Citation: Hirt, C (2015), Digital Terrain Models. In: Encyclopedia of Geodesy (Ed. E.W.

Grafarend),
doi 10.1007/978‐3‐319‐02370‐0_31‐1, Springer, Berlin, Heidelberg.

Digital Terrain Models

Christian Hirt 1,2
1
Department of Spatial Sciences, Curtin University, Perth, WA, Australia
2
Institute for Advanced Study, TU Munich, Germany

Definition
Digital Terrain Model. Digital description of the terrain surface using a set of heights over 2D
points residing on a reference surface
Introduction
A Digital Terrain Model (DTM) approximates a part or the whole of the continuous terrain
surface by a set of discrete points with unique height values over 2D points. Heights are – in
approximation ‐ vertical distances between terrain points and some reference surface (e.g.,
mean sea level, geoid and ellipsoid) or geodetic datum. Mostly arranged in terms of regular
grids, the 2D points are typically given as geodetic coordinates (latitude and longitude), or
planar coordinates (North and East values). DTMs usually assign a single unique height value
to each 2D point, so cannot describe vertical terrain features (e.g., cliffs). DTMs are
therefore “2.5D” rather than truly 3D models of the terrain (Weibel and Heller, 1991).

Fig. 1. Left: DTM, Right DSM. Data obtained from airborne laser scanning, resolution 1 m, area covers
1000x1000 m and shows a part of Germanys high‐speed train network between Nuremberg and Ingolstadt.
Image courtesy Wolfgang Reinhardt and Walter Henninger (Munich).

