Geodatabase of Sinkholes in the the Post-Mining Area of the Brown Coal Mine “Babina” (W Poland)

Post-mining sites are prone to complex processes related to the ceased mining and disturbance of the rock mass around the excavations. Therefore, such sites require continuous monitoring to minimize threats associated with, e.g. occurrence of often unexpected discontinuous deformations such as sinkholes. This study focuses on the development and analysis of a database of sinkholes in the former “Babina” brown coal mine in Western Poland. The mine site is located in the SW part of the complex glaciotectonic area of the “Muskau Arch”. It was subjected to long-term open-pit and shallow underground mining. The primary objective of the study is to create a comprehensive database of sinkholes, based on analysis of differential digital elevation model and derivatives of digital elevation model such as slope and hillshade maps. The structure of the database includes dependent variables such as geographical location and dimensions of sinkholes, as well as parameters representing potential causative factors including: geological, mining, geophysical and topographical characteristics (exploratory variables). It will be used to analyse and model the relationship of sinkhole occurrence with potential causative factors of their occurrence in the project no. 2021/43/B/ST10/02157.

The geodatabase was developed using ArcGIS software from ESRI, encompassing information on more than 230 identified sinkholes. Each sinkhole in the database is comprehensively described by a range of attributes. The exploratory variables include total depth of mining, distance to the first underground level, distance to shafts and adits, location of brown coal outcrops locally named gizers, proximity to coal seams (geological mining factors). Among the topographical factors the following attributes have been stored: slope of the terrain, distances to former open pits, anthropogenic lakes and waste heaps, land cover types. The geophysical data include results of gravimetric observations (anomalies in the gravitational field). Whereas, the hydrogeological data include results of underground water modelling.

The construction of the database was done by using advanced spatial data processing tools such as Map Algebra Statistics and Surface Functions, as well as extract value to feature tools. These functions were used to calculate and to extract raster values associated with location of sinkholes in addition distance tools where used to determine parameters derived from vector data that include for example database of underground working.

The dataset was subjected to a comprehensive statistical analysis, which included developing descriptive statistics encompassing histograms of the values of dependent variables (sinkhole parameters) and independent variables (factors potentially influencing the formation of deformations). An exploratory data analysis was also conducted to determine correlations between variables.

The results of the study have allowed analysing weighted spatial distribution of sinkholes in the post-ming area. The weights included parameters of sinkholes. Further research is aimed at developing predictive models with a machine learning approach. The models will be used to identify areas prone to future sinkhole formation.

The results of the study confirm the complexity of post-mining impacts and the necessity for further detailed analysis of the changes taking place in the study area.

The research has been financed from the OPUS National Science Centre projects grants no. 2019/33/B/ST10/02975 and no. 2021/43/B/ST10/02157.