Subsidence determination by aerial photogrammetry

Description

The demand for extensive exploitation of natural resources has led to an increase in underground mining activities. After the extraction of material the overlying strata settle to fill the empty spaces causing cavings and crevices. Consequently, subsidence may occur on the ground surface. The effects of surface ground subsidence have caused serious problems to society, including the loss of human lives, extensive property damage, disruption of communications, as well as environmental concerns. Monitoring of subsidence can provide important information about both magnitude and trends. Therefore, effective preventive methods must be developed to control the subsidence phenomena. Among other surface monitoring methods, photogrammetry offers a useful tool for accurate monitoring of mining subsidence. In this thesis, the causes of subsidence, as well as some interrelated factors influencing the amount and rate at which subsidence occurs, are presented. An overview of the various monitoring methods – physical and geometrical – is also given. The examination of the advantages and limitations of the different approaches shows why, why and how photogrammetry can play a major role in a monitoring system. Depending on the purpose and accuracy of monitoring, four photogrammetric techniques are investigated in this study, which focusses on their applicability in highly mountainous terrain. These include comparison of contour lines, determination of displacements by mathematically defined object points, determination of displacements by natural points and actual ground displacements from model coordinate differences. The identification of discrete natural points is based on a ‘cross-identification’ procedure where the advantages of stereo-vision are fully utilized. Even though the first method is purely analogues, an error analysis of the uncertainty of the contour lines gives an indication of where it can be applied. Digital elevation model techniques and the application programs of the analytical plotter AP-2C (O.M.I.) are combined to study subsidence of mathematically defined object points. Problems encountered on steep slopes are examined. In the third technique, the performance of two bundle adjustment programs with self-calibration are studied, while the displacements are computed as coordinate differences between two epochs. Due to extreme topographical conditions of the mining site, surveying and resurveying of control – and check points is quite problematic. The need to know the displacements with respect to the same reference datum led the author to develop a fourth photogrammetric technique, presented in this thesis. Displacements are determined in the model space and then transferred to the ground via four elements of the absolute orientation of one epoch only. Finally, an example of a study area is given, with statistical tests applied to evaluate the usefulness of the photogrammetric monitoring system.