Digital elevation models
Digital Elevation Models, also referred to as Digital Terrain Models (DTM) or Digital Surface Models (DSM) record a raster representation of ground surface elevation; almost invariably the raster cell is square. Unlike contours, DEMs deliver an elevation measurement for every cell in the raster, so there are no intermediate points for which interpolation is necessary. DEMs are used for a wide variety of purposes, including studying visibility and intervisibility, exposure, hydrology, ease of access and, using high resolution data, minor surface variations that may indicate the presence of buried features. This section first discusses some readily available DEM data for Great Britain and then sources for other areas. It then discusses some of the potential uses in an achaeological context.
DEMs for Great Britain
The primary provider of spatial data for Great Britain is the Ordnance Survey of Great Britain (OSGB). At the time of writing, OSGB offer two DEM products, both of which are available to users whose institutions are registered for the JISC Digimap service:
- Landform Panorama. This product is derived from 1:50000 contour mapping to generate a DEM with a 50m cell size and an integer (whole number) elevation value. This product gives a generalised surface model but which is unsuited for hydrological studies. It is good for general visibility and landscape studies.
- Landfrom Profile. This product includes material derived from 1:10000 contour mapping and spot height control for interpolating between contours and has a cell size of 10m and integer elevation value. It is better than Landform Panorama for hydrological studies but will not necessarily yield a hydrologically correct model. It is suitable for similar studies to those using Landform Panaorama but reqquiring greater detail.Users of DEM products derived from contours should be aware of significant potential artefacts that are directly related to the contour source. These are fully described in Jo Wood’s PhD thesis (1996).
There are a number of DEMs for Great Britain which are derived from Remote Sensing methodologies. These include:
- Shuttle RaDAR Topography Mission (SRTM). These data were aquired by NASA using a Synthetic Aperture Radar (SAR) instrument mounted on the Space Shuttle. Details of the coverage and product options are discussed in connection with other geographic areas, below, however 3ArcSecond SRTM data for the British Isles (including Ireland) are available from the JISC Landmap service for academic and other users permitted access to this service. These SRTM data have a cell size of 75m and a vertical resolution of 10m. SRTM data are suitable for larger scale landscape studies and for visibility, intravisibility and exposure studies; they are not suitable for hydrological studies. Whilst generally of good quality, SAR data suffer from a problem termed ‘loss of phase coherence’, where the radar signal cannot be effectively processed and which can lead to ‘holes’ in these data. Such ‘holes’ are rare and are most common over water; they can also occur where the direction of the radar beam is the same as the slope of the ground. The data available for download from the Landmap service are projected onto the British and Irish National Grids respectively.
- Landmap ifSAR DEM. This product, with a 25m cell size is derived from European Space Agency European Radar Satellite (ERS) data using a process with similarities to the photgrammetric process used with stereoscopic aerial photography called interferometry. This product gives a surface representation that include man-made objects, such as the upper surface of embankments, bridges, etc. These data are more detailed than the SRTM materials and generally have a better vertical resolution. There are similar processing artefacts as the source instruments are similar. Whilst unsuited to use in hydrological applications, these are good data for landscale level work and for visibility, inter-visibility and exposure studies. These data are projected to the British and Irish National Grids respectively.
- Bluesky DTM. This commercial DTM has a cell size of 5m and an integer elevation value which has been rounded to the nearest metre. These data are photogrammetrically interpolated from stereoscopic aerial photography and adjusted to record the ground surface elevation: a separate, Building Heights, dataset is available which records the difference between the measured and surface elevations. At the time of writing, these data are available for England and Wales from the JISC Landmap service for academic and those users eligible for access to this facility. These data, registered to the British National Grid, are suitable for more detailed landscape, visibility, inter-visibility and exposure studies; they are also of a suitable resolution for larger scale hydrological studies, generalised flooding models, etc. The cell size means that some larger archaeological features may be discernable from these data.
- LiDAR (Light Direction and Ranging). These highly detailed products are generated from airborne laser instruments and can deliver ground cell sizes down to 25cm and a sub-centimetre vertical resolution. Expensive to collect and store, these very high resolution data are primarily collected in the UK for flood risk modelling and monitoring and so tend to be collected for ‘at risk’ urban environments. These data are capable of yielding highly accurate hydrological models and are capable of supporting study of buried structures that result in variations in ground surface elevation, such as archaeological remains. UK data are collected by the Environment Agency and the Geoinformation Group: data from the GeoInformation Group product range are available through the JISC Landmap service for eligible users. These Landmap service datahave a cell size of 1m, a vertical resolution of 15cm and coverage at the time of writing for Birmingham, Edinburgh, Glasgow, Liverpool, London, Manchester and Newcastle-upon-Tyne.
