Prospecting archaeological archives in South Africa through hyperspectral image processing and field spectroscopy
Domain:
geospatialRecord type:
paper
Creator:
Publisher:
fig
Host:
Presentation held at the session S6 "The archaeological perspective on the use of satellite data" at the Annual Conference of Computer Applications and Quantitative Methods in Archaeology (CAA) 2021 on 2021-06-15.
Abstract:South Africa and the province KwaZulu-Natal play an important role in prehistory of Homo sapiens. Modern human archaeological record of the Middle Stone Age (MSA) and Later Stone Age (LSA) is known to be exceptionally rich and detailed in rock shelters, although the total number of sites is very low. While open sites are abundant in Southern Africa, the stable and old surfaces (African surface and Post-African I) bear little contextual information and render dating difficult. This is not the case for the Late Pleistocene Masotcheni Formation, which covers the area from the Drakensberg Mountains in southern KwaZulu-Natal, through the Free State Province and into Eswatini. The colluvial deposits accreted at the footslopes of mountain ranges during multiple phases throughout the Late Pleistocene and are interbedded with layers of buried soils. This gives not only a valuable insight into the geomorphic history of the region and cyclical climate changes, but makes this formation also a well-datable archaeological archive populated with MSA/LSA artifacts.
We present a method to map this landscape feature through hyperspectral remote sensing and digital landscape analysis. Therefore, we sampled the spectral properties of local surface and soil profile materials in situ through field spectroscopy (250-2500 nm wavelength), yielding high resolution reflectance curves that give insight to their physio-chemical properties. Thereupon we developed spectral indices, which enable the discrimination of different surface types and applied these to VIS, Near Infrared (NIR) and Shortwave Infrared (SWIR) reflectance bands of WorldView-3 to map the formation based on its specific spectral properties. Finally, we demonstrate how we used UAV and historical aerial imagery to estimate the age and erosion rates of soil degradation affecting the Masotcheni today, thereby affecting not only the archaeological archives but also the livelihood of the modern population.
Abstract:South Africa and the province KwaZulu-Natal play an important role in prehistory of Homo sapiens. Modern human archaeological record of the Middle Stone Age (MSA) and Later Stone Age (LSA) is known to be exceptionally rich and detailed in rock shelters, although the total number of sites is very low. While open sites are abundant in Southern Africa, the stable and old surfaces (African surface and Post-African I) bear little contextual information and render dating difficult. This is not the case for the Late Pleistocene Masotcheni Formation, which covers the area from the Drakensberg Mountains in southern KwaZulu-Natal, through the Free State Province and into Eswatini. The colluvial deposits accreted at the footslopes of mountain ranges during multiple phases throughout the Late Pleistocene and are interbedded with layers of buried soils. This gives not only a valuable insight into the geomorphic history of the region and cyclical climate changes, but makes this formation also a well-datable archaeological archive populated with MSA/LSA artifacts.
We present a method to map this landscape feature through hyperspectral remote sensing and digital landscape analysis. Therefore, we sampled the spectral properties of local surface and soil profile materials in situ through field spectroscopy (250-2500 nm wavelength), yielding high resolution reflectance curves that give insight to their physio-chemical properties. Thereupon we developed spectral indices, which enable the discrimination of different surface types and applied these to VIS, Near Infrared (NIR) and Shortwave Infrared (SWIR) reflectance bands of WorldView-3 to map the formation based on its specific spectral properties. Finally, we demonstrate how we used UAV and historical aerial imagery to estimate the age and erosion rates of soil degradation affecting the Masotcheni today, thereby affecting not only the archaeological archives but also the livelihood of the modern population.
Similar
A real-world hyperspectral image processing workflow for vegetation stress and hydrocarbon indirect detectionHyperspectral field spectroscopy and SENTINEL-2 Multispectral data for minerals with high pollution potential content estimation and mappingSoil salinity prediction using a machine learning approach through hyperspectral satellite imageCoffee Leaf Plant Disease Identification through Image Processing and Machine-Learning Techniques in Ethiopia.Saving Archives Through Digitisation: Reflections on Endangered Archives Programme Projects in AfricaVee995/Image-Processing-Project