Abstract
The importance of terrain evaluation techniques in engineering geology is demonstrated by the instigation of two Engineering Group working parties on the subject that reported in 1982 and 2001. Since the second working party report ground models have emerged as one of the key requirements of ground investigations for engineering practice and this paper suggests that terrain evaluation is fundamental to the development of ground models in providing the spatial framework to facilitate interpretation. However, terrain evaluation has also emerged as the basis for creating more extensive generalized models that are required when constructing long linear structures such as roads and pipelines that cross very varied landscapes. This variation in application has led to confusion over the nomenclature used in terrain evaluation. In this paper terrain evaluation is regarded as a broad over-arching term that encompasses a range of techniques and it is suggested that its application in engineering practice should not be restricted by narrow or ambiguous definitions.
In 1982 a Working Party of the Engineering Group of the Geological Society published their report on ‘Land Surface Evaluation for Engineering Practice’ (Anon. 1982). The 1982 report stated that ‘Land surface evaluation should be regarded and used as a normal method of investigation for ground engineering’ (Anon. 1982, p. 266). The report went on to recommend the use of geomorphological ground mapping and remote sensing, especially aerial photograph interpretation, based on a framework of land classification as the main methods of use for land surface evaluation. As discussed in a Second Working Party Report on the topic (Griffiths 2001) ‘the expression “land surface evaluation” was used in preference to “terrain evaluation” or “terrain classification” because the varied uses of the terms had created confusion and led to misunderstandings’ (Griffiths & Edwards 2001, p. 3). However, in the context of engineering geology the terms land surface evaluation and terrain evaluation are synonymous, although in the present publication the term terrain evaluation is preferred. The philosophy of the 1982 Working Party Report was endorsed by the 2001 report and terrain/land surface evaluation was defined as ‘the evaluation of the land surface and near surface features using techniques that do not involve ground exploration by excavation or geophysics’ (Griffiths & Edwards 2001, p. 3). The range of techniques that fell under the heading of terrain/land surface evaluation had expanded considerably since 1982 (Brunsden & Griffiths 2001) to include digital photogrammetry, stereo-satellite imagery, multispectral aerial survey, geographic information systems, synthetic aperture radar, airborne laser scanning (LiDAR), digital elevation models (DEMs), multibeam swathe bathymetry in offshore studies, engineering classification of environmental performance, natural hazard assessment and ground behaviour mapping. Since 2001 there have been further developments in techniques but more importantly there are more examples of the application of terrain evaluation on engineering projects. However, the major change in ground investigation since 2001 has been the recognition that a synoptic appreciation of ground conditions at a site can be provided by the development of a ground model (Fookes 1997; Parry et al. 2014; BSI 2015). In the present paper the development of terrain evaluation in engineering is discussed, including the advances that have occurred since the publication of the Second Working Party Report, and an attempt is made to provide clarification of the sometimes confused nomenclature associated with the subject. In addition, a clear link between ground models and terrain evaluation is established that requires changing the 2001 definition of terrain evaluation so that it does not preclude the use of ground exploration by excavation or geophysics.
Development of terrain evaluation
Terrain evaluation for engineering utilizes a suite of techniques to assess ground conditions and the techniques help organize and communicate earth science information (Fookes & Lee 2005), which is particularly helpful when addressing client requirements. The techniques that can be utilized and fall under the broad term ‘terrain evaluation’ include geomorphological mapping (Smith et al. 2011), aspects of geological mapping (Barnes & Lisle 2003), engineering geological mapping (Dearman 1991; Griffiths 2002), remote sensing analysis and interpretation (Teeuw 2007), terrain classification and landsystems mapping (Lawrance et al. 1993), and the creation of digital terrain models and airborne laser scanning (Martin et al. 2011).