While DTMs represent the bare ground of the terrain, a Digital Surface Model (DSM)
describes heights of vegetation (e.g., trees) and of man‐made features (e.g., buildings) too
(Fig. 1). It is thus important to distinguish between DTM and DSM over vegetated or built
areas. A closely related term is Digital Elevation Model (DEM), which is sometimes used
synonymously with DTM, but often as an umbrella term to describe both DTM and DSM
(Wood, 2008; Hutchinson and Galant, 2009; Shinagre and Kale, 2013). DEM is often used for
elevation models from remote sensing (e.g., radar or photogrammetry). These models are
rather DSM than DTM unless vegetation and building heights are removed.
The concept of DTM is not only limited to Earth’s visible terrain surface. It also finds
application in bathymetry (digital bathymetry models describing the geometry of the sea
floor), polar geodesy (digital bedrock models to describe the rock below the ice sheets), and
planetary sciences (digital elevation models of the planetary surfaces), among many other
areas of application.
Representations
Common mathematical representations for DTMs include regularly spaced grids (2D raster
or matrix form), irregularly distributed 2D points (variable point distances), 1D‐profiles,
contours (i.e., lines of constant heights), and Fourier series (Li et al., 2004). Triangulations of
irregularly distributed 2D points, known as Triangulated Irregular Networks (TIN) allow
incorporation of terrain break lines (ridges, valleys), and extreme locations (e.g., summits) to
better represent the terrain, but are more complex to handle than gridded DTMs. Gridded
DTMs are derived from TINs through computation of height values at 2D raster locations
using some interpolation technique (e.g., fitting of linear or curved surfaces, least‐squares
prediction, kriging). With increasing spatial resolution, gridded DTMs tend to approximate
break lines and extreme locations better, while containing more redundant information in
flat terrain. DTMs are most commonly provided in terms of regularly spaced grids which
allows simple data handling, manipulation and storage (cf. Li et al., 2004, El‐Sheimy et al.,
2005; Peckham and Gyozo, 2007). Another efficient but less common way to present a DTM
is through a set of surface spherical harmonic coefficients computed from a high‐resolution
2D raster. DTM heights at arbitrary locations are then obtained through series expansion of
the coefficients. Currently, those spherical harmonic DTM representations reach spatial
resolutions of about 2 km, equivalent to harmonic degree 10,800 (Balmino et al., 2012; Hirt
and Rexer 2015).
Sources of DTM data
From a range of sources relying on ground‐based, airborne and space‐borne surveying
techniques DTM data is available for parts and the whole of the Earth’s surface at different
spatial resolution. Traditional ground‐based methods (e.g., Kennie and Petrie, 1990) are
very accurate, but tedious, and therefore mostly limited to small areas. Field surveying
based on tachymetry delivers terrain heights at selected locations. Satellite surveying
techniques (GNSS) provide terrain heights at discrete locations, or along profiles (e.g.,
roads). GNSS heights are sometimes used as a check on DEMs from remote sensing. Another
source of terrain heights are digitized contour lines from existing topographic maps.
Modern airborne‐ and space‐borne sensors are considerably more efficient for terrain
surface mapping than ground methods. This is because the terrain is sampled at a vast
number of locations in little time, and the sensor movement in air or space allows mapping
of regional or even global profiles. Important mapping sensors on flying platforms are (i)
image‐based (photogrammetry), (ii) laser‐based (lidar), and (iii) radar‐based (e.g., radar
interferometry).
Photogrammetric methods (e.g., Baltsavias, 1999) use overlapping pairs of images showing
the terrain from different angles (stereo principle). Stereoscopic processing yields terrain
heights for ground points captured and identified in both images. Aerial photogrammetry is
used for terrain mapping at national scale, and satellite‐borne imagery (e.g., SPOT, ASTER or
ALOS satellites) globally. Dense vegetation cover and clouds reduce the completeness of
DTM data from photogrammetry.
To establish terrain heights with laser‐based methods (lidar), travel time of short light pulses
‐ emitted from ha laser measurement system and reflected by the ground – is measured.
Rotating mirrors spread series of laser pulses into swaths allowing to sample terrain height
profiles with high spatial resolution. Over vegetated areas, there are usually multiple return
pulses (e.g., reflection at the top of the canopy and at the ground), which provide
information on the bare ground elevation and vegetation height. Airborne lidar (airborne
laser scanning, cf. Baltsavias, 1999) is a standard in generating highly‐accurate and detailed
regional DTMs and DSMs in an efficient manner. Space‐borne lidar, deployed e.g., on the
ICESAT satellite, is well‐suited to obtain DTMs over the Earth’s ice sheets. Similar technique
used aboard planetary missions provided highest‐resolution DTMs for Mars and Moon.
Different to lidar, radar‐based techniques operate with microwave signals which are cloud‐
penetrating. Signal travel time, phase, of phase differences are used as measure. Commonly
used radar variants are SAR (synthetic aperture radar) with increased imaging resolution
through use of artificial apertures, and InSAR (interferometric SAR), where the stereo
principle is applied with two nearby antennas recording the same radar pulse (Rabus et al.,
2003). Differences in signal phases depend on the terrain height, so can be used to derive
DTM data. In steep terrain, radar‐based topographic mapping may produce voids in the DTM
caused by so‐called radar shadows or signal layover. The most prominent example for space‐
borne InSAR is the Shuttle Radar Topography Mission (SRTM). Radar systems are being
operated aboard planes for special applications too, e.g., to obtain elevation data of the rock
beneath Earth’s ice‐sheets. In a planetary context, radar was used to obtain a DTM of Venus,
a permanently cloud covered planet.
Earth elevation products
Publically and freely available DEMs with continental or near‐global coverage originate from
space‐borne topographic mapping missions, notably from the SRTM and ASTER missions.
The characteristics of selected DEMs with near‐global coverage are reported in Table 1.
The first near‐global high‐resolution DEM was acquired in the year 2000 from SAR
interferometry carried out aboard the Space Shuttle (Rabus et al., 2003). The SRTM DEM
covers most of Earth’s land areas between 60°N‐56°S. Originally produced at 1 arcsec (30 m)
spatial resolution, the DEM data has been released at lower resolution level of 3 arcsec (90
m) to the public in 2003. Recently, USGS announced to release the SRTM data set at 1 arc‐
sec resolution globally. The initial SRTM release was subject by numerous voids in steep
terrain, which were filled e.g., based on interpolation techniques, or auxiliary elevation data
sets and published by CGIAR‐CSI as version v4.1 (e.g., Jarvis et al., 2008). The most recent
SRTM USGS release (v2.1) still contains voids.
A higher‐resolution global DEM data set was released to the public by NASA/METI based on
stereoscopic imagery from the ASTER satellite. The global ASTER DEM extends towards the
pole regions (83° latitude) at 1 arc‐sec resolution. The current ASTER DEM release is
GDEM2 (Tachikawa et al., 2011). It describes the terrain surface with mostly greater detail
than SRTM, but is in places subject to artefacts such as diagonal stripes and peaks (e.g.,
Rexer and Hirt, 2014) which impede its direct use in some applications.
Table 1. Characteristics of selected near‐global DEMs. Acronyms: NGA = National Geospatial Intelligence
Agency, NASA = National Aeronautics and Space Administration, USGS = United States Geological Survey, METI
= Ministry of Economy Trade and Industry (Japan), CGIAR CSI = Consortium for Spatial Information, DLR =
Deutschen Zentrum für Luft‐ und Raumfahrt (Germany), SRTM = Shuttle Radar Topography Mission, ASTER =
Advanced Spaceborne Thermal Emission and Reflection Radiometer, GDEM = Global Digital Elevation Model,
JAXA = Japan Aerospace Exploration Agency

