Dual-scale undulation characterisation and modelling of sedimentary bauxite deposits

Abstract

Abstract Sedimentary bauxite deposits exhibit both regional-scale orebody undulations and irregular local ore–rock interfaces, which cannot be adequately represented using models constructed solely from borehole data. Taking the Aojiawan II bauxite mine in Lüliang, Shanxi Province, China, as a case study, this study proposes a dual-scale method for characterising and modelling orebody undulations using borehole data and roadway-exposure data. An ore-seam undulation coefficient is introduced to quantify the dispersion of the ore seam relative to its macro-scale trend surface, while a local interface undulation coefficient is used to describe the relative dispersion of the roof and floor interfaces. Scale-sensitivity analysis is conducted to determine an appropriate macro-scale evaluation grid. A dual-scale orebody model is then constructed by superimposing local stochastic residual fields derived from roadway-exposure data onto the macro-scale trend generated from borehole data. The results show that a 50 m × 50 m evaluation grid provides a suitable balance between data support and spatial resolution. The classification thresholds for macro-scale undulation are 24% and 26%, while the local interface undulation coefficient ranges from 23.66% to 46.78%, indicating marked spatial variations in interface geometry. Across the four independent validation profiles, the dual-scale mean model reduced the mean absolute error (MAE), root mean square error (RMSE), and maximum absolute error (MaxAE) by 41.02%–76.26%, 33.47%–72.30%, and 28.11%–40.93%, respectively, relative to the macro-scale trend model. The model integrates regional ore-seam geometry with local ore–rock interface undulations and may provide a refined geometric basis for subsequent ore–rock boundary delineation and stope-layout assessment.