The Crandall Mine collapsed in August 2007 and resulted in the death of 6 miners. The collapse induced surface subsidence is visible by satellite geodesy. We processed data from ALOS satellite acquired before and after the collapse to quantify the subsidence. Our InSAR results show a localized oval shaped (1000 x 500 m2) pattern of subsidence with a maximum vertical displacement of 29 cm. Profiles across the subsided area show a steep V-shaped pattern and slight asymmetry. We use these profiles for two-dimensional modeling of the surface deformation.We first model the underground collapse (sudden roof-floor closure) using an elastic halfspace model with a horizontal tensional dislocation with a negative displacement. However, we find poor agreement between the elastic model solution and the InSAR observations, mainly because the elastic solution is characterized by a broad subsidence pattern. The plane strain assumption in our two-dimensional model does not influence this result.Modelling a combination of collapse and normal faulting reduces the misfit, in particular leads to a more narrow subsidence trough (Lu et al., 2009). The motivation behind the partitioning of normal faulting is the observed asymmetry in the subsidence pattern. Furthermore the seismic data supports the concept of a secondary subsidence source, as motion on a steeply dipping fracture (normal fault). This motion can have been stimulated by stress changes associated with the cavity collapse itself (Pechmann et al., 2008). However, in the elastic halfspace model geometry, either the normal fault does not intersect the collapse region, or much higher partitioning of faulting would be required, than allowed by the focal mechanism decomposition (Ford et al., 2008). This raises the question if the large misfit of the elastic collapse model is really explained by the missing contribution of normal faulting, or if other physical processes are neglected?Similar to underground collapse, surface settlement occurs above tunneling, as the tunnel walls move inward in response to the lithostatic pressure (leading to volume loss of the tunnel). In order to explain the lateral extend of surface settlements, one of these models considers the materials internal friction angle (Loganathan and Poulos, 1998). We apply this two-dimensional model, modifying the geometry from a circular shape to an elliptical, to account for the flat shape of the collapse strata. Our best fitting model results provide a good fit to the narrow subsidence trough, and shows reasoable model parameter values. Although our simple model does not explain all the observed features (in particular not the asymmetry), we find it more suitable than elastic halfspace models to explain the localized deformation pattern.