generated mapping (Brabb, 1984). The factor maps used to generate the landslide hazard map can be encoded to a geographic information system (GIS) and manipulated by a computer. (See Chapter 5 for a discussion of computer mapping applications and GIS.) This enables the rapid preparation of tables showing the area for different factor combinations. In some cases, data maps used in landslide hazard assessment may be part of the GIS created for generated land-use planning, for instance a vegetation map. A second advantage of this approach is that the scales for maps to be overlaid in a landslide hazard assessment can be matched regardless of their original scale. For example, the scale of a published bedrock map may differ from the other factor maps. Using manual techniques, redrafting the bedrock map at the same scale would be necessary, whereas a computer-based system permits the matching of map scales regardless of their original scale and the maps can be overlaid.

Computerized matching of map scales requires that certain reference points on each map be identified to ensure proper registration of points between maps. Once maps are computerized, they are capable of being updated or used to improve landslide hazard assessments. A more detailed landslide inventory map could be encoded and used to produce an improved hazard zonation map with the already encoded data maps.

The single major limitation of using a computer-based system is the amount of time and expense that is required to encode the maps and establish the data base for a landslide hazard assessment at a scale sufficiently large to permit the calculation of the percentage of the area covered by existing landslides. Creating such a data base usually dictates that a major project or series of projects be planned to justify this commitment of resources, or that a data base of computerized maps already exists. One final consideration is the ability to gain access to computer equipment, since computers may be scarce or in great demand for many uses. Nevertheless, readily available and relatively inexpensive micro-computers and software programs which are adequate for a landslide hazard assessment make it possible for some planning studies to have their own system.

CONCLUSION

Areas susceptible to landslides can be projected, based on the physical factors associated with landslide activity: past landslide history, bedrock, slope steepness, and hydrology. Predicting where and when landslides are going to occur is not possible even with the best available information. It is, however, possible to identify landslide-susceptible areas. This chapter has discussed some of the concepts related to landslide susceptibility: the different types of landslides; the relative nature of landslide hazard zonation; its relationship to development activity; and ways to mitigate the effects of landslides. The essential point has been to demonstrate the importance of considering landslides early in the planning study and to provide one technique which can be used at all stages of the planning process. The different questions that need to be asked at the different planning stages were highlighted. Many answers can be generated from the use of landslide hazard zonation at each stage of the planning study. The step-by-step combined factor analysis to prepare hazard maps was presented. All of this will enable the planner to have a working knowledge of terms, concepts, and the important considerations related to landslides and landslide hazard mapping.

REFERENCES

Anderberg, M.R. Cluster Analysis for Applications (New York: Academic Press, 1973).

Brabb, E.E. "Innovative Approaches to Landslide Hazard and Risk Mapping" in IV International Symposium on Landslides, vol. 1 (Toronto, 1984), pp. 307-323.

Campbell, R.H. Landslide Maps Showing Field Classifications, Point Dume Quadrangle, California, U.S. Geological Survey Field Studies Map MF-1167 (Reston, Virginia: U.S. Geological Survey, 1980).

Carrara, A., and Merenda, L. "Landslide Inventory in Northern Calabria, Southern Italy" in Geological Society of America Bulletin, vol. 87 (1976), pp. 1153-1162.

Cotecchia, V. "Systematic Reconnaissance Mapping and Registration of Slope Movements" in Bulletin of the International Association of Engineering Geology, no. 17 (1978), pp. 5-37.

DeGraff, J.V. "Regional Landslide Evaluation: Two Utah Examples" in Environmental Geology, vol. 2 (1978), pp. 203-214.

- "Initiation of Shallow Mass Movement by Vegetative-type Conversion" in Geology, vol. 7 (1979), pp. 426-429.

- "Quantitative Approach to Assessing Landslide Hazard to Transportation Corridors on a National Forest" in Transportation Research Record 892 (1982), pp. 64-68.

DeGraff, J.V., and Romesburg, H.C. "Regional Landslide-Susceptibility Assessment for Wildland Management: A Matrix Approach" in D.R. Coates, and J. Vitek (eds.), Thresholds in Geomorphology (Boston: George Alien & Unwin, 1980), pp. 401-414.

Hutchinson, J.N., and Kogan, E. 'The Mayunmarca Landslide of 25 April 1974" in UNESCO Serial No. 3124/RMO.RD/SCE (Paris: UNESCO, February, 1975).

Hopkins, L.D. "Methods for Generating Land Suitability Maps: A Comparative Evaluation" in American Institute of Planning Journal, vol. 43 (1977), pp. 386-400.

Kockelman, W.J. "Some Techniques for Reducing Landslide Hazards" in Bulletin of the Association of Engineering Geologists (vol. 22, 1985).

Morgenstern, N.R., and Sangrey, D.A. "Methods of Stability Analysis" in R.L Schuster, and R.J. Krizek (eds.), Landslides, Analysis, and Control, Special Report 176 (Washington, D.C.: Transportation Research Board, 1978), pp. 155-171.

Organization of American States. Integrated Regional Development Planning: Guidelines and Case Studies From OAS Experience (Washington, D.C.: Organization of American States, 1984).

Pomeroy, J.S. Isopleth Map of Landslide Deposits, Washington County, Pennsylvania, U.S. Geological Survey Field Studies Map MF-1010 (Reston, Virginia: U.S. Geological Survey, 1978).

- Map Showing Landslides and Areas Most Susceptible to Sliding in Beaver County, Pennsylvania, U.S. Geological Survey Miscellaneous Investigations Series Map 1-1160 (Reston, Virginia: U.S. Geological Survey, 1979).

