The use of ambient seismic noise to investigate internal changes in a tailings storage facility and to image the subsurface geology in the Cradock area of the Eastern Cape

Passive seismic interferometry is a process by which ambient noise data recorded at different seismic stations can be cross-correlated to estimate Green's functions. In the past, both surface waves and body waves have successfully been extracted by cross-correlation of ambient noise data on both regional and global scales. Recent advancements in ambient seismic noise techniques have the potential to provide new methods for subsurface imaging and monitoring. The ambient noise data processing procedure divides into four principal phases: (1) single station data preparation, (2) cross-correlation and temporal stacking, (3) measurement of dispersion curves and (4) inversion of dispersion curves to obtain 1-D shear wave profiles and computation of 2-D shear wave velocity cross-section. The purpose of this study is to investigate whether ambient seismic noise can be used to image and detect internal changes within a mine tailings dam wall and to image the subsurface geology of part of the Eastern Cape Karoo near Cradock. In the first experiment, the investigation consisted of continuous ambient noise data recordings over a period of 3 days with 20 three-component short period geophones. The geophones were deployed over a survey wall of roughly 100 m in length at the Harmony Gold mine tailings dam in Welkom. In the second experiment, the investigation consisted of data recordings over a period of 35 days. The geophones were deployed in Cradock. The first phase of the data processing procedure included de-trending, de-meaning and band-pass filtering the data. This was done to ensure that any long period trends associated with instrument glitches are removed from the data. A spectrogram was then computed to view the spectrum of frequencies in the signal and to check if the filter that was designed was able to cut off the unwanted frequencies. The horizontal and vertical components of the ambient noise data were cross-correlated and picked between sensor pairs to create surface wave dispersion curves. Subsequently, the dispersion curves were inverted to estimate the shear wave velocity of the dam wall and subsurface as a function of depth. The computed cross sections of shear wave velocity indicated a low-velocity zone between 2 and 10 m below the surface on the dam wall, this suggested that the phreatic surface is much closer to surface in this area. In the second experiment, the interpolated shear wave velocity profiles indicated that there is a layer of low velocity zone between depths 250 to 300 m below the surface. The cross-correlations were also used to compute group velocity maps from periods 1.5 seconds to 30 seconds. The group velocity maps showed various high and low velocity anomalies. The high velocity zones observed on the eastern section of the map were interpreted as evidence of dolerite intrusions. The low velocity zones observed in the western and southern sections of the map interpreted as Karoo sediments that belong to the Adelaide Subgroup which is dominated by mudstones.Thesis (MSc) -- Faculty of Science, Department of Geosciences (Geology and Geography), 202