Search

Measuring 3-D mining-induced displacements is essential to understand mining deformation mechanisms and assess mining-related geohazards. In our previous work, we proposed a method for estimating 3-D mining-induced large displacements with the surface deformation along the radar line-of-sight (LOS) direction derived from a single amplitude pair (SAP) of synthetic aperture radar (SAR) using the offset tracking (OT) procedure (hereafter referred to as OT-SAP). The OT-SAP method effectively reduces the strict requirements on SAR data of the previous OT-based methods for 3-D mining-induced displacement retrieval. However, OT-SAP is not robust to errors in the LOS deformation, due to the lack of redundant observations. In this paper, we present an alternative approach (hereafter called AOT-SAP) to OT-SAP. The AOT-SAP method involves estimating the 3-D mining-induced large displacements with OT-derived 2-D deformation observations along the LOS and azimuth directions from an SAP of SAR, instead of just the LOS deformation in the OT-SAP method. Consequently, more redundant observations are incorporated in the AOT-SAP method compared with the previous OT-SAP method. The theoretical analysis and experiments based on both simulated and real data sets suggest that AOT-SAP can effectively improve the accuracies of the estimated 3-D displacements compared with the OT-SAP-estimated ones. [ABSTRACT FROM AUTHOR]

This paper presents a novel method for modeling and robustly estimating the time-series 3-D mining-induced large displacements from a single imaging geometry (SIG) synthetic aperture radar (SAR) amplitude data set using the offset-tracking (OT) technique (hereafter referred to as the OT-SIG). It first generates multitemporal observations of 3-D mining-induced displacements from the single-geometry SAR amplitude data set with the assistance of a prior model. Then, a functional relationship between mining-induced time-series 3-D displacements and the multitemporal 3-D deformation observations generated is constructed. Finally, the time-series 3-D displacements are robustly estimated based on the constructed function model using the M-estimator. The proposed OT-SIG provides a robust and cost-effective tool for retrieving time-series 3-D mining-induced large displacements, relaxing the basic requirement of the traditional method that at least two different viewing geometries’ SAR data are needed. Finally, we tested the proposed OT-SIG with descending TerraSAR-X SAR amplitude data set over the Daliuta coal mining area in China. The results show that the root-mean-square errors (RMSEs) of OT-SIG-estimated time-series displacements are about 0.22 and 0.11 m in the vertical and horizontal directions, respectively. These RMSEs are around 5.7% and 10.9% of the maximum in situ deformation measurements in the corresponding directions, which can meet the accuracy requirements of practical applications. [ABSTRACT FROM AUTHOR]

It is crucial to locate underground goafs (i.e., mined-out areas) resulting from coal mining and define their spatial dimensions for effectively controlling the induced damages and geohazards. Traditional geophysical techniques for locating and defining underground goafs, however, are ground-based, labour-consuming and costly. This paper presents a novel space-based method for locating and defining the underground goaf caused by coal extraction using Interferometric Synthetic Aperture Radar (InSAR) techniques. As the coal mining-induced goaf is often a cuboid-shaped void and eight critical geometric parameters (i.e., length, width, height, inclined angle, azimuth angle, mining depth, and two central geodetic coordinates) are capable of locating and defining this underground space, the proposed method reduces to determine the eight geometric parameters from InSAR observations. Therefore, it first applies the Probability Integral Method (PIM), a widely used model for mining-induced deformation prediction, to construct a functional relationship between the eight geometric parameters and the InSAR-derived surface deformation. Next, the method estimates these geometric parameters from the InSAR-derived deformation observations using a hybrid simulated annealing and genetic algorithm. Finally, the proposed method was tested with both simulated and two real data sets. The results demonstrate that the estimated geometric parameters of the goafs are accurate and compatible overall, with averaged relative errors of approximately 2.1% and 8.1% being observed for the simulated and the real data experiments, respectively. Owing to the advantages of the InSAR observations, the proposed method provides a non-contact, convenient and practical method for economically locating and defining underground goafs in a large spatial area from space. [ABSTRACT FROM AUTHOR]