Other Space Activities

Digital Elevation Models (DEMs)

Last updated:Sep 19, 2025

Measurement Types

Digital Elevation Models (DEMs) are three-dimensional digital representations of the topography of a geographic area in the form of a georectified point-based or area-based grid, encoding the height and slope of the Earth’s surface. DEMs are necessary for radiometric terrain correction, and assist in georeferencing Earth Observation (EO) imagery, identifying geological structures for use in mapping and terrain analysis, hydrology and flood modelling, and monitoring natural disasters. They can be created using traditional in situ techniques, such as ground surveying or manually digitising map contour lines, aircraft or drones, or using modern space-based techniques, such as Synthetic Aperture Radar (SAR) and Interferometric SAR (InSAR).

Colourised Global DEM from ASTER GDEM 003 (Image credit: NASA)

DEMs can be classified into Digital Terrain Models (DTMs), which provide a ‘bare Earth’ model, and Digital Surface Models (DSMs), which include additional surface features such as buildings or vegetation. Typically they use rectangular grids of pixels, each of which contains a single value representing the elevation. Alternative forms of digital topography, such as triangulated irregular networks (TINs), contours, and point clouds are not classified as DEMs because they are not grid based. 1) 2) 3)

Applications

DEMs are commonly used to help correct distortions in satellite data products caused by terrain elevation in a process known as orthorectification. They are also used to identify geological and other structures of interest.

DEMs are commonly used in Geographic Information Systems (GIS) applications to perform spatial analysis of terrain and other geographic features. Viewshed analysis is one such technique that uses DEMs to identify visible areas from specific points (i.e. analyse terrain occlusion); it is used in military planning and urban development. Furthermore, they can be used to develop 3D simulations of natural phenomena such as landslides and mass-movement events. 1)

Different types of DEMs are essential for use in agriculture and forestry. DSMs show vegetation, while DTMs can be used for applications in hydrology and soil erosion including delineating drainage networks and determining catchment areas for water resource management. 1) 4)

Techniques

DEMs can be created using traditional topographic maps or space-based EO data. In situ techniques include ground surveying using a theodolite or manually digitising map contour lines and interpolating them into a raster format. However, these processes can be labor-intensive and time-consuming, especially in rugged terrain.

Stereo photogrammetry can be used, whereby corresponding pixels in overlapping images are found, enabling 3D reconstruction via triangulation based on known exterior and interior orientations. This can be done with aerial photography and satellite imagery. Differential Global Positioning System (DGPS) measurements can be made with specialised devices, including a stationary reference station and mobile receivers, that use satellite information to survey points across an area and determine GPS positioning. These have to be interpolated to yield a continuous raster. 1) 5)

Contemporary techniques for the development of DEMs make frequent use SAR and InSAR data due to their ability to provide complete global coverage, and sensitivity to change. InSAR can be employed in combination with deep learning to correct for urban influence, such as is done in Climate Central’s CoastalDEM.

Lidar measures the time it takes for laser pulses to be reflected back to the satellite and therefore derive elevation data. The resulting data point clouds are processed to create a continuous raster of the Earth’s surface, from which high-resolution DTMs can be derived. NOAA’s Digital Coast provides coastal area lidar data and USGS Earth Explorer serves U.S. Geological Survey (USGS) lidar data. 1) 4) 6)

Figure 1: Removing vegetation by processing lidar point clouds (left) to bare-earth DTMs (right) to reveal past landslides and steep slopes at risk of failure (Image credit: USGS)

Figure 2: NOAA’s Digital Coast (Image credit: NOAA)

Example Products

Digital Surface Models (DSM) and Digital Terrain Models (DTM)

A DSM is a type of DEM that represents the top surface of the Earth, including vegetation and human-made structures. A DTM represents the bare-earth surface, with vegetation and structures removed. Hybrid DEMs likewise exist, which extract the height of different elements for particular user needs. These include non-vegetated surfaces, which exclude vegetation while including man-made structures, and the converse, non-urbanized surfaces, which includes man-made structures while excluding vegetation. 3)

Figure 3: Terrain being represented by a digital surface model and digital terrain model

Table 1: Features of DTMs and DSMs
Feature	DTM	DSM
Surface	Bare-earth, excluding surface features	Includes all surface features
Key applications	Hydrology, flood risk, infrastructure	Urban planning, telecoms, forestry

Digital Surface Models (DSM)

As DSMs display all above-ground features, they are useful for applications in urban planning. DSMs are also used to assess zone encroachment in aviation, visualisation, disaster management, navigation, and vegetation management. DSMs with a larger coverage area are useful for the design of telecommunications networks that requires line-of-sight planning for longer distances. 3) 4) 5)

Digital Terrain Models (DTM)

DTMs focus solely on the natural elevation contours, creating a ‘bare Earth’ model that excludes vegetation and human-made structures. It is a three-dimensional digital representation of the topographic surface showing elevations and natural features such as rivers and ridgelines. The development of DTMs involves digitally removing non-ground features to isolate the bare Earth terrain elevation from traditional DSMs, and are often developed using algorithms and manual editing techniques. Additional steps like interpolation are also applied to create a smooth, continuous surface in areas with sparse data. 1) 3) 4)