Mission/DTM Institution Coverage Spatial Source

*3
product Resolution
SRTM 1 arc‐ NASA, NGA, 60°N‐56°S 1 arc‐sec http://earthexplorer.usgs.gov/
second global USGS

SRTM USGS v2.1 NASA, USGS, 60°N‐56°S 3 arc‐sec http://dds.cr.usgs.gov/srtm/

JPL

SRTM CGIAR CSI CGIAR‐CSI 60°N‐56°S 3 arc‐sec http://srtm.csi.cgiar.org/

v4.1

ASTER‐GDEM2 METI, NASA 83°N‐83°S 1 arc‐sec http://asterweb.jpl.nasa.

gov/gdem.asp

Tandem‐X/ World‐ DLR 87°N‐87°S 3 arc‐sec*1 not yet available

DEM or more

ALOS World 3D JAXA 82°N‐82°S 1 arc‐sec*2 not yet available

or more
*1
0.4 arc‐sec resolution as commercial product, *2 0.15 arc‐sec as commercial product,
*3
An angle of 1 arcsec equates to about 30 m in North‐South direction on the Earth’s surface

It is important to note that the SRTM or ASTER DEMs cannot be considered as pure DTMs. In
case of ASTER the underlying optical stereo methods applied did not probe the bare ground
over built or vegetated areas, but rather the top of terrain features (DSM). In case of SRTM,
radar reflections cannot be unambiguously attributed to either the ground or the top of
canopy, which is why SRTM is probably best characterized as mixed DSM/DTM. Overall both
the SRTM and ASTER DEMs represent terrain heights at the 10 m accuracy level. The
accuracy, however, can be better in non‐vegetated and flat terrain (few metres), while being
worse in mountain areas (e.g., Rexer and Hirt, 2014).
Using improved InSAR technology, the TanDEM‐X satellite mission (Krieger et al., 2007) has
now mapped the Earth’s global topography with a spatial resolution of 0.4 arc‐sec (about 12
metres). While TanDEM‐X is a commercial mission, a down‐sampled elevation model
(WorldDEM) with 3 arc‐sec resolution (commensurate to SRTM) and global coverage is
intended to become freely available for science. As a commercial product, a pure DTM (with
building heights and vegetation removed) will be produced. Further global DEM data sets
can be expected to become available from remote sensing in the future. As an example, a
new 3D model of the Earth’s surface of up to 0.15 arc‐sec (5 m) resolution will be generated
from optical stereoscopy carried out aboard the ALOS satellite (ALOS World 3D, Tadano et
al., 2014).
A number of composite DEMs exist that describe Earth’s surface in the absence of water and
ice masses, by providing bathymetric depths over the oceans and major lakes, and
elevations of bedrock over Antarctica and Greenland. Examples include NOAA’s ETOPO1 (60
arc‐sec global resolution) and Scrippts Institution of Oceanography’s SRTM_30PLUS (30 arc‐
sec global resolution, no ice information).
Complementary to global terrain models from remote sensing, surveying agencies of many
countries have generated DTMs from a composite of data sources, e.g., ground surveys,
contour scans, and airborne photogrammetry. These national models are mostly commercial
products. For parts of some countries and regions, extremely detailed DTMs have been
produced from airborne laser scanning with often 1‐meter‐resolution and sub‐m‐precision.
Applications
Terrain models play a fundamental role in geosciences and engineering, and have numerous
applications. They can be used to calculate derived quantities, such as volumes, slope,
curvature, sun exposure, hill shade, contours, visibility from given sites, drainage, and
gravitational attraction. Application examples for DTM include its use as a base layer in
geographic information systems (GIS), e.g. for planning of engineering structures (roads,
railways, canals), hydrology (drainage and catchment area analysis), coastal protection
(inundation), mass movements in mountain areas, rendering visualisations and topographic
maps, planning of radio networks and alternative energy power plants, and rectification of
photogrammetric imagery (orthophotos).
In the narrower field of gravity field modelling and physical geodesy, DTM data is a pivotal
data source providing geometry information of the topographic masses. Using gravity
forward modelling techniques, the gravitational attraction of the masses is computed from
DTM data, and can be (i) subtracted from observed gravity values to highlight signatures of
mass anomalies in the Earth’s interior, (ii) used as reduction in geoid determination, or (iii)
utilized to predict a detailed gravity field over otherwise less surveyed areas. For these
applications, the availability of DTM rather than DSM data is important.
Summary
DTMs, conceptionally introduced more than half a century ago, are today in wide use in
geodesy and beyond to approximate the Earth’s relief for a broad range of applications.
DTMs are publically available with near‐global coverage from satellite‐based mapping
missions (notably SRTM and ASTER) at 1 to 3 arc‐sec resolution and 10‐20 m accuracy. In the
near future, new commercial DTMs from the TanDEM‐X and ALOS missions will provide
more detailed terrain information than currently available. At a national level, DTMs exist for
many countries at varying resolution, mostly from ground‐based, airborne lidar and
photogrammetric surveys. On a planetary scale, high‐resolution DTM information is
available for Moon, Mars, Venus and other bodies.
Cross references
Gravity Forward Modelling, Topographic Effects, Regional Gravity Field Determination
Bibliography
Balmino, G., Vales, N., Bonvalot S., and Briais, A. 2012. Spherical harmonic modelling to ultra‐high degree of
Bouguer and isostatic anomalies. J. Geod., 86, 499‐520.