Rib, H.T., and Liang, T. "Recognition and Identification" in R.L. Schuster and R.J. Krizek (eds.), Landslides, Analysis, and Control, Special Report 176 (Washington, D.C.: Transportation Research Board, 1978), pp. 34-80.

Sharpe, C.F.S. Landslides and Related Phenomena (New York: Columbia University Press, 1938).

Sidle, R.C., Pearce, A.J., and O'Loughlin, C.L. Hillslope Stability and Land Use, Water Resources Monograph Series No. 11 (Washington, D.C.: American Geophysical Union, 1985).

Swanson, F.J., and Dyrness, C.T. "Impact of Clearcutting and Road Construction on Soil Erosion by Landslides in the Western Cascade Range, Oregon" in Geology, vol. 3 (1975), pp. 393-396.

U.S. Geological Survey. Goals and Tasks of the Landslide Part of a Ground-Failure Hazards Reduction Program, U.S. Geological Survey Circular 880 (Reston, Virginia: U.S. Geological Survey, 1982).

Varnes, D.J. "Slope Movement Types and Processes" in R.L. Schuster and R.J. Krizek (eds.), Landslides, Analysis, and Control, Special Report 176 (Washington, D.C.: Transportation Research Board, 1978), pp. 12-33.

- Landslide Hazard Zonation: A Review of Principles and Practices, UNESCO Natural Hazards Series No. 3 (Paris: UNESCO, 1985).

Wieczorek, G.F. "Preparing a Detailed Landslide-Inventory Map for Hazard Evaluation and Reduction" in Bulletin of the Association of Engineering Geologists, vol. 21 (1984).

Wright, R.H., Campbell, R.H., and Nilson, T.H. "Preparation and Use of Isopleth Maps of Landslide Deposits" in Geology, vol. 2 (1974), pp. 483-385.

APPENDIX

SAMPLE COMPUTATION OF W FUNCTION

As noted in Section C-3, Factor Analysis, the W function is computed from the formula:

where:
Xij = jth observation ith group
ni = number of observations in the ith group

For the example, it is assumed the combined factor analysis yielded the following sixteen proportions:

.53, .01, .19, .03, .39, .04, .05, .88, .11, .01, .21, .03, .61, .01, .04, .11

Step 1: The proportions are then arranged in ascending order:

.01, .01, .01, .03, .03, .04, .04, .05, .11, .11, .19, .21, .39, .53, .61, .88

The data range from 0.1. to .88. This range is divided equally to form three groups based on an equal interval partition: .01£X<.29, .29£X<.58, and .58£X<.88.

Step 2: The W factor is computed using the values in each group formed under the initial equal interval partition:

[.01£X<.29]

[.29£X<.58]

[.58£X<.88]

.01,.01,.01,

.39,.53

.61,.88

.03,.03,.04,



.04,.05,.11,



.11,.19,.21



X1 =.07

X2 =.46

X3=.745

W1 =.0534

W2 =.0098

W3=.0365

W = W1 + W2 + W3 =.0534 +.0098 +.0365=.0996

The objective is to minimize the value of W. In other words, find the smallest W values that can be computed for three groups of the proportional values. This applies the principle of least squares, a common statistical approach, to this one-dimensional problem through minimizing the sum of squared deviations about the around means.

Step 3: The boundary is shifted to the right to seek the desired decrease in W function:

[.01£X<.39]

[.39£X<.58]

[.58£X<.88]

.01,.01,.01,

.53

.61,.88

.03,.03,.04,



.04,.05,.11,



.11,.19,.21,



.39



X1 =.0946

X2 =.53

X3 =.745

W1 =.1479

W2 = 0

W3 =.0365

W = W1 + W2 + W3 =.1479 + 0 +.0365 =.18435

Because the recomputed value is more than the W value initially computed, this is the wrong direction move. The boundary will be shifted to the left of the initial boundary seeking a decrease in the W value.

Step 4: The left-most boundary is moved to the left by one value. The W function is recomputed and compared to the initial W value to determine whether the desired decrease occurred:

[.01£X<.19]

[.19£X<.58]

[.58£X<.88]

.01,.01,.01,

.21,.39,.53

.61,.88

.03,.03,.04,



.04,.05,.11,



.11,.19



X1 =.0573

X2 =.3767

X3 =.745

W1 =.0320

W2 =.0515

W3=.0365

W = W1 + W2 + W3 =.0320 +.0515 +.0365=.12

This is not a decrease. Therefore, the partition for the left-most boundary is kept at the initial value.

Step 5: Now the second or right-most boundary is moved to the right:

[.01£X<.29]

[.29£X<.61]

[.61£X<.88]

.01,.01,.01,

.39,.53,.61

.88

.03,.03,.04,



.04,.05,.11,



.11,.19,.21



X1 =.07

X2 =.51

X3 =.88

W1 =.0534

W2 =.0248

W3 = 0

W = W1 + W2 + W3 =.0534 +.0248 + 0 =.0782

This is a decrease in the W value. If any other values remained in the third group, the boundary would be shifted in single moves to the right until no further decrease in W values was obtained. With no other values present, this minimizes the sum of squared deviations about the group means to the greatest extent possible and retains three groups. If the shift to the right had resulted in an increased W value, a move to the left on the right-most boundary would have been tried. Having determined the boundaries for obtaining the smallest W value, the best grouping of the proportional values present is achieved.

As a result of this iterative process, the initial partition into groups with the following ranges:

.10£X<.29
.29£X<.58
.58£X<.88

is changed to a grouping more consistent with the proportional values involved based on the range of values below:

.01£X<.29
.29£X<.61
.61£X<.88

Previous Page Top of Page Next Page