Figure 4: Comparison of the processing levels of 5 m resolution DSM: Level 1 - Raw (left), Level 1 - Voids infilled (middle), Level 2 - DSM (right) (Image credit: AW3D)

DTMs are often developed from DSMs based on data from radar and lidar missions, because their sensing beams can penetrate some vegetation, and reach between structures. Their derivation involves removing surface features from the point cloud and sampling the ground elevation in uniform increments. 6) 7)

DTMs are essential for applications requiring accurate terrain representation, for example, hydrological modelling, slope analysis, investigation of soil erosion, and infrastructure planning. 4)

Figure 5: Comparison of DSM (left) and DTM (right) (Image credit: AW3D)

Point-based and Area-based DEMs

For area-based sampling, each DEM value represents the average elevation across a pixel. This reduces undersampling and ensures the data is representative of the average ground surface within that pixel. Meanwhile, point-based sampling techniques result in grids with each DEM value representing the elevation at a specific point. This makes point-based DEMs more variable and sensitive to pixel grid alignment. 3)

Digital Elevation Models (DEMs) are often stored in GeoTIFF format. In this format, pixels can be defined in two ways:

RasterPixelIsArea – the elevation value represents the average for the area covered by the pixel.
RasterPixelIsPoint – the elevation value represents a single point location, typically at the pixel’s centre (see Figure 6).

When combining DEMs from different sources, it is important to account for differences in pixel alignment, sampling method, and storage convention to avoid mismatches or artefacts. Digital Elevation Models (DEMs) are often stored in GeoTIFF format. In this format, pixels can be defined in two ways. 3)

Figure 6: GeoTIFF RasterPixelIsArea DEM (A) GeoTIFF RasterPixelIsPoint DEM (B) The difference between the elevation storage in the GeoTIFF files for the ASTER GDEM and ALOS AW3D30 DEMs (C) 3)Figure 6: GeoTIFF RasterPixelIsArea DEM (A) GeoTIFF RasterPixelIsPoint DEM (B) The difference between the elevation storage in the GeoTIFF files for the ASTER GDEM and ALOS AW3D30 DEMs (C) 3)

Table 2: Freely available quasi-global DEMs 3)
DEM	Resolution	Data Source	Provider	Sampling method	Storage method	Acquisition date
SRTM (v3)	1″, 3″	C band radar	NASA	Area-based	RasterPixelIsPoint	2000 (11 days)
ASTER GDEM (v3)	1″	Stereo NIR imagery	NASA/METI	Area-based	RasterPixelIsArea	2000–2013
ALOS World 3D AW3D30 v3.2	1″	Stereo pan imagery	JAXA	Area-based	RasterPixelIsArea	2006–2011
NASADEM	1″	Reprocessed C band radar	NASA	Point-based	RasterPixelIsPoint	2000 (11 days)
Copernicus DEM GLO30 and GLO90	1″, 3″	X band radar, Edited WorldDEM	ESA/Airbus	Area-based	RasterPixelIsPoint	2011–2015
TanDEM-X DEM	3″	X band radar	DLR	Area-based	RasterPixelIsPoint	2011–2015
MERIT	3″	Radar and Stereo pan imagery	University Tokyo	Area-based	RasterPixelIsArea	2000–2013

Related Missions

Optical Constellation in Three Dimensions (CO3D)

CO3D (Optical Constellation in Three Dimensions) is a high resolution optical imaging and three dimensional mapping mission of the French National Centre for Space Studies (CNES). The four CO3D satellites launched in July 2025 with the aim of capturing imagery for the production of a worldwide DEM, with relative 3D accuracy of 1 m.

Figure 7: Example of CO3D GLOBAL-DEM tiles production strategy (Image credit: CNES)

Terra (EOS/AM-1)

Terra (EOS/AM-1), launched in December 1999, is a joint mission between the US, Japan, and Canada. Terra’s Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER), is a high resolution optical imager, providing stereoscopic observation of local topography (VNIR: 15 m, stereo: 15 m horizontally and 25 m vertical). ASTER has a near-infrared backward-looking telescope that captures stereo image pairs along the same orbital path

The ASTER GDEM (Global Digital Elevation Model), first released in June 2009, was generated using stereo-pair images collected by the ASTER instrument. The latest version, ASTER GDEM Version 3 (GDEM 003), was released by the Ministry of Economy, Trade, and Industry (METI) of Japan and NASA in August 2019. The coverage of ASTER GDEM spans from 83° N to 83° S, encompassing 99 percent of the Earth's landmass.

