Levees provide vital functions for water delivery and flood protection. However, they present unique challenges for seismic design because their great length makes engineering evaluation of stability at closely spaced regular intervals impractical. Accordingly, relatively broad, empirically-driven risk assessment tools have the potential to serve as effective screening tools. We are undertaking a large data collection and synthesis effort to support the development of such tools, with the initial focus being on levee performance from the 2007 M_w6.6 Niigata Chuetsu-oki earthquake in Japan. Naturally, ground shaking is a key variable in this process, so the reliable estimation of ground shaking hazards fromseismic networks is an essential element of the case history analysis. We postulate thatdirect application of Krigingtechniques can produce biased ground motion estimates due to variable site conditions. Accordingly, we apply Kriging to residuals of ground motion prediction equations (GMPEs), which remove the average site effect. The resulting maps of residuals can be readily applied with the GMPE to produce ground motion maps that properly reflect spatial variations of geologic conditions.The proposed procedure produces ground motions near levees that are lower in some areas than those produced bydirect Kriging.

Ground motions at the foundation levels of structures differ from those in the free-field as a result of inertial and kinematic interaction effects. Inertial interaction effects tend to produce narrow-banded ground motion modification near the fundamental period of the soil-structure system, whereas kinematic effects are relatively broad-banded and concentrated at high frequencies. Kinematic interaction effects can be predicted using relatively costly finite element analyses with incoherent input or simplified models. The simplified models are semi-empirical in nature and derived from California data. These simplified models are the basis for seismic design guidelines used in the western United States, such as ASCE-41 and a pending report published by NIST. We compile some available data from building and ground instrumentation arrays in Japan for comparison to these two sets of models. We demonstrate that the model predictions for the sites under consideration are very similar to each other for modest foundation sizes (equivalent radii under about 50 m). However, the data show that both approaches overestimate the transfer function ordinates relative to those from Japanese data. This indicates that the semi-empirical models currently in use are conservative relative to these data sets. We speculate as to possible causes for the observed discrepancies.

We are working towards developing risk assessment tools for levees to identify conditions that correlate with ground failure rates. Our initial data set is from the 2007 M_w6.6 Niigata Chuetsu-oki earthquake in Japