8.3 Availability of the terrain models

There was a long period, when the national geodetic/geoinformation data providers offered the elevation models, developed on the base of contours their own topographic maps. Nowadays, the availability of models, resulted from laser scanning is more and more frequent. These models show the elevation in the horizontal coordinate system of the given country, and similarly, the elevations are represented in the local vertical datum. The quality characteristics are also determined by the technology level of the providing country and by the scale and quality of the available topographic maps. In most cases, the accuracy and the resolution of the laser scanned models are better than the ones of the contour-based models. However the national data providers are constantly working on actualization and quality improvement of their data, there are no such data available for the huge majority of the Earth’s surface.

The situation is different, and sometimes surprisingly better in case of the medium-resolution terrain models. Different international groups were formed in the 1990s to compile global models using the local ones. At the end of the last millennium, such datasets (e.g. the GTOPO30) were issued and widespread used. However, according to their edited/mosaicked being, the data quality of these models are heavily varies from place to place. The situation was fundamentally improved by the Shuttle Radar Topography Mission (SRTM) dataset, published in 2003.

This program was started in 1996 by the American NASA (National Aeronautic and Space Administration), aiming to mapping the relief of cca. 80% of the Earth’s surface, using a radar system, onboard of the Space Shuttle (Fig. 49). After some delays, the space shuttle Endeavour has been launched in 11 February, 2000, onboard with all necessary instruments for the measurement. The whole survey campaign lasted 11 days. The space measurements were completed and supported by extent surface GPS-measurements as well as placing many (around 70 thousands) artificial radar reflectors at pre-set positions, to provide geo-reference. The data processing took 18 months, led by the NIMA (National Imagery and Mapping Agency) of the US Ministry of Defense. According to the agreement between the NASA and the NIMA, with the permission of the NASA, the dataset is archived and published by the USGS (United States Geological Survey).

SRTM measurement overview

Fig. 49. The settings of the SRTM measurement onboard of the Space Shuttle.

In the frame of the project, the digital elevation model of the mapped area was completed in two different resolutions: the pixel size of the finer version is one arc second (available publicly only for the territory of the United States) while the general version has the pixel size of 3 arc seconds (cca. 90-100 meters is mid-latitudes). Thus, such a public database was created, whose existence and use should be known for any specialists, working with geo-information technology (Fig. 50).

For the measurement, onboard radar equipment was used. As the orbit inclination of the space shuttle in the experiment was 57 degrees, it didn’t fly over the polar regions. In the frame of the SRTM program, therefore, was between the 60th degrees of northern and the 57th degrees of southern latitudes. For example, the database is not covering Finland; its topography is not available in it. The resulted 3-arcdegree resolution data is available for everybody on the Internet. The latitude-longitude grid follows the parallels and meridians, the horizontal datum is the WGS84. The elevations are interpreted above the level of the EGM96 global geoid model.

Color and shaded part of the SRTM elevation model

Fig. 50. The elevation model of the Székely Land (eastern Transylvania) in the SRTM dataset.

While using the dataset, we shall keep in mind that it was constructed with radar technology. We have uncertain signals from water surfaces (because of the unavoidable waves), so at the seas, lakes and rivers, we have false data. Majority of them was filtered out during the data processing, and these pixels have NULL cell values. Similar NULL value have been arranged for many mountainous pixels, mainly in deep valleys, which were in radar shadow, according to the survey geometry, and we don’t have radar backscatter signal from. This kind of data absence is more frequent in the high mountains. If necessary, the missing data can be completed from other, lower resolution models. The 5.6 centimeter wavelength radio signals are not penetrating the dense or even the medium foliage and, of course, scattered back from the solid roofs or walls of the buildings. Thus, the elevation values of the model represent the geoid height of the layer that is the reflector for the 5.6 centimeter wavelength electromagnetic signal. In the regions of cities or forests, the effect of the buildings and the trees is in our data.

On the Mars, the thin atmosphere enables to survey the surface elevation by laser altimetry. The resulting MOLA (Mars Orbiter Laser Altimetry) project provides an SRTM-like elevation model with a horizontal resolution around half kilometer, of course without any artificial of vegetation ‘noise’ (Fig. 51). In the last decade, the altimetry of the Mars has been significantly improved.

Color and shaded part of the MOLE model

Fig. 51. The southern foreland of the Huygens crater in the Mars, shown by the MOLA elevation dataset.