Early Detection, Permanent Monitoring and Documentation of Critical Locations at the Surface in Mining Areas

Abstract

[EN] Being responsible for the legacy risks of the underground hard coal mining in Germany, RAG is developing a remote sensing-based monitoring workflow for the early detection and long-term monitoring of ground movements in areas affected by near-surface mining in the southern Ruhr region, Saar region and Ibbenbueren. The workflow to detect surface fractures like sinkholes is based on high resolution airborne imagery and simultaneously recorded ALS data. As a part of RAG's overall concept in the field of multi-sensor monitoring of ground movements this workflow has been steadily improved and will be further on. The complete monitoring area that is covered by the workflow is about 415 km². This paper first outlines the procedure of detecting ground movements by remote sensing. RAG uses an approach with annual simultaneous multispectral image and laser scanning flights. Due to the often small scale phenomena of sinkholes of only a few meters in diameter, a high accuracy of the sensor technology as well as all preceding processes like the determination of control points and reference surfaces is necessary. The process of accuracy assessment is done using continuous GNSS measurements for tie points and the projection centers of the aerial imagery as well as classic photogrammetric stereoscopic measurement. The derivation of geospatial products such as high resolution ortho imagery and digital elevation models will be covered as well as the processing of these datasets to reveal information necessary for risk analysis. The approach is based on the detection of subsidence structures in single date elevation models by pattern recognition as well as the calculation of differences between those datasets. The processing is followed by a step of data enrichment from different mining related data and stored in a specially created database. This enriched data source is provided through a modern GIS application which enables the employees to do the risk analysis and to document their findings. Ongoing research projects are evaluating the future use of artificial intelligence and machine learning to further automate the workflow.