Surface subsidence induced by underground coal mining, together with the associated damage to surface infrastructure and a range of ecological impacts such as hydrological disturbance and vegetation degradation, collectively hampers the sustainable development of coal resource-based cities worldwide. What remains unclear, however, is how discontinuity density influences mining-induced surface subsidence (mining subsidence), which may render conventional analogy-based analysis methods inadequate for interpreting the anomalously low surface subsidence observed in the Ordos coalfield in China. In this study, we demonstrate that the stiffness of the overburden, represented by the uniaxial compressive strength of intact rock (\(\:{\sigma\:}_{ci}\)), is not a sufficient indicator for subsidence estimation, and the discontinuity density may play a more influential role. To substantiate this claim, a series of UDEC models with varied discontinuity density and rock stiffness were built and calibrated, and the modeling results were compared with practical cases to assess their validity. The results indicate: (1) increasing discontinuity density leads to a systematic shift in the dominant overburden failure mode from intact rock deformation to block sliding, thereby reducing the reliability of \(\:{\sigma\:}_{ci}\) as an indicator for subsidence estimation; (2) the applicability of \(\:{\sigma\:}_{ci}\)-based analogy methods is conditional, being reliable primarily when discontinuity densities are comparable; and (3) higher discontinuity density reduces the width-to-depth ratio required for critical mining, which helps to explain the delayed and attenuated subsidence observed in deep coal mining areas of Ordos, China. The study provides useful insights for subsidence prediction and mining design, and contributes to a better understanding of subsidence mechanisms relevant to the sustainable development of coal resource-based cities.
