Digital elevation models (DEMs) are needed for mapping and modelling natural hazards and risks that are influenced by topography, for example floods and landslides.
Since water only runs downhill and the flow of water is channelled by objects that are higher than the water level, elevation information is used to model the flow of water and to determine possible inundation areas in case of floods. Step-by-step procedures how to use DEMs for flood modelling are provided in the practice recommended by University of Purdue (link to the practice).
Slope is a decisive factor for landslides. Generally, landslides occur only in areas with a certain initial relief contrast in the source area of the landslide. Slope can easily be derived from DEMs; any standard remote sensing software provides tools to automatically derive a slope image from a DEM. In a study performed at the GFZ Helmholtz Center Potsdam and published in August 2014, a DEM of 30 m spatial resolution derived from the X-band data of the Shuttle Radar Topography Mission in February 2000 is used in combination with multi-temporal NDVI trajectories derived from 5m resolution RapidEye optical satellite imagery to automatically detect landslides over large areas. The approach can be used for generating multi-temporal landslide inventories and regular monitoring of landslide activity. It was applied successfully in a study area in Kyrgyzstan with an average landslide size of approximately 13,000m2. According to the authors of the study, further methodological development is needed to adapt the approach to regions where the natural environments largely differ from the conditions in Central Asia. (link to the open access paper)
Furthermore, DEMs are necessary for orthorectifying satellite imagery. Satellite imagery taken from different view angles can only be compared when orthorectified, otherwise the pixels do not align.
How is elevation measured from space?
Elevation can be measured from space using different approaches designed for different sensors. For radar sensors interferometric SAR (InSAR) is applied. For optical sensors stereo images are analyzed to derive elevation information.
Elevation measurements on the ground also make use of space technology. Handhelds with built-in receivers for Global Navigation Satellite Systems (GNSS) like GPS are used to estimate elevation on the ground. With Differential GPS (D-GPS), elevation can be measured on the ground with high precision. Such measurements are used to model and to validate elevation based on satellite imagery.
Very precise and high resolution elevation mesurements are possible using airborne laser-ranging (lidar) instruments. The first space-borne lidar instrument for continuous global observations of Earth was the Geoscience Laser Altimeter System (GLAS) onboard ICESat (2003-2009), which was designed to measure ice-sheet topography and associated changes. The resulting DEMs for Antarctica and Greenland have a horizontal spatial resolution of 500m and 1km. ICESat-2 is slated for launch in 2017 with the ATLAS instrument onboard, also desigend for the observation of ice-sheet elevation change (cf. this video of NASA to understand how the ATLAS instrument will work).
How can I access digital elevation models?
The most widely used digital elevation models derived from satellite data are the products of the Shuttle Radar Topography Mission (SRTM), which was flown aboard the space shuttle Endeavour February 11-22, 2000. They are freely available in different versions:
- The first datasets of SRTM 1 arc-second global (approximately 30m horizontal resolution) have been newly released in September 2014 covering regions of Africa. Until then, 1 arc-second SRTM data was only availble for the United States. Further datasets for different regions will be released in consecutive phases. (link to the data)
- SRTM 3 arc-seconds non void filled and void-filled by NGA (approximately 90m horizontal resolution). The vertical error of the DEMs is reported to be less than 16m (source: JRC) or plus/minus 10m (source DLR). The data are available for the entire land surface between 60 degrees North and 60 degrees South. (link to the data)
- SRTM 3 arc-seconds void filled by CGIAR-CSI and JRC. (link to the data)
- While the above DEMs are based on C-band data of the SRTM, the German Aerospace Centre (DLR) also provides DEMs based on X-band data of the SRTM. The horizontal resolution is 25m and according to DLR, the vertical precision is plus/minus 6m. The data are available for the whole globe between 60 degrees North and 60 degrees South but not covering the entire land surface. (link to the data)
The SRTM X-band mission served as a precursor for the TerraSAR-X add-on for Digital Elevation Measurement (Tandem-X) mission of DLR and Astrium. DEMs derived of TanDEM-X data have a horizontal resolution of 12m and a vertical accuracy better than 2m (source: DLR). The Tandem-X mission has started in June 2010 and the expected lifetime is 5.5 years. Currently, TanDEM-X data are only available via the TanDEM-X Science Service System for scientific use; all Announcements for Oppartunity (AOs) are published there. In May 2014, the first demo data were published for two selected areas, namely the Badlands National Park in South Dakota and the Flinders Ranges in Australia (link to the demo data).
An example of DEMs derived from optical sensors is the ASTER Global DEM. It covers the entire land surface of the Earth between 83 degrees North and 83 degrees South with a horizontal resolution of 1 arc-second (approximately 30m). The vertical accuracy is generally between 10 and 25m. The ASTER DEM contains some artifacts which could affect its utility incertain applications (cf. validation summary report). (link to the data)
Other freely available DEMs includes GTOPO30 at 1km horizontal resolution derived from different raster and vectot sources covering the land surface of the whole globe (link to the data) or the Global Multi-Resolution Topography (GMRT), which inlcudes Sonar Seafloor Bathymetry (link to the data).
A nice and esay to use tool to look at elevation profiles of any are area of interest are the Google Earth elevation profiles (how to). The image below shows an elevation profile for the landslide in Indonesia for which the International Charter on Space and Major Disasters was activated on 15 December 2014.