A note of caution
With raster data, such as these, an increase in horizontal detail is accompanied by a significant increase in the volume of the data. For example, going from a cell size of 30m to 15m will quadruple the data volume. These data are, therefore, very large: the Bluesky DEM for England and Wales, as delivered, occupies around 50Gb of disk storage. Some software packages and/or computer operating systems may impose limits on the maximum dataset size that can be manipulated or stored. It is beyond the scope of this guide to provide details or advice on dealing with issues relating to dataset size.
Rest of the world
This list is in no way complete: there are data products for nations and regions which cannot be listed here. Listed here are some easily accessible datasets with near global coverage and some pointers for looking for more detailed products.
- SRTM. The Shuttle RaDAR Topography Mission covers landmasses between 60 degrees North and 56 degrees South. The mission was flown in February 2000 as a joint international venture. The most detailed SRTM products are restricted to military users; 1ArcSecond, approximately 30m cell size, data are available within the USA to USA users for the co-terminus USA. 3ArcSecond data, which varies in cell size depending on the location of the study area on the surface of the earth, 90m on the Equator, is available for download for any location within the data collection limits; there is also a 30ArcSecond, or approximately 1km, product and a range of related imagery products. The vertical resolution of published SRTM data are 10m. These data are available for download from USGS and a number of other servers.
- ASTER GDEM. The product of a joint Japanese Ministry of Environment, Trade and Industry and NASA initiative, these data are produced from measurements collected by a Japanese instrument on board the ‘terra’ satellite. Data have been collected since 2000 for landmasses between 83 degrees North and South. These data have a cell size of 30m and a vertical resolution between 7 and 14m. The ASTER instrument operates in the visible light spectrum, so no data can be collected from areas of the Earth subject to continuous cloud cover. ASTER data are supplied in 1 degree tiles and are available for download from ASTER GDEM.
- SPOT DEM. The French SPOT Earth observation satellite series are capable of collecting stereoscopic imagery, which can be processed photogrammetically, like stereoscopic aerial photography, to generate a DEM. These data can deliver a planimetric accuracy of 15m and a vertical accuracy of 10m and are supplied as either 1ArcSecond or 20m datasets. This is a commercial service delivered by SPOT Image and details can be found on the SPOT Image website.
- National services. There are national and commercial services supplying DEM data in many countries and beyond the scope of this guide to deliver a detailed list. LiDAR data, as well as photogrammetic DEMs are increasingly available and local sources must be consulted for details.
Uses of DEM data
Surface topography has a direct impact on many natural and human processes, in addition to providing evidence of past or present natural or human activities. In high latitudes, the angle and direction of slope can be the controling factor in snow melt; shallow slopes and the similarity of surface elevation to river level are factors in assessing flood risk; and the elevation of one location relative to another is a major factor in controlling whether the one location is visible from the other.
Working with DEM data
The nature of DEM data is such that it is necessary to employ a GIS or Remote Sensing package in order to use these data. Although similar to image data, DEM data have values from the lowest elevation, which may be below sea level and thus negative, to the highest and, as a result, are not suited to processing with general purpose image processing packages, such as Photoshop or the GIMP. Most GIS packages, including Open Source products, include at least some basic tools for computing hillshade and visibility, direction and angle of slope and to be able to associate specified locations with the elevation at those points. In combination, these basic tools enable a wide range of study opportunities. In archaeological terms, such tools support an interpretive analysis: they are not deterministic in themselves.
Visibility and exposure
These approaches concern whether one location is visible from another (visibility), the extent to which an area is visible from elsewhere and the extent to which an area is exposed to the sun, wind directions, etc (exposure). These can be measured using similar tools. Most GIS packages offer a hillshading tool: this is designed to model light shone from a light source and illuminate those parts that are not in shadow. By adjusting the location and elevation of the light source, it is possible to determine the field of view, or viewshed, from a given location and, thus, the area exposed to that location. In a similar manner the light source can be programmed to follow the passage of the sun and to identify areas illuminated or in shadow, or their exposure: obviously in an extreme climate, exposure to the sun or shade that prevents the sun penetrating to a location may have a significant impact on the behaviour of natural and human phenomena.
Some packages have tools to support the determination of the visibility of individual and specific locations from another. An example might be a chalk-covered barrow or for a more defensive purpose. This is the type of work referred to here as inter-visibility. Where such tools are not included, a similar effect can be achieved by reversing the viewshed analysis, to determine whether the specified location is in view from the target.
The surface morphology of a landscape can also have a controlling impact on natural and human phenomena: flood risk is one example, as might also be the availability of suitable land for some activity. These are usually measured in terms of slope, the angle of slope, and aspect, the direction in which the ground slopes. Areas to which the ground slopes from all adjacent cells may, for example, be damp hollows; the combination of slope, aspect and elevation may determine the practicality of some activity, such as the growing of some crop. There are many tools for studying landscape morphology, but these are outside the scope of this guide.