Jobs:Job-01020
Apply before: 6 January 2020
|
Posting:
Position: PostDoc position
Apply before: 6 January 2020
Apply online:
|
Whereas soil and root strength properties are often estimated for past landslide events using back-analyses, current methods for predicting these parameters across different regions are lacking. In particular, simple methods are needed to inform better parameterization of landslide models based on existing geodatabases and remotely sensed data. Previous work in the hydrologic sciences has successfully identified empirical relationships to establish “transfer” functions that translate readily available soils information (e.g., textural classification, organic carbon content, vegetation cover) to the hydraulic properties of interest that are more difficult to obtain (i.e., porosity, hydraulic conductivity, water retention). These transfer functions (e.g., Schaap et al., 2001) have facilitated regional and even continental scale simulations with physics-based hydrologic models. However, the transfer function approach used successfully in hydrology – relating readily available data to parameters of interest – has not been attempted for estimating geomechanical properties (i.e., angle of internal friction, cohesion, and root strength).
The primary objective of the research opportunity is to identify and develop simple or complex transfer functions for assigning geomechanical inputs into regional slope stability models. Extensive data on soil mechanical properties and root strength parameters are dispersed throughout the literature in journal articles (e.g., Schwarz et al., 2010) and various USGS reports (e.g., Mirus et al., 2016), but these have yet to be assembled into a comprehensive database for further quantitative analysis or prediction of landslide triggering. Numerous machine learning approaches have been applied in the Earth Sciences to identify empirical relations between measurements and parameters of interest and have even been used to map landslides from lidar data (Bunn et al., 2019). However, these computational methods have not been applied rigorously to explore geomechanical properties. Relationships between strength properties and existing geospatial data on soils, geology, topography, and landcover could be explored quantitatively with a variety of techniques. The proposed post-doctoral research opportunity therefore provides many avenues for innovative research to identify suitable properties and relations that can help improve the predictive power of landslide initiation models.