A number of publications have reviewed the history of development of terrain evaluation techniques: Webster & Beckett (1970) showed how an approach arose out of the work by the Oxford–MEXE group for military purposes (Brink et al. 1966), Lawrance et al. (1993) provided a range of examples on its use for highway projects, Cooke & Doornkamp (1990) demonstrated how it is a core element in environmental management, Rowan (1977) referred to the use of terrain classification by the Forestry Commission, and Dearman (1991) established that it is a component of engineering geological mapping. The best early examples of the use of terrain evaluation techniques for engineering are from the work by TRRL (Transport and Road Research Laboratory) in establishing highway alignments in the humid tropical conditions of West Malaysia (Lawrance 1972). However, many of the basic methods of terrain evaluation had been developed much earlier in Australia by CSIRO under the title Terrain Classification as a means of compiling basic natural resource data on larger areas where logistics made it impossible to undertake the work using conventional field mapping techniques (Christian & Stewart 1952).
In the UK, following the publication of the First Working Report (Anon. 1982) case studies appeared in the literature that showed that the techniques were being used in engineering practice. One of the most innovative was by Waller & Phipps (1996), on relict periglacial conditions along the alignment of the Channel Tunnel Link in Kent. Further case studies were provided in the Second Working Party Report, including such subjects as pipeline route selection in Algeria (Fookes et al. 2001), terrain hazard assessment around the Ok Tedi copper mine in Papua New Guinea (Hearn et al. 2001), ground conditions for Suez City development (Fig. 1; Jones 2001), land use planning in unstable areas in Ventnor, Isle of Wight (Lee & Moore 2001), and ground subsidence mapping in Cheshire (Lee & Sakalas 2001). However, the most comprehensive use of terrain evaluation techniques for infrastructure planning and development was undertaken in Hong Kong by the Geotechnical Control Office (GCO), where a systematic geomorphological, geological and geotechnical mapping programme of the whole territory was carried out (Brand et al. 1982; Koirala et al. 1986). Although this study was subsequently superseded by more extensive ground investigation studies, the concept of terrain evaluation has continued to form the basis for natural hazard assessment (Parry & Ruse 2002; Ng et al. 2003), in particular for the investigation of landslides (Parry et al. 2002; Parry & Ng 2010; Parry 2011; Styles & Sin 2014).
Geomorphological map of the Suez area produced by aerial photograph interpretation and ground mapping - note the black line identifies the planned limits of the new city development (Jones 2001; reproduced with permission of the Geological Society of London).
The important applications in terrain evaluation since the Second Working Party Report
In the period following the publication of the Second Working Party Report three areas of engineering have placed a strong emphasis on the value and effectiveness of using of terrain evaluation techniques: onshore oil and gas pipelines, low-cost roads in mountainous roads and planning of seabed oil and gas developments. These engineering situations involve the design and construction of long linear structures, often in remote or difficult to access areas including submarine sites, where there are few pre-existing data on the ground conditions and therefore extensive use has to be made of remote sensing data (Manning 2007).
Fookes et al. (2001) had already demonstrated the importance of the terrain evaluation for pipelines in deserts but the full scope of its application for all onshore pipelines became apparent in the conference book Terrain and Geohazard Challenges Facing Onshore Oil and Gas Pipelines (Sweeney 2005). In this collection of papers Charman et al. (2005) explained how ground conditions and terrain affect the alignment and construction of oil and gas pipelines, thus making the use of terrain evaluation as part of site investigations very cost-effective. Amongst a number of case studies two papers on the In Salah Gas project in Algeria most effectively demonstrate the value of terrain evaluation: Fookes et al. (2005) discussed the technique in relation to pipeline alignment, and Shilston et al. (2005) went to explain how it is used for actual site design. Subsequently, Sweeney (2017) in the 15th Glossop Lecture specifically identified terrain evaluation as working well in oil pipeline investigations. The approach employed at In Salah has been used for other desert construction projects, and Shilston et al. (2012) have discussed the critical role of desk studies, remote sensing, GIS and field evaluation of the terrain in relation to all aspects of engineering geology in hot deserts.
