Ministry of Environment Terrain

Terrain Stability Maps

This Chapter summarizes various aspects of terrain stability maps: uses (Section 3.1), types (Section 3.2), terrain attributes (Section 3.3), map scales (Section 3.4), and map units (Section 3.5). The following Chapter summarizes various methods of terrain stability mapping.

3.1 Uses

Terrain stability maps are used for a variety of purposes that can be compiled into three groups:

• resource development planning;
• land use and development planning; and
• linear project planning.

3.1.1 Resource Development Planning

Presently in British Columbia, of all resource industries, the forest industry makes the greatest use of terrain stability maps for development planning.

The forest industry uses these maps:

• to assist with establishing cutblock boundaries, road alignments, and timber harvesting systems to minimize future landslides;
• to predict areas where landslides may occur and/or to predict the severity of landslides in response to road construction, or during or following logging;
• to assist with road deactivation plans; and
• to locate forestry camps, mill sites and other facilities.

The first systematic method of terrain stability mapping in British Columbia was developed by MacMillan Bloedel Ltd for coastal areas (Bourgeois 1978). Over the years the BC Ministry of Forests and other forest companies have adopted and adapted similar mapping methods. The "Mapping and Assessing Terrain Stability Guidebook" associated with the Forest Practices Code (BC Ministry of Forests 1995a) has recently summarized and standardized this method of mapping. Similar methods are used elsewhere in the Pacific Northwest. See for example Duncan (1989).

Since the early 1980s, research has been carried out in coastal British Columbia on the prediction of landslides in clear cut areas based on the study of a variety of terrain attributes. Examples include Rollerson and Sondheim (1985), Howes (1987) and Rollerson (1992). It is anticipated that such 'terrain attribute studies' will help refine the methods of pre-logging terrain stability mapping.

Gully sidewall and channel stability, as well as erosion and sediment delivery to streams, are also concerns of the forest industry. Methods have been suggested in the province and elsewhere to map and assess gully sidewall and channel stability, and erosion potential and the potential for sediment delivery of the eroded material to streams. Examples include Hogan and Wilford (1989) and the California State Board of Forestry (1990). These methods, although not yet routine, are gaining general acceptance. The "Gully Assessment Procedure Guidebook" associated with the Forest Practices Code (BC Ministry of Forests 1995b) is an initial attempt to standardize the assessment of gully sidewall and channel stability.

Recently terrain stability mapping in the forest industry has been included as a component of the multi-disciplinary approach to 'cumulative effect' assessments that also include sediment production, effects on wildlife and vegetation, fire hazards, soil degradation and other environmental concerns. Examples include Acres International Ltd (1993) and Washington Forest Practices Board (1993). Landslide hazards constitute one of the most important components of cumulative effects of forestry activity.

3.1.2 Land Use and Development Planning

The purpose of terrain stability mapping in land use and development planning is to delineate areas where existing and/or land development may be affected by landslide hazards, and where land development may affect slope stability. These areas include land on steeper slopes, at the breaks in slope, along the base of slopes and land on colluvial and alluvial fans. The ultimate aim of such mapping is to assist with the planning or regulation of land use and development.

Many regional terrain stability mapping programs for land use planning were implemented in Europe and the United States in the 1970s and 1980s. Tippett and Roberts (1992) reviewed such programs in Austria and France. Since 1975, Austria has required the preparation of plans of natural hazards including debris flows, snow avalanches and flooding. France in the late 1970s published a series of terrain stability maps at 1:10,000 to 1:25,000 scales, locally referred to as ZERMOS (zones exposed to risk of soil and sub-soil movement) maps. Since 1982 a French law has required natural hazard prediction (PER) maps at 1:2,000 to 1:10,000 scales be prepared for areas prone to natural hazards. Each map is accompanied by a report and four additional maps: a land use map, a process inventory map, a probability of occurrence map, and a 3-class zoning map (Perrot 1988).

In San Mateo County, California in the early 1970s a detailed terrain stability map using a 7-class zoning system was prepared to reduce new development in hazardous areas and to encourage site specific studies (Brabb et al 1972). The program was extended to include the entire San Francisco Bay region. Separate zoning maps addressed landslides (Brabb 1991), debris flows (Mark 1992) and seismically triggered landslides (Wieczorek et al 1985).

In British Columbia, the BC Ministry of Transportation and Highways, which has jurisdiction over subdivisions in unorganized areas, carried out a number of pilot terrain stability mapping projects in the 1970s and 1980s. For examples see Buchanan (1977, the South Thompson Valley) and Haughton (1978, the Columbia Valley).

