USING A 3D SURFICIAL GEOLOGY MODEL TO MAP SHEAR-WAVE VELOCITY AND FUNDAMENTAL SITE PERIOD FOR SEISMIC SITE RESPONSE IN THE OTTAWA AND ST. LAWRENCE VALLEYS

Shear-wave velocity (V_S) and fundamental period of vibration (T_O) of surficial sediments are important parameters for predicting the intensity of ground shaking. The St. Lawrence and Ottawa valleys are partially underlain by large basins of soft Late Pleistocene clays and include several large and mid-size urban centers exposed to significant intra-plate seismic hazard. As the available Vs measurements are unevenly distributed and cover only a small part of this region, a simplified 3D geological model of the Quaternary sediments was used as ancillary information for predicting the spatial variation of V_S, V_S30 and T_O. The applied method consisted of: 1) updating the surficial geology maps; 2) classifying the surficial units with similar physical properties into three broad categories: upper sandy sediments, intermediate clayey sediments and basal glacial deposits; 3) delineating the spatial thickness of each category by way of 3D geologic modeling; 4) sorting available geophysical data with respect to each individual category; and 5) assigning statistically representative V_S values to each of the categories. A power V_S-depth function was assumed for the combined sand and clay units, whereas a constant interval V_S equal to the geometric mean was assigned for glacial sediments. The bedrock velocities were drawn from the literature for Precambrian, Paleozoic, and intrusive rocks. The final velocity values were computed as travel-time weighted averages down to bedrock for V_S and to a depth of 30 m for V_S30. T_O values were computed as the ratio between soil thickness and average V_S. The model was validated against results from detailed seismic zonation studies in Ottawa-Gatineau, Montreal and Quebec City. While the expected V_S and To prove regionally satisfactory, the V_S30 model was occasionally unable to adequately reproduce situations where the upper 30 m consisted mainly of bedrock. Regional trends of decreasing velocity toward the center of the study area were identified in both glacial and postglacial units; further investigations are needed to address this issue. Uncertainties resulting from converting the continuous model information to a 500 m raster and from the observed variability in velocity measurements are also presented.