EXPRESSING UNCERTAINTY AND VARIABILITY IN GEOLOGIC MAPPING AND GEOTECHNICAL DATA FOR HAZARD EVALUATION

Natural hazard evaluations require understanding of surface geologic conditions and subsurface geotechnical characteristics, which differ substantially in the way data are obtained and in their uncertainty and variability. Federal geologic-map guidance in the United States defines spatial accuracy of points and lines plotted with GIS tools in terms of buffer zones with non-specified distances expressed in ground units (zone of confidence). Numerical values of zones of confidence are based on geologic judgment which depends on geologic complexity, terrain, exposure quality, and mapping objectives. Geologic maps are artistic representations of individual interpretations of geologic features and relationships inferred from limited observations: Geologic maps are geologic models. Surficial geologic maps generalize bedrock formations, whereas bedrock geologic maps generalize surficial deposits.

Geotechnical data are based on measurements, obtained mainly from boreholes or soundings. Geophysical methods applied at the ground surface (i.e., seismic refraction) are used for some measurements. Geotechnical data are used for stratigraphic positions and values of geotechnical parameters. Stratigraphic positions typically are depth to groundwater and depth to top of bedrock; geotechnical parameters typically are thicknesses of subsurface "layers" and their compression- and shear-wave velocities. Other parameters (e.g., unit weight, unconfined compressive strength) are measured or based on correlations with seismic velocity. Uncertainty and variability of geotechnical parameters typically are based on linear interpolation of "layers" and measurement statistics. Geologic contacts separating subsurface units fundamentally are non-planar, making the "layer" concept and linear interpolation oversimplifications of reality.

Recognition of sources of uncertainty and variability is the first step in quantifying them. Fundamental differences in the nature of geologic and geotechnical data require different quantification approaches, which must be suitable for combining to express uncertainty and variability needed for hazard mapping.