In organized areas of the province, residential subdivision and building permit approval procedures are mandated by the Municipal Act. This act is implemented by the appropriate ministries, regional districts and/or municipalities. Currently joint provincial-community funding supports terrain stability mapping for communities that adopt Official Community Plans (OCP). Terrain stability maps are used to delineate areas where restrictions should be applied under the act and where site specific assessments are required. For example see Cave (1992, Fraser-Cheam Regional District).

The extensive 1:50,000 scale terrain mapping carried out by the BC Ministry of Environment in the 1970s and 1980s, following Environment and Land Use Secretariat (1976) and Ryder and Howes (1984), has been used to derive terrain stability information for land use planning. For examples see Maynard (1979) and Ryder and MacLean (1980). This derived information is intended for guidance, without direct administrative or regulatory implications.

Many site-specific terrain stability mapping projects have also been carried out throughout the province, usually by geotechnical engineers and geoscientists, to aid land use planning and regulation.

3.1.3 Linear Project Planning

Terrain stability maps are usually produced prior to the location of a linear project, such as a road, railway, pipeline or transmission line, to help choose the optimal alignment. Examples include Buchanan (1990, a highway), Thurber Engineering Ltd (1989, a resource road), and Pacific Hydro Consultants Ltd (1989, a transmission line). These types of terrain stability maps often include project-specific interpretative comments.

Terrain stability maps are also often produced for planning purposes along linear geomorphic features such as streams, shorelines and reservoirs. Such mapping has been applied to BC Hydro's reservoir shorelines since the early 1970s to assess existing terrain stability and to predict terrain stability after reservoir flooding. Examples include Morgan (1982), Enegren and Moore (1990) and BC Hydro (1993). These maps often delineate a 'safe line', 'break line', and/or 'impact line' (described in Section 6.3) along the crests or toes of bluffs, or along existing or future shorelines.

Snow avalanche 'atlases' serve similar planning purposes. Several have been prepared for specific highways in the province by the BC Ministry of Transportation and Highways, for example the Coquihalla Highway (BC Ministry of Transportation and Highways 1980), and by Environment Canada, for example the Rogers Pass section of the Trans-Canada Highway (Schleiss 1989).

Linear terrain stability maps are also prepared during the operation of certain facilities to help monitor landslide activity and establish priorities for maintenance. An example is the rock fall hazard map for an active French road, produced by the French Road Research Laboratory (Einstein, 1988).

3.2 Types

There are a number of different types of maps that can provide information on terrain stability. For the purpose of this study they are grouped into seven types. Types 1 through 5 are maps that delineate the distribution of particular landslide data, or terrain attributes, and may be accompanied by some form of data base. When interpreted or combined with other information these map types can become, or can be used as, terrain stability maps that address either landslide hazards or risks. Type 6 is a specific type of terrain stability map that addresses landslide hazards. Type 7 is a specific type of terrain stability map that extends the landslide hazard assessment by considering the consequences of the hazards, and therefore is a landslide risk map.

Type 1 -- Geology maps delineate bedrock and/or surficial geology units, usually on the basis of relative geological age. Certain map units, specific symbols and/or marginal notes may indicate, or be used as a rough guide to indicate, the distribution of landslide hazards.

Bedrock structure and lithology are often significant in controlling the character of large bedrock landslides. In glaciated terrain, landslide hazards commonly occur in association with certain surficial geology units.

Type 2 -- Terrain maps delineate surface units based on a number of terrain attributes, including material genesis and texture, surface expression and geomorphic process. Several terrain mapping systems have been developed in Canada, beginning with Fulton et al (1974), and in other countries, for example Finlayson (1984, Australia).

In this province, the BC Terrain Classification System (Howes and Kenk 1996; Resources Inventory Committee 1996a) is the provincial standard and is a versatile system to produce terrain maps from a scale of 1:10,000 to 1:250,000. An early data base format for this system was introduced by Kenk et al (1987). Resources Inventory Committee (1996b) summarizes the most recent data base format.

Medium scale terrain maps (1:20,000 to 1:50,000) can be used as preliminary terrain stability maps with suitable annotation of landslide hazards for each type of unit. Ryder and MacLean (1980), Howes and Swanston (1994) and Resources Inventory Committee (1996a) provide examples for use with the BC Terrain Classification System.

Type 3 -- Engineering geology maps delineate, interpret and annotate surficial or bedrock units or terrain units to provide information relevant to engineering issues, such as material usability, soil plasticity, foundation conditions, groundwater conditions, swelling potential, and/or landslides.

Engineering geology maps are produced at a variety of scales, from medium (1:20,000 to 1:50,000) to detailed (1:5,000 or larger). Methods for engineering geology mapping including symbols to represent landslide areas have been summarized by the International Association of Engineering Geology (1976, 1981a and 1981b).