The way terrain evaluation has been used as part of site investigations for low-cost mountain roads has been summarized by Hearn (2011a), in a volume that built on the work originally presented in Overseas Road Note 16 (TRL 1997). Hearn (2011b) highlighted how desk studies utilizing all forms of remote sensing data and compiled in a GIS can produce an initial evaluation of the terrain that can summarize ground conditions for construction, produce 3D models and maps of natural hazard susceptibility (Fig. 2), cut–fill balance, and assess the drainage requirements. Hearn (2011c) went on to demonstrate how the initial desk studies need to be supplemented by field mapping at larger scales to produce information that can be fed directly into the design process.
Terrain classification for route selection in Nepal (Hearn 2011b; reproduced with permission of the Geological Society of London).
According to M. Lee (pers. comm.) one of the most sophisticated and comprehensive applications of terrain evaluation since the 2001 Working Party Report has been for the development of seabed oil and gas resources, notably in the West Nile Delta, the Caspian Sea and offshore of Angola. Much of this work remains confidential but the papers by Evans et al. (2007), Hough et al. (2010) and Power et al. (2011) illustrate the scope and nature of this work. It is the refinement of multibeam swathe bathymetry systems that has been one of the most important technical developments in facilitating this application. Data retrieved using this system provide the basis for high-resolution DEMs of the seafloor that are suitable for geomorphological mapping and interpretation (Dunlop et al. 2011) and when combined with geophysical seismic reflection information allow 3D models of the geology down to kilometre depths to be created (Micallef 2011).
Apart from the above relatively new areas of application, terrain evaluation has continued to be part of engineering geological practice. For example, in the Engineering Geology Working Party book on hot deserts (Walker 2012), landscape classification, hazards and the desert ground model have been brought together by Griffiths & Stokes (2012). Building on this philosophy, a complete integration of terrain evaluation, engineering geology and geomorphological landsystems for engineering practice has been provided by Griffiths & Martin (2017), in the Working Party Report on the engineering geology and geomorphology of relict glacial and periglacial terrain. Work also has continued in Hong Kong in utilizing terrain evaluation as part of the continuing investigation of landslide hazard (Anon. 2014).
The above examples are not a comprehensive review of the literature on the subject that has been published since the Second Working Party Report. Nevertheless, these examples demonstrate that the original aim of the First Working Party on Land Surface Evaluation for Engineering Practice (Anon. 1982) of making the techniques a normal method of ground investigation has gone a long way to being fulfilled.
Terrain evaluation, terrain classification and landsystems
Lawrance et al. (1993) described terrain evaluation as a method for summarizing the physical aspects of a landscape through classification, but this is more correctly called terrain classification, although it is also referred to as landsystems mapping. Terrain classification is more specific than terrain evaluation and, as noted above, is a technique developed by CSIRO that was used to classify large areas, at relatively small scales, to assist in the identification of potential agricultural and economic resources in Australia and Papua New Guinea. Terrain classification subdivides the landscape into the following elements.
Landsystems. These are large areas (e.g. 100 km2) with a recurring pattern of landforms, soils, vegetation, geology and hydrological regimes (Mitchell 1991; Lawrance et al. 1993). Normally landsystems are mapped at small scales (e.g. >1:100 000), and Table 1 provides an example of landsystems classification compiled for a proposed road alignment in Papua New Guinea.
Land facets. These are terrain units of uniform slope, parent material, soils and hydrological conditions. Lawrance et al. (1993, p. 57) described a land facet as ‘sufficiently homogeneous to be considered uniform for most practical purposes’. Land facets combine to make up a landsystem and can be mapped at scales ranging between 1:10 000 and 1:100 000. Land facets can often be given simple morphographic descriptors such as dune field, glaciated valley or tombolo. Features of this scale can be shown on geomorphological maps specifically designed for engineering (Brunsden et al. 1975) and a range of morphological, geological and geotechnical data can be identified for each facet.
Land elements. These are the smallest unit of the classification and combine to make a land facet. These could include simple morphological descriptors such as concave slope, cliff and convex slope. Land elements can be mapped in the field using a system known as morphological mapping (Savigear 1965) that underpins all detailed geomorphological mapping for engineering (Griffiths & Marsh 1986).
