The Need for Better Information
Many remote areas of the World are now being opened to exploration and development, generating a growing demand for up-to-date maps of land topography, land-use and other information products. Such maps are of great importance for activities ranging from managing and planning land-use development, to natural-resource management and engineering studies. In this context, more and more users have started to integrate cartographic and thematic mapping products into their management and decision-making systems.
The mapping industry worldwide is currently experiencing rapid technologicaland organisational change. A prerequisite for obtaining a complete knowledge of any landscape is to rely on as many data sources as possible. Today’s mapping projects are therefore relying more and more on multi-source remote-sensing techniques capable of providing highresolution data sets over wide geographical areas. In addition, map products available in digital format are of a great value in the rapidly expanding market of Geographical Information Systems (GISs). These are now used extensively to integrate data from different sources in domains like land management, monitoring and planning.
Moreover, map producers are looking to the latest and least time-consuming methods to ensure the best product quality at the lowest cost.
The Limitations of Traditional Techniques
Aerial surveys and ground measurements are the conventional means of producing maps. These traditional map-making techniques rely on scanning and digitising processes for their updating and injection into modern decision management systems. However this type of process does not allow the fast and cheap production of maps covering large geographical areas such as entire regions.
Consequently, there is a lack of regular information updating and many maps quickly become obsolete and are therefore of little practical use. information updating and many maps quickly become obsolete and are therefore of little practical use. Furthermore, statistics provided in 1987 by a United Nations Secretariat Survey showed that more than 40% of the World’s land surface is still not covered by 1:100 000 scale maps. Almost 50% is not covered by 1:50 000 scale maps, and 80% is not covered by 1:25 000 scale maps.
The Beneﬁts of Space-Based Monitoring
Because of its ability to provide fast, up-to-date information and wide spatial coverage, spaceborn imagery is being used more and more by the mapping industry. Optical remote-sensing data products are used to produce space maps on a wide range of scales to serve many different needs. These include 1:50 000 scale (one of the standard scales) map compilation and map updating, as well as Digital Terrain Models (DTMs) based on stereoscopic imagery. Space maps are geocoded products, annotated in the same way as traditional maps.
Although optical remote-sensing information is widely used particularly in remote regions, it has two major limitations:
• excessive cloud cover often precludes its utilisation, and
• the location accuracy of individual points within a space map is sometimes not sufﬁcient for some applications.
The Contribution of ERS SAR
The Synthetic Aperture Radar (SAR) instrument carried by the European Remote Sensing satellites ERS-1 and ERS-2 has proved extremely valuable in developing markets dependent on largecoverage maps on scales ranging from 1:1 000 000 to 1:50 000. A large archive of SAR data has been constructed since the launches of ERS-1 and ERS-2 in July 1991 and April 1995, respectively, and this database is continually being updated with new acquisitions.
The SAR is an active instrument that produces images under all weather conditions by analysing the echoes (Cband and VV polarisation) transmitted from the satellite and backscattered by the Earth’s surface. An ERS SAR scene covers an area of 100 km by 100 km and has a high geographical location accuracy. Because of the speciﬁc interaction between the radar wave and the ground surface, the information content of SAR images is different from that of optical images.
Ground visibility is much improved in the latter thanks to the radar’s cloud penetration, while the topographic features stand out clearly as a result of the SAR’s oblique viewing angle. Both images have been post-processed and geo-referenced by CEGN (Cellule d’Etudes en Geographie Numerique), the geographic research department of the French armament agency.
The synergy between ERS SAR and optical images, for both cartographic map generation and updating, increases the ability to extract thematic information. Thematic and topographic maps compiled using ERS SAR data allow one to detect and identify speciﬁc features such as hydrographic networks and structures, which are particularly important in geomorphologic analysis and geology.
In regions such as the tropics, where optical satellite information is either unavailable or not usable due to excessive cloud cover, space maps generated from SAR data are a precious tool, and sometimes represent the only solution. SAR images can be a unique source of information for compiling highresolution space imagery at a continental scale, and for providing highly valuable thematic information as the sensor can also discriminate between a wide variety of land-cover types.
The capabilities of ERS for large-coverage image mosaicing and thematic mapping worldwide are well-established. They are based on state-of-the-art processing techniques which allow improvement of both the radiometric and geometric quality of the data. For example, the speckle induces alterations in the radiometric resolution of SAR data. This effect, which is inherent in the SAR system and due to the coherent nature of the SAR signal, gives a noisy effect in the images .
Filtering techniques applied on the SAR image can reduce this noisy effect, and thereby enhance image quality. The example shows how multi-temporal ﬁltering techniques have been applied to SAR data by the French company SERTIT, specialised in image processing and GIS systems, in order to reduce the speckle and thus tobe able to use enhanced radarimagery as an input to Geograﬁcal Information System. The information content of ERS temporally ﬁltered imageis highly valuble in this context and facilitates the interpretation.
ERS SAR image maps can be exploited for various types of applications:
• cartography: for map compiling and updating taking advantage of the thematic information provided by the radar for broad applications and its capabilities to extract topographic information.
• localisation: to detect or identify control points and localise targets on the ground surface, as a complement to GPS measurements for remote areas.
• rectiﬁcation: to rectify old or inaccurate maps, space maps generated using remote-sensing data with inaccurate localisation accuracy, or even to rectify a Digital Terrain Model (DTM).
In addition, ERS SAR data products can be used for topographic mapping as Digital Terrain Models can be provided using techniques such as radargrammetry and interferometry (INSAR). Radargrammetry allows one to generate DTMs using stereopairs of radar images with different viewing angles, as for optical stereo imaging.
INSAR is based on the combination of two ERS SAR images acquired with slightly different geometrical conﬁgurations. Using INSAR, highly accurate DTMs can be produced, depending mainly on the stability of the observed surface over time with respect to radar signal phase and atmospheric effects that may affect the phase information during the acquisitions.
Under favourable conditions, the altimetric accuracy achievable can be a few metres.
Differential interferometry allows one to measure surface movements with a sensitivity of the order of a few centimetres over large surfaces. This technique can be applied for subsidence monitoring and in the observation of active volcanoes, earthquakes and faults.
Moreover, interferometry can be used to extract thematic information from ERS SAR data for land-use mapping. By computing the correlation of the SAR signal between the two images, valuable thematic information is obtained. The interferometric correlation, or coherence, mainly depends on radar wave interactions with the target, and its temporal stability. Such derived imagery is used as an additional channel allowing the generation of multi-band radar products.
(Source – http://www.esa.int/esapub/br/br128/br128_2.pdf )