Type 4 -- Terrain attribute maps delineate the distribution of one or more specific terrain attributes, such as overburden depth, soil type or soil moisture. For terrain stability mapping, two terrain attributes, slope gradient and drainage network, are most useful and discussed further below. Using a GIS, individual terrain attribute maps can be combined as a series of overlays to produce a multi-terrain attribute map.

Slope maps show the distribution of slope gradients. These maps can be prepared at scales as small as 1:50,000 but larger scales are preferable. Simple slope maps may have only two slope categories, for instance areas <50% and areas >50%. More complex slope maps may have several slope classes that are determined to be significant. Slope maps can be used to prioritize areas for more detailed mapping, especially if background data and/or funding are limited.

Traditionally slope maps are produced by measuring the distance between topographic contour lines. Recently Digital Terrain Models (DTM) and Digital Elevation Models (DEM) have become available. To produce satisfactory results, however, DTMs and DEMs must be sufficiently detailed. It is important to note that photogrammetrically derived contours on heavily forested slopes often underestimate local slope angles.

Drainage network maps show permanent and ephemeral drainage paths and can be produced at a variety of scales. A potential use in terrain stability mapping, besides the depiction of drainage density, is to indicate potential paths of debris or sediment movement, and potential locations of erosion. Drainage network maps have traditionally been produced manually, and may require considerable field checking if produced at a large scale especially in a heavily forested area. Drainage network maps produced from DTMs require a very high quality terrain model.

Type 5 -- Process inventory maps delineate the distribution of one or more geomorphic processes, such as snow avalanching, erosion and landsliding. Such processes can be shown by polygons, feature outlines, linear symbols and/or point symbols. Landslide inventory maps are produced for a variety of purposes including:

• to delineate different types or sizes of landslides;
• to distinguish active, or recently active, landslides from those which are dormant;
• to document landslide damage incurred in a region from a specific event, such as a rain storm or earthquake;
• to guide research or mitigation spending within a region; and/or
• to calibrate and provide detail to other types of terrain stability maps.

Landslide inventory maps, at scales larger than 1:5,000, often use elaborate systems of feature outline symbols (see Section 3.5) to indicate internal detail of large landslides, such as scarps, ridges, and tension cracks.

Landslide density maps, an extension of landslide inventory maps, use contours to join areas with equal densities of landslides (isopleths), and are useful in areas that contain a relatively large number of relatively small landslides. See for example DeGraff (1985).

Some countries are attempting to establish a national landslide inventory mapping program accompanied by a data base. A world-wide standard for the reporting and inventorying major landslides is being prepared by the UNESCO Working Party on World Landslide Inventory (1990a, 1990 b, 1991, 1993a, 1993b, 1995 and in prep).

Other examples of landslides inventory maps include Kienholz (1978, a single valley in Switzerland), VanDine and Evans (1992, a regional study of Vancouver Island), and Radbruch-Hall (1982, the entire United States).

Type 6 -- Landslide hazard maps can be derived by interpreting one or more of the map types described above. They are often produced, however, specifically for use as landslide hazard maps and have a wide range of forms and present a wide range of information.

Landslide hazard maps are most commonly directed toward landslide initiation zones and are sometimes referred to as 'landslide initiation maps'. They often delineate areas of equal probability of landslide initiation, such as the probability of occurrence, or the probability of occurrence combined with magnitude and/or some other characteristics of the landslide. In runout zones, landslide hazard maps delineate the probability of certain areas being affected by the runout of landslide debris. Some landslide hazard maps consider both the initiation and runout zones. The probability in both the initiation and runout zone can be expressed either qualitatively or quantitatively (refer to Section 2.3).

Type 7 -- Landslide risk maps extend the information shown on landslide hazard maps to include consequence associated with the hazard. They are usually prepared for the runout zones and are sometimes referred to as 'landslide runout maps'. Occasionally landslide risk maps are prepared for the initiation zone.

Risk may be expressed either in qualitative or quantitative terms. In some cases spatial distribution functions are used to determine the variability of landslide risks over a specific geographic area. Landslide hazard maps that are annotated with simple consequence ratings can be considered as simplified qualitative landslide risk maps.

3.3 Terrain Attributes of Landslide Hazards

Landslides are exceedingly complex phenomena that are controlled by, or associated with, many physical factors, or terrain attributes. Hutchinson (1992) summarized the main 59 terrain attributes, both preparatory and triggering, into 13 categories. His summary is reproduced as Table 3.1. Popescu (1994) formulated a similar list.

The ideal terrain stability map would record information on all the terrain attributes in Table 3.1. The resulting map, however, would of course be unrealistically complex. Furthermore, not all terrain attributes are important in all circumstances. Therefore, any practical terrain stability mapping system must select a relatively small group of relevant terrain attributes.