Figure 8: A visualisation of ASTER Global Digital Elevation Model (GDEM) data over Mount Everest and the Himalayas (Image credit: LP DAAC)Figure 8: A visualisation of ASTER Global Digital Elevation Model (GDEM) data over Mount Everest and the Himalayas (Image credit: LP DAAC)

SRTM (Shuttle Radar Topography Mission)

SRTM (Shuttle Radar Topography Mission) was a joint mission between NASA, US Department of Defense (DoD)/NGA (National Geospatial-Intelligence Agency), DLR (German Aerospace Center), and ASI (Italian Space Agency) that was carried as a payload on the Space Shuttle Endeavour (STS-99) in February 2000. The data collected from the two onboard interferometric Synthetic Aperture Radars during the 11 days in orbit was used to develop DEMs with a resolution of 30 m, with vertical accuracy of approximately 15 m.

Figure 9: An SRTM hillshade image of the Mount Everest region captured in February 2000 using radar interferometry equipment operating from inside Space Shuttle Endeavour's payload bay (Image credit: NASA)

ALOS (Advanced Land Observing Satellite) / Daichi

JAXA’s (Japan Aerospace Exploration Agency) ALOS (Advanced Land Observing Satellite) mission was launched in January 2006 for applications in cartographic mapping, regional observation, disaster monitoring and resource surveying. JAXA’s ALOS World 3D is a 5 m resolution DEM developed using stereo mapping based on data from the onboard PRISM (Panchromatic Remote-sensing Instrument for Stereo Mapping) instrument.

Figure 10: Example of the PRISM DSM in the Kingdom of Bhutan. (Left: 5 m resolution, middle and right: simulated DEMs with 30 m and 90 m resolutions) (Image credit: JAXA)

TanDEM-X

TanDEM-X is an X-band Synthetic Aperture Radar (SAR) satellite mission, launched in June 2010, and operated by the German Aerospace Centre (DLR)

The Copernicus DEM provides three different instances of global and European DSM based on data acquired through the TanDEM-X mission: EEA-10, GLO-30 and GLO-90. EEA-10 provides coverage of 39 European countries at a resolution of 10 m, with a surface coverage of approximately 6 million km², while GLO-30 and GLO-90 provide worldwide coverage at 30 m and 90 m, together totalling approximately 149 million km².

Figure 11: Copernicus DEM land surface coverage – EEA-10 (left) and Copernicus DEM – GLO-30 & GLO-90 (right) (Image credit: CDSE)

Figure 12: Copernicus DEM data at the three available spatial resolutions: 10 m (left), 30 m (centre) and 90 m (right) (Image credit: CDSE)

Airbus 3D World

Airbus Defence and Space provides a 3D and Elevation Portfolio that provides DTMs and DSMs with worldwide coverage at four resolutions ranging from 0.5 m to 5 m. Elevation 0.5, Elevation 1 and Elevation 4 are developed from stereo and tri-stereo optical data from Pléiades, a two-satellite constellation of CNES (French Space Agency) launched in December 2011. WorldDEM™ Neo is based on data collected worldwide by the DLR TerraSAR-X and TanDEM-X radar satellite missions, launched in June 2007 and 2010, respectively. The WorldDEM™ Neo has a 5 m spatial resolution, with 1.4 m vertical accuracy and 6 m horizontal accuracy.

Figure 13: The Airbus 3D and Elevation Portfolio (Image credit: Airbus)

For larger-scale applications like mapping or exploration of natural resources, Airbus Elevation 8 provides a DEM derived from stereo and tri-stereo optical satellite data from Satellite Pour l'Observation de la Terre 6 (SPOT 6) and SPOT 7, launched in 2012 and 2014, respectively. It provides a 8 m spatial resolution and a vertical accuracy of 3 m.

TanDEM-X: Read more

TerraSAR-X: Read more

Pleiades-HR: Read more

SPOT-6 and SPOT-7: Read more

References

1) O'Donohue, D., “What are digital elevation models,” 9 December 2022, URL: https://web.archive.org/web/20251112024124/https://mapscaping.com/digital-elevation-models/

2) GIS Navigator, “Understanding Raster Data – How does it help in GIS?” 21 August 2024, URL: https://gisnavigator.co.uk/understanding-raster-data/

3) Guth, P. et al., “Digital Elevation Models: Terminology and Definitions,” 8 September 2021, URL: https://www.mdpi.com/2072-4292/13/18/3581

4) Dreusicke, M., “Digital Elevation Models,” 25 January 2022, URL: https://web.archive.org/web/20250914135730/https://www.cloudeo.group/blog/cloudeo-blog-space-1/digital-elevation-models-12

5) Marwaha, N. & Duffy, E., “Everything you need to know about Digital Elevation Models (DEMs), Digital Surface Models (DSMs), and Digital Terrain Models (DTMs),” 12 March 2021, URL: https://up42.com/blog/everything-you-need-to-know-about-digital-elevation-models-dem-digital

6) USGS, “What is the difference between lidar data and a digital elevation model (DEM)?” 17 Feburary 2022, URL: https://www.usgs.gov/faqs/what-difference-between-lidar-data-and-a-digital-elevation-model-dem

7) Fotias, V., “Applications of digital elevation models,” 7 February 2022, URL: https://web.archive.org/web/20251017080147/https://www.cloudeo.group/blog/cloudeo-blog-space-1/applications-of-digital-elevation-models-8