Impact of malfunctioning drainage systems on landslide initiation and propagation

Despite extensive literature on the effectiveness of drainage systems and documented cases of landslides linked to leaking and burst pipes, malfunctioning drainage systems are rarely incorporated into hazard modelling. Porewater pressure is a major control of landslides, commonly managed through the installation of drainage systems. Paradoxically, such an intervention can act as a landslide driver rather than a mitigation measure. Malfunctioning drainage systems and burst pipes may disrupt slope hydrological connectivity and, in some cases, lead to oversaturation, thereby elevating localized pore water pressure. This study aims to quantify in data scarce regions, through numerical simulations, the influence of malfunctioning drainage systems on landslide dynamics, from initiation to propagation.

The complex Hofermühle landslide in Lower Austria is an illustrative case of the interplay between natural and anthropogenic processes. Previous activities of this landslide and its proximity to a stream have led to the installation of subsurface drainage for decades to dewater the hillslope. We develop a landslide conceptual model of a mudflow event that occurred on 21 April 2013 using long-term monitoring data, residents' reports, 2D seepage and slope stability analyses, and propagation modelling (r.avaflow). The seepage and slope stability analysis demonstrates that the malfunctioning drainage scenario is the most plausible trigger of the reported landslide event.

Our results indicate that a drainage capacity of less than 40% and an antecedent malfunction of at least 68 days before failure were likely factors in the 2013 landslide. The mudflow results from sufficient water storage, localized porewater pressures, seepage emergence, and, thus, slope failure which transformed into a flow. The preliminary propagation analysis showed that a minimum volume of 100 m-3 is necessary to propagate the initial landslide mass downstream. Our findings suggest that abandoned drainage infrastructure may play a crucial role in landslide occurrences. The backward simulation is a demonstrative example of a process that may become increasingly important as projected future urbanization and associated hillslope modifications unfold.