**1. Introduction**

Landslides are common natural hazards in many parts of the world, often triggered by increased pore pressure after heavy rainfall. Due to climate change, more intense rainfall events are expected and the frequency of destructive landslides may increase [1–3]. In order to evaluate hazards associated with landslide-prone hillslopes, several modeling approaches can be used to determine critical precipitation events that may lead to slope failure. Common limit-equilibrium models tend to overestimate the factor of safety as a measure of slope stability in complex geometrical setups [4]. The concept of a local factor of safety in Coulomb stress-field based finite element models is one approach used to overcome this limitation [5]. Studies applying limit-equilibrium and continuous finite element models have partly found good agreemen<sup>t</sup> between the results for stability analysis

**Citation:** Moradi, S.; Heinze, T.; Budler, J.; Gunatilake, T.; Kemna, A.; Huisman, J.A. Combining Site Characterization, Monitoring and Hydromechanical Modeling for Assessing Slope Stability. *Land* **2021**, *10*, 423. https://doi.org/10.3390/ land10040423

Academic Editor: Enrico Miccadei

Received: 29 March 2021 Accepted: 14 April 2021 Published: 15 April 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

e.g., [6]. However, variation between models depends strongly on the methods used and the application scenarios [7,8].

The incorporation of spatial variability of soil and rock types into models for slope stability evaluation is crucial to assess the structural and hydrological state of a hillslope e.g., [9]. Slope morphology and spatial distribution of the material properties affect the slope failure potential e.g., [10,11]. Especially bedrock topography and soil depth were identified as important factors with respect to slope hydrology and stability [9,12,13]. Geophysical characterization e.g., [14–16] and monitoring e.g., [17,18] are increasingly used for structural and hydrological assessment of landslide-prone hillslopes. Such geophysical studies often combine multiple methods to study the subsurface structures and seismics and electric resistivity tomography (ERT) are the most commonly used e.g., [19,20]. Seismic refraction is particularly useful to identify lithological layers and slip surfaces e.g., [21,22], whereas ERT is able to provide information on the water content distribution in the subsurface e.g., [23,24] by using the correlation between bulk electrical resistivity and saturation e.g., [25,26]. Compared to drilling methods or point sensors, geophysical measurements usually provide information with a higher spatial resolution at lower cost and are only minimally invasive [27]. The combination of geophysical methods allows collection of complementary information and can be analyzed using data fusion methods [28]. Typically, supporting laboratory studies are used to improve geophysical monitoring concepts and to evaluate innovative geophysical methods, such as self-potential measurements, for detecting critical hydrological conditions [29,30]. An extensive review of geophysical monitoring methods for failure-prone hillslopes is given in Whiteley et al. [31]. In addition, the development of cost-effective sensor networks for monitoring soil water content and slope movement has gained momentum in recent years, which allows us to bridge the gap between costly boreholes and extensive geophysical monitoring and surveying e.g., [32,33].

A few case studies have combined hydrogeological and geomorphological site characterization, geophysical monitoring and hydromechanical modeling. For shallow landslides in pyroclastic soils, water content, derived from electric resistivity profiles, were combined with statistical modeling using a cellular automaton to derive a relationship between the factor of safety of a hillslope with in situ measurable quantities [34]. For the La Clapiere landslide in France, vertical electrical sounding was used to obtain the underground structure used in a geomechanical model [35]. An extensive ERT survey was combined with groundwater measurements and meteorological data to study the groundwater dynamics of a translational landslide and to develop a conceptual model [36]. To assess the slope stability of the Brzozowka landslide near Cracow (Poland), ERT monitoring was combined with drilling and laboratory tests to improve a stability analysis in which critical conditions for slope stability were derived from simulations for extreme precipitation events [37].

In this study, we present a slope stability analysis of a failure-prone hillslope in Germany based on the combination of thorough site investigation and monitoring with a hydromechanical finite element model. The site investigation includes a geological, geomorphological, hydrological, and geophysical characterization and soil water content and groundwater monitoring using a sensor network as well as geoelectric measurements. We introduce a work flow incorporating the different data sources into the model setup and for validating the model results. The data used in this study are partly obtained from former studies at the study site, and include (i) groundwater level [38], (ii) borehole logs [38], (iii) laboratory tests for soil hydraulic and mechanical properties [38,39], (iv) slope movement and (v) precipitation [40], as well as recent (vi) seismic refraction surveys, (vii) electric resistivity tomography, (viii) soil water content monitoring network and (ix) a digital elevation model. The parameterized, constrained and verified hydromechanical model was subsequently used to study potential hazardous precipitation events. We compared the results to the observations in the field and discuss how the developed approach could be extended towards a site-specific early warning system.
