This PIRE research project, carried out by a US team from Colorado School of Mines, the University of Mississippi and Mississippi State University in collaboration with dozens of international partners, focused on advancing the sustainability of earth dams and levees, and other critical geotechnical infrastructure. Four prominent goals were addressed, namely: (a) advancing characterization through geophysics and remote sensing; (b) using remote sensing to investigate sustainable earth dam and levee interaction with the natural environment; (c) advancing health assessment and performance prediction using data driven modeling; and (d) understanding, measurement and communication of failure risk across social and political contexts.

The project made it possible for US researchers to participate in the unique full scale IJKdijk levee experiments being conducted in the Netherlands, as well as the Livedijk levee monitoring on seaside levees in the Netherlands. Participation in these and access to rich data enabled significant advancement by the US researchers. Our testing of the IJKdijk levee allowed extensive multi-geophysics and remote sensing investigation of the levee. This led to novel advances in passive seismic monitoring, acoustic emission and electrical self potential monitoring of internal erosion.  

The project also enabled considerable advances in geophysical and remote sensing based assessment of levees along the Mississippi river, advances that can be implemented into practice for future health assessment of US levees. Interferometric synthetic aperture radar (from satellites) was successfully used to detect subsidence on Mississippi levees, and polarimetric synthetic aperture radar was successfully used to detect slump slides.

The project made considerable strides in health assessment and performance prediction using data driven modeling. Highlights of advances include: (1) A study of the full-scale Netherlands IJkdijk pore water pressure data set revealed that the progression of internal erosion could be tracked using inversion; (2) successful development and demonstration of a time lapse structural joint inversion technique to improve the estimation of internal erosion from geophysical data; (3) the development and implementation of a Bayesian technique to improve the estimate of levee reliability using sand boil and pore water pressure data collected during flooding events; and (4) numerous improvements in predicting anomalies using machine learning techniques. Field testing at a full scale sea levee in Colijnsplaat, Netherlands allowed for the novel and successful implementation of passive seismic monitoring wherein internal erosion within an active levee was illuminated using the vibrations created by passing vehicle traffic and sensed by an array of permanently deployed sensors.

An extensive amount of real estate is protected from floods by levees in the US. A detailed study of one data set in St. Louis county, Missouri, showed that commercial properties in the area protected by 500-year levees are not price discounted compared to properties that are not in a floodplain, suggesting that public perception of levees is high. The study also found that the selling prices of properties in levee-protected areas tend to be higher than those in a floodplain without such protection.

The project produced dozens of journal and conference proceedings publications. The results were disseminated by dozens of project participants throughout the US and internationally.

Last Modified: 12/30/2018
Modified by: Michael A Mooney