Baltsavias, E.P., 1999. A comparison between photogrammetry and laser scanning. ISPRS Journal of
Photogrammetry & Remote Sensing, 54, 83‐94.

El‐Sheimy, N., Valeo, C., and Habib, A., 2005. Digital Terrain Modeling: Acquisition, Manipulation And
Applications. Norwood: Artech House.

Hirt, C., Rexer, M., 2015. Earth 2014: 1 arc‐min shape, topography, bedrock and ice‐sheet models ‐ available as
gridded data and degree‐10,800 spherical harmonics. International Journal of AppliedEarth Observation and
Geoinformation, 39, 103–112.

Hutchinson, M.F., and Gallant, J.C., 2005. Representation of Terrain. In Longley, P.A. et al. (ed.) Geographical
Information Systems: Principles, Techniques, Management and Applications, Hoboken: John Wiley and Sons.

Jarvis A., Reuter H.I., Nelson A. and Guevara, E. 2008. Hole‐filled SRTM for the globe Version 4. Available from
the CGIAR‐CSI SRTM 90m database. http://srtm.csi.cgiar.org

Kennie, T.J.M., and Petrie, G., 1990. Digital terrain modelling. In Kennie, T.J.M., and Petrie, G., (ed). Engineering
Surveying Technology, Oxon: Taylor and Francis.

Krieger, G. A. Moreira, et al., 2007. TanDEM‐X: a satellite formation for high resolution SAR interferometry. IEEE
Trans. Geosci. Remote Sens., 45, 3317–3341.

Li, Z., Zhu, C., and Gold, C., 2004. Digital Terrain Modeling: Principles and Methodology. Boca Ration: CRC
Press.

Peckham, R.J., and Gyozo, J., (Eds) 2007. Digital Terrain Modelling, Development and Applications in a Policy
Support Environment. New York: Springer.

Rabus, B., Eineder, M., Roth, A., and Bamler, R., 2003. The Shuttle Radar Topography Mission – A New Class of
Digital Elevation Models Acquired by Spaceborne Radar. ISPRS Journal of Photogrammetry & Remote Sensing,
57, 241‐262.

Rexer, M., and Hirt, C., 2014. Comparison of free high resolution digital elevation data sets (ASTER GDEM2,
SRTM v2.1/v4.1) and validation against accurate heights from the Australian National Gravity Database;
Australian Journal of Earth Sciences, 61, 213‐226.

Tachikawa, T., et al., 2011. ASTER Global Digital Elevation Model Version 2 – Summary of Validation
Results,https://www.jspacesystems.or.jp/ersdac/GDEM/ver2Validation/Summary_GDEM2_validation_report_f
inal.pdf.

Tadono, H., Ishida, H., et al., 2014. Precise global DEM generation by ALOS PRISM. ISPRS Annals of the
Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume II‐4, 71‐76.

Shingare, P.P., Kale, S.S., 2013. Review on Digital Elevation Model. International Journal of Modern Engineering
Research, 3: 2412‐2418.

Weibel, R., and Heller, M., 1991. Digital Terrain Modelling. In Maguire, D.J. et al. (ed.) Geographical Information
Systems: Principles and Applications. London: Longman.

Wood, J., 2008. Digital elevation model (DEM). In Kemp, K. (ed.), Encyclopedia of geographic information
science. Thousand Oaks, CA: SAGE Publications, pp 107‐110.