Visual interpretation of ASTER satellite data, Part 1: Geologic mapping in the Cradle of Humankind World Heritage Site

Abstract

Since the first earth observing satellite was launched in 1972, remote sensing has become a powerful tool in the arsenal of geoscientists. This satellite became known as Landsat 1 and carried the Multispectral Scanner (MSS) delivering imagery at a spatial resolution of 80, and spectral resolution from blue to near infrared. Ongoing satellite and sensor development to the end of the century produced the Landsat Thematic Mapper (TM) with improved spatial and spectral resolution, as well as the SPOT series of satellites delivering the highest spatial but limited spectral resolution. These developments culminated in the SPOT 4 (1998) and Landsat Enhanced Thematic Mapper (1999) sensors. While Landsat ETM in particular provided much improved spatial and spectral resolutions, on the basis of which a large amount of geoscientific remote sensing was conducted world wide, the data did not provide adequate spectral and spatial sensitivity to be optimally effective for geological mapping at the local scale. On 18 December 1999 the Terra platform was launched, carrying five remote sensing instruments, including ASTER (Advanced Space borne Thermal Emission and Reflection Radiometer). ASTER consists of three separate instrument subsystems, each operating in a different spectral region, and using separate optical systems. These are the Visible and Very Near Infrared (VNIR) subsystem with a 15m-spatial resolution, the Short Wave Infrared (SWIR) subsystem with a 30m-spatial resolution and the Thermal Infrared (TIR) subsystem with a 90m-spatial resolution. ASTER effectively offers an improvement on Landsat MSS, Landsat TM, Landsat ETM+ and SPOT spectral and spatial resolutions. Given the paucity of published research on geological remote sensing at the local scale in South Africa, and particularly on the use of ASTER for geological mapping in South Africa, it is imperative that the value of ASTER be investigated. This article reports on the improved detail and scale achieved in the mapping of litho-stratigraphy, geological structures and mining-related features by the visual interpretation of processed ASTER images. ASTER imagery obtained from the EOS website was subjected to a range of image enhancement and analysis techniques including colour composites, band ratios, normalised difference indices, regression and decorrelation, in order to obtain optimal visual interpretability. Eight images thus obtained could be used for visual analysis, and it became evident that litho-stratigraphy, faults, fracture zones and elements of the regional seam system, as well as remnants of mining activities, were readily identifiable. Some of these were in accordance with the most recent and accurate geological map of the area, but many of them had apparently not been mapped. These features were annotated and were verified by field checks. In all cases the accuracy of detection and location from satellite imagery was confirmed on the ground. The improved detail and accuracy obtained by visual interpretation of processed ASTER satellite data for mapping a section of the Cradle of Humankind World Heritage Site demonstrated the potential value of this data for a variety of other geoscience applications. It appears that the improved accuracy can be ascribed jointly to the higher spatial and spectral resolution provided by ASTER data.