ASSESSMENT OF LANDSLIDE HAZARD POTENTIAL IN PARKER COUNTY, NORTH TEXAS USING FDM-BASED FLAC AND GIS SOFTWARE TECHNOLOGIES

This study describes a new methodology for predicting landslides using empirical evidence from recent slope failures within Lower Cretaceous sedimentary rocks in north Texas. Typical failure surfaces lie within the Paluxy Formation of the Trinity Group near its contact with the overlying Walnut Formation of the Lower Fredricksburg Group. Direct shear testing was conducted on samples collected from tworecent landslides (Willow Park and Pottsboro) in order to obtain cohesion and friction angle data for slope stability modeling using FLAC/Slope software.

A parametric study involving Factor of Safety (FS) calculations was conducted for both varying slope angles and groundwater table depths. Regression analyses of FS data from FLAC/Slope were performed to obtain representative equations for FS determination, in which the slope angle and groundwater depth are independent variables. Cohesion and friction angle were statistically insignificant due to lack of data variability. Rock layer thicknesses were assumed to be constant wihtin the area, due to shallow dip. Additional cohesion & friction angle data were used to introduce data variability for slope stability modeling. Regression analysis of this modeling generated a representative equation for FS determination with cohesion, friction angle, slope angle and groundwater depth as independent variables. The two sets of equations were imported into ArcGIS to generate Landslide Hazard Zonation Maps for the area around Willow Park landslide in eastern parker County, Texas. The maps indicate that slope failures are likely to occur when groundwater table depth is less than eleven meters. Based on these maps, the areas most vulnerable to slope failures lie along slopes greater than or equal to 14.5 degrees.

The results show that the equation without cohesion & friction angle may be used for site-specific cases only, where these data may be limited. The equation with all four variables overestimates FS, thus may be refined with addition of other variables such as layer thicknesses and actual dilation angle and tensile strength data for each rock layer.