The Need for Better Information
Development and exploration in previously remote parts of the world are creating an increasing demand for current land topography and land-use maps, as well as for other information products. These maps are critical to a wide range of activities, such as planning and managing the development of land-use, management of natural resources, and engineering studies. In these areas, a growing number of users have begun to implement thematic and cartographic mapping solutions in their systems for decision-making and management.
Organizational and technological changes are rapidly occurring worldwide throughout the mapping industry. In order to acquire thorough knowledge of any landscape, users must obtain the widest possible range of data sources. Therefore, the mapping projects seen today are coming to depend on multi-source techniques for remote sensing that are able to give users data sets in high resolution that cover broad geographical areas. Additionally, digitally-formatted mapping products provide highly useful information with respect to the Geographical Information Systems (GIS) market, which is experiencing rapid growth. GIS systems are currently widely used to incorporate information derived from a range of sources, such as land management planning and monitoring.
Also, producers of maps are seeking the most up-to-date and efficient ways to achieve the best possible quality of their products at the lowest possible cost.
The Limitations of Traditional Techniques
Traditionally, maps are produced using ground measurements and aerial surveys. These typical techniques for map-making depend on the processes involved in scanning and digitization in order to be updated and incorporated into current systems of decision management. However, this does not allow large-scale, regional maps to be produced quickly and inexpensively.
As a result, information used to produce maps is not updated regularly, causing those maps to rapidly become obsolete and of little use. For example, a 1987 survey by the United Nations demonstrated that more than 40% of the land surface of the world is still uncovered by maps at the scale of 1:100 000, nearly 50% is not yet covered by 1:50 000 scale maps, and 80% is not documented using 1:25 000 scale maps.
The Beneﬁts of Space-Based Monitoring
Because of its capacity to quickly provide current information and its broad spatial coverage, space-born imagery gaining in popularity within the mapping industry. Data products which use optical remote-sensing are utilized to create space maps on a wide spectrum of scales in order to meet a large variety of user needs. Such maps include Digital Terrain Models (DTMs) created from stereoscopic images, along with standard 1:50 000 scale map compilation and updating. Space-derived maps are geocoded and are annotated in the same ways as their traditional map counterparts.
While information derived from optical remote-sensing is particularly useful in remote regions, it has two primary drawbacks:
• it cannot be utilized in conditions of excessive cloud cover
•for some applications, the space map does not provide sufficient accuracy in location of individual points
The Contribution of ERS SAR
The European Remote Sensing Satellites, ERS-1 and ERS-2, carry an instrument known as the Synthetic Aperture Radar (SAR). SAR has provided highly valuable information to those growing markets which depend on broad-coverage maps on scales from 1:1 000 000 to 1: 50 000. After ERS-1 and ERS-2 were launched in July 1991 and April 1995, respectively, the SAR data collected has been used to create and regularly maintain a database with that information.
Under any weather conditions, the SAR is able to actively produce images by analyzing the Cband and VV polarization echoes sent from the satellite that the Earth’s surface then backscatters. The area covered by an ERS SAR scene measures 100 km by 100 km and is highly accurate to a particular geographical location. This interaction between the ground’s surface and radar waves makes SAR images quite different from those derived optically.
The SAR’s indirect viewing angle makes topographic features much clearer, while the optical radar’s ability to penetrate clouds greatly improves ground visibility. Both images are post-processsed and geo-referenced by the geographic department of research, Cellule d’Etudes en Geographie Numerique (CEGN), the geographic research department of the French armament agency.
The compatibility and integration of optical and ERS SAR images increases users’ ability to derive thematic information, with respect to cartographic map generation and updating. Topographic and thematic maps constructed using data from ERS SAR gives users the capability of detecting and identifying particular features like hydrographic structures and networks, which are of particular importance in geomorphologic geology and analysis.
Where cloud cover prevents accessing optical satellite information or where that data is otherwise unavailable, such as in tropical regions, SAR data can be used to generate valuable space maps that are not available using any other means. SAR images may be the sole source of information used to create continental-scale space imagery at high resolution, and can provide extremely valuable thematic data due to the sensor’s ability to discriminate among a broad spectrum of land-cover types.
ERS’ ability to provide large-scale thematic mapping and image mosaicing on a worldwide basis is well-known. Those are based on modern processing technologies that improve both the geometric and radiometric quality of the information. For instance, alterations in the SAR data’s radiometric resolution are caused by the speckle, a “noisy” effect inherent to the SAR system resulting from the coherent quality of the SAR signal.
This noisy effect can be reduced by the application of filtering techniques to the SAR images, which also enhances the quality of the images. One example demonstrates how techniques of multi-temporal filtering have been used on SAR data by SERTIT in France, a company specializing in GIS systems and image processing, reducing the speckle and thus allowing for enhanced radar imagery to be used in the context of GIS. Temporally-filtered ERS images provide highly useful data in this context and make interpretation easier and more accurate.
ERS SAR image maps can be used in a number of applications:
• rectiﬁcation: to correct inaccurate or out-of date maps, space maps created with inaccurate localization information obtained from remote-sensing, or even to fix a Digital Terrain Model (DTM).
• cartography: for map updating and compiling, utilizing the thematic information obtained by radar for wide application and its ability to provide topographic information
• localization: to identify or detect control points and locate targets on the surface of the ground, complementary to measurements taken by GPS measurements in remote areas.
Furthermore, information obtained from ERS SAR can be used in the context of topographic mapping; radargrammetry and interferometry (INSAR) provides DTMs. Stereo pairs of radar images with varying viewing angles are used to generate DTMs for radargrammetry, just as in optical stereo imaging.
A combination of two ERS SAR images with slight variations in their geometrical configurations forms the basis of INSAR. This system produces extremely accurate DTMs, which primarily depend on the overall stability of the observed surface with respect to atmospheric effects and radar signal phase, both of which can affect image acquisition.
The altimetric accuracy can be as close as a few meters, under favorable conditions. Surface movements can be measured with a sensitivity of even a few centimeters over large surfaces using differential interferometry. This technique proves highly useful in the observation of active volcanoes, faults, and earthquakes as well as in subsidence monitoring.
Additionally, thematic information can be extracted from ERS SAR data for land-use mapping using interferometry. This data is obtained by calculating the correlation of the SAR signal between the pair of images. This interferometric correlation, or coherence, primarily depends on the interactions of radar waves with the desired target and its temporal stability. That derived imagery is useful as another channel that allows the generation of multi-band radar products.