An evaluation of airborne laser scan data for coalmine subsidence mapping

Abstract

The accurate mapping of coalmine subsidence is necessary for the continued management of potential subsidence impacts. The use of airborne laser scan (ALS) data for subsidence mapping provides an alternative method to traditional ground‐based approaches that affords increased accessibility and complete spatial coverage. This paper evaluates the suitability and potential of ALS data for subsidence mapping, primarily through the examination of two pre‐mining surveys in a rugged, densely vegetated study site. Data quality, in terms of mean point spacing and coverage, is evaluated, along with the impact of interpolation methods, resolution and terrain. It was assumed that minimal surface height changes occurred between the two pre‐mining surfaces. Therefore any height changes between digital elevation models of the two ALS surveys were interpreted as errors associated with the use of ALS data for subsidence mapping. A mean absolute error of 0.23 m was observed, though this error may be exaggerated by the presence of a systematic 0.15 m offset between the two surveys. Very large (several metres) errors occur in areas of steep or dynamic terrain, such as along cliff lines and watercourses. Despite these errors, preliminary subsidence mapping, performed using a third, post‐mining dataset, clearly demonstrates the potential benefits of ALS data for subsidence mapping, as well as some potential limitations and the need for further careful assessment and validation concerning data errors.

Acknowledgements

The authors would like to thank the following individuals for providing data, information, and advice: Mark Freeburn from AAMHatch for information on data collection, the ALS sensors and for providing the relevant metadata; Rod Doyle for depth of cover data; Richard Walsh for operational data and comments on the manuscript and Dr Alexandra Golab for her advice and comments on various drafts. The authors also gratefully acknowledge funding from the Australian Coal Association Research Program (Project C14031), which made this research possible, and the helpful comments of two anonymous reviewers.