Simulation of abandoned mining induced surface movements for estimating DInSAR detection limits

Abstract

Underground mining activities very often cause changes to the surface. Beside the strong influences of active mining, former mining activities with still existing cavities in the underground do influence the surface as well, causing a risk for people and infrastructure. The main risk is induced by sudden falls of the surface. In contrast to surface deformations induced by active mining these deformations are difficult to predict. The risk justifies the monitoring of ground movements in historical mining areas. One technique evaluated is Differential Interferometric Synthetic Aperture Radar (DInSAR). The main difference to other applications of DInSAR is, that location, subsidence rate and spatial extent can just roughly be estimated using related historic information about former mining activities. In addition, the expected character of the subsidence complicates the separation between deformation signal and atmospherical artifacts. In order to get a first idea about the limits of the technique, circular shaped deformation models with different extents and maximum vertical displacements have been modeled and integrated into the DInSAR processing chain. The final results have been evaluated qualitatively and quantitatively in order to get an idea about the detection limits of DInSAR using ERS-1/2 data. The qualitative assessment shows that subsidence cones of circular shape with radii below 400 m cannot be reliably identified in a single interferogram in the investigated case. In the quantitative assessment the highest relative RMS errors between model and result were found for subsidence cones with small radii and small maximum vertical displacements. Furthermore it could be shown that the use of multiple interferograms by applying stacking techniques leads to a better detection limit of the DInSAR method.