Example of a landsystems classification for a proposed highway in Papua New Guinea based on aerial photograph interpretation (Griffiths 1994)
Terrain classification should be seen as a key component of terrain evaluation for engineering purposes (Shilston et al. 2012). As demonstrated by Waller & Phipps (1996), in geomorphological studies landsystems can be developed at scales that have direct engineering geological relevance as they link the geomorphological response of a landform to the processes of formation. Indeed, Brunsden (2002) proposed that, in the context of the geomorphological applications to engineering, the Earth's surface comprises ‘process–response [landform] systems that are specified by the stress fabrics given by tectonic history and setting, the properties of the rocks, superficial deposits, landform geometries, climates and biomes arranged in patterns across the Earth's surface in a hierarchy of landscape types and scales’ (Brunsden 2002, p. 103). It is this definition that establishes the link between geomorphological landsystems and the creation of ground models that is now a fundamental part of engineering geological practice (Parry et al. 2014). This is exemplified by the example of a process–response model that Brunsden (2002) presented of a sea cliff that highlights both the geomorphological systems and the geotechnical factors that control the stability of the cliff face (Fig. 3). If all the factors shown in Figure 3 could be quantified perhaps the behaviour of the cliff face over time could be accurately forecast, although it is also true that geotechnical investigation has not yet reached the stage where this deterministic approach to modelling is possible. Indeed, with respect to coastal cliff recession, Lee (2008) argued that it could be a chaotic process, therefore deterministic modelling may never be possible and probabilistic methods may be the only way to proceed. Nevertheless, if the ground model presented in Figure 3 was site specific it would fulfil the requirements for an observational ground model as defined by Parry et al. (2014).
The sea-cliff geomorphological process–response system (modified from Brunsden 2002; reproduced with permission of the Geological Society of London). MHWL, mean high water level; MLWL, mean low water level.
Terrain evaluation output
The output from a terrain evaluation should be a suite of maps and block diagrams that in the past were produced as hardcopy but now should be held as a series of overlays in a geographic information system (GIS) (Turner 2003). The GIS should contain all the data collected in desk studies that are needed for an engineering geological assessment, and all these data must be georeferenced. Although the end-product produced for a client must be project specific, it can be helpful for document control if the map or GIS overlays are classified into three categories, as follows.
Element ‘maps’ that record the actual measurable and/or mapped ground conditions (i.e. the factual basis, such as topography, bedrock geology, soils, morphology, land use, and location of previous and planned ground investigations).
Derivative ‘maps’ and ‘diagrams’ obtained by either interpreting or combining various element maps and, where available, incorporating ground investigation data (e.g. landforms based on an interpretation of the morphology; slope steepness derived from the topographic data; depth to bedrock compiled by combining ground investigation data, superficial deposit and bedrock maps; and block diagrams showing the landscape and subsurface geology in three dimensions).
Summary ‘maps’ and ‘diagrams’ that integrate a range of derivative and element maps to identify combinations of hazards, resources or land use issues that act either as constraints to any development or indicate the potential of the land for exploitation and development.
An extended legend (as exemplified by Table 1) should be attached to each map and diagram, and linked to GIS layers. The terrain evaluation must include an interpretative report that explains the basis for developing the derivative and summary maps. It must be emphasized that the output of most value to the end-users (whether they are planners, developers, accountants, engineers or lawyers) is the derivative and summary maps, and these must meet their requirements (e.g. why build here or what are the ground problems in this area?). However, the author's experience as an expert witness is that if anything goes wrong and construction claims or litigation are initiated then the background to the creation of derivative and summary maps and how these guided design decisions can become critical. Therefore, it is recommended that even if background data are not immediately required by an end-user these data are compiled and archived in a form that provides a justification for any design decisions based on the work.
