Sponsor: University of Arkansas
PI: Thomas Oommen
Michigan Tech will aid in developing an empirical model to predict the probability of mudflow/rockslide and a Remote Sensing Based Decision Support System to evaluate the risk to transportation infrastructure following wildfires.
The empirical modeling will include developing a probabilistic model using logistic regression that relates the remote sensing inputs to the occurrence and non-occurrence of mudflows/rockslides. The remote sensing derived parameters for logistic regression will include soil suction, soil volumetric water content, soil temperature, soil density, predominant mineral type, clay content, etc.
The RSBDSS will be built upon the existing decision support system that is being developed for Phase 5 Geotechnical Asset Management project (PI: Thomas Oommen). Specifically, following the development of the RSBDSS, input parameters obtained from existing data or newly collected remote sensing data will be entered into the logistic regression model to determine the amount of risk associated with mudflows or rockslides following wildfire events. This RBDSS is anticipated to aid highway managers in determining the risk to transportation infrastructure, which would help in developing plans for required road closures and ensuring safety of the users of transportation infrastructure.