The selection of such attributes can be based on individual judgment, or on an analysis of actual landslides. This study examined twelve different terrain stability mapping projects from different parts of the world to determine which terrain attributes were selected and why. In eight of the twelve projects the terrain attributes were selected subjectively. In the other four, statistical analyses were used to select statistically 'significant' terrain attributes. The mapping projects examined are listed in Table 3.2.

The projects examined show that the only terrain attribute common to all projects is slope gradient. Evidence and frequency of past instability were used in seven of the twelve projects. Besides slope gradient and past instability, a wide variety of terrain attributes were selected by the mappers as being relevant. The revelance of certain terrain attributes was found to be very different, and sometimes contradictory. To some extent this was dependent on the local or regional terrain and geology. For example, bedrock lithology tends to be important in regions of weak argillaceous rocks, but insignificant in areas of stronger rocks, or areas covered by glacially derived soils. It was noted that different groups of terrain attributes were sometimes responsible for different types of landslides in the same geographic area. Rollerson (1992), for example, determined that a different group of relevant terrain attributes were responsible for landslides triggered by road building, than for landslides associated with clear cut logging.

3.4 Map Scales

The scale of presentation of the terrain stability map is very important to communicate the appropriate level of detail for the intended use. The presentation scale should be dependent upon the actual scale of mapping and the methods and intensity of field checking -- sometimes referred to as the terrain survey intensity level (TSIL) (discussed further in Chapter 5).

The following classification of scales of map presentation is modified from Van Westen (1993).

Synoptic or territorial scale maps (>1:50,000) are often process inventory maps, used by planning agencies to direct allocation of funds, develop emergency preparedness plans and similar tasks. An example is the overview map of the United States produced at a scale of 1:7,500,000 (Rudbruch-Hall 1982).

Medium scale maps (1:20,000 to 1:50,000) are generally used for preliminary or regional landslide hazard assessments and feasibility studies, to be followed by more detailed work. The 1:50,000 scale terrain maps using the BC Terrain Classification System are examples.

Large scale maps (1:5,000 to 1:20,000) are generally used for planning of land use in urban areas or resource development in rural areas. Depending on the use, large scale maps quite often must be supplemented by detailed site investigations or on site assessments. In British Columbia, detailed terrain stability maps prepared for forest management are prepared at 1:20,000 scale (BC Ministry of Forests, 1995a)

Detailed scale maps (1:5,000 to 1:500) are usually prepared as part of a landslide hazard assessment of a specific site and should be accurate enough to guide layout of individual structures or specific operations, or to plan mitigation. Engineering plans at these scales can also be used to derive design parameters. In the BC forest industry, areas with moderate or high probabilities of landslide hazard, identified from the 1:20,000 scale mapping, are examined on the ground at the 1:5,000 scale . The findings from these field assessments are used to help locate cutting boundaries and roads (BC Ministry of Forests, 1995a).

3.5 Map Units

Map units on terrain stability maps delineate the terrain attributes, such as slope, soil drainage, and material texture, and/or the hazard and risk parameters, such as probability of occurrence, magnitude and specific risk. They can be referenced to the map in a number of different ways.

A regular rectangular map grid is sometimes used to approximate a continuous variation of terrain attributes or parameters over an area defined by a regular grid. This is similar to the mathematical process of making a continuous function discrete by assuming that the function is constant within each elementary area.

A regular grid of points can be used to display the average terrain attributes or parameters at a point, usually defined on the basis of a regular grid.

Polygons are used to delineate areas which are approximately uniform in terms of one or more terrain attributes or parameters. Polygons can be delineated subjectively or objectively by using a map overlay process. The BC Terrain Classification System (Howes and Kenk 1996; Resources Inventory Committee 1996a) uses subjectively delineated polygons based on material genesis and texture, surface expression, and geomorphic process. The polygon system provides greater flexibility and scope for use of geological knowledge than the above grid systems.

Linear segments are used to map linear geomorphic features such as shorelines, stream channels or gullies.

Contours of a terrain attribute are used to display the variation of a specific terrain attribute or parameter. For example, slope maps are often presented as polygons with a range of specific slope gradients.

Feature outline symbols are used to display detailed features of landslides such as tension cracks, slump blocks, and headscarps, and are therefore usually limited to large landslides and/or detailed scale maps.

Linear and point symbols are generally used to delineate small features. The definition of a small feature is determined by the map scale.

Table 3.1 Main Terrain Attributes Associated with Landslides

Table 3.2 References Examined for Selection of Terrain Attributes

Terrain attributes selected subjectively

Terrain attributes selected by statistical analysis