Once a terrain evaluation has been carried out it should not be seen as a task completed but should be part of any continuing development on a project and must be regarded as a live document that can be revised and updated as more data become available. This is most easily carried out if the maps and accompanying data are compiled within a GIS framework. As ground investigation and excavation data become available they should be added to the database and the interpretation of ground conditions updated, but careful quality control on the documentation is needed and it is important to keep an accurate record of why revisions were made and what new data were used to update the work.
Nomenclature
It is apparent from the above discussion that the terminology employed in the overall subject of terrain evaluation for engineering practice can be confusing. Therefore, an attempt needs to be made to provide an agreed nomenclature that brings together the various facets of terrain evaluation to ensure all parties know what is expected when a project requiring its use is commissioned.
Terrain evaluation is the overall term to describe the suite of techniques employed in an assessment of ground conditions. This includes all the field and remote methods of evaluating the land surface plus, when available, the results from exploratory holes and geophysics. All terrain evaluations must be updated when additional data become available. The scale of the terrain evaluation will vary with the project but for large-scale detailed engineering design (≥1:1250) terrain evaluation will provide the spatial framework for designing and interpreting ground investigations and form the basis for creating site-specific observational ground models. At smaller scales (≤1:10 000) terrain evaluation will be used for general planning, decisions on the best alignment for linear structures, identification of geohazards and the production of ground models that will be more conceptual in nature. However, in practice, terrain evaluation is a continuum and there is no boundary between these two approaches, and the methods employed and output produced will be a function of the project requirements.
Based on its history of development terrain classification is more specific than terrain evaluation and is primarily a reconnaissance technique based on remote sensing interpretation linked to limited ground truth checking. Terrain classification will be used as a method of terrain evaluation and initially divides a landscape into mega-scale landsystems that comprise a number of macro-scale land facets, which in turn are made up of meso-scale land elements (Table 2).
Definition of mega-, macro- and meso-scales in geomorphology (Griffiths & Stokes 2008)
Using the terminology proposed by Parry et al. (2014) for describing ground models in engineering, landsystems can be both conceptual and observational. In Figure 4 a conceptual landsystems model of a glacial margin is provided based on an understanding of glacial geomorphology (Evans 2017). It should be noted that this conceptual model shows geomorphological processes and landforms and hence can be regarded as a process–response model albeit at a smaller scale than the sea-cliff model shown in Figure 2. The landsystems model is not location specific but identifies the type of features that might be encountered in such an environment. Terrain classification, therefore, can be regarded as an early phase of terrain evaluation that can produce conceptual ground models although often at small scales. Subsequent detailed engineering geological and geomorphological mapping based on the terrain classification linked to ground investigations will result in a more detailed understanding of the nature of the landscape that can facilitate the production of ground models designed to meet the project requirements.
Active glacial margin landsystem (Evans 2017; reproduced with permission of the Geological Society of London).
Conclusions
The utilization of terrain evaluation in engineering practice should not be restricted by issues around narrow or sometimes ambiguous definitions; therefore terrain evaluation should be seen as an all-encompassing term that utilizes a suite of techniques to collect the data about the surface and near-surface landscape for engineering purposes. Under this definition, unlike in the 1982 and 2001 Working Party Reports, terrain evaluation should be taken to include using results from ground investigations and can be used to create the spatial framework for designing and interpretation of ground investigations at any scale; as the basis for developing site-specific observational ground models; in natural hazard analysis; for assessing alignments of long linear structures such as roads, railways or pipelines; and for interpreting subsea conditions.
Fookes (1997, p. 380) stated that ‘the maximum value of the geological model is in the earlier stages of the study as the benefits it would bring are enhanced quality of design and improved data from the main ground investigations’. On this basis the early application of terrain evaluation techniques will provide substantial benefits to a project and when used in conjunction with ground investigation will provide the basis for safe and economical design.
Acknowledgements and Funding
Although the views expressed in this paper are the author's own he wishes to acknowledge the valuable input from M. Lee in his review of the draft paper and to thank him for the innumerable helpful discussions held over the past 30 years.
- © 2017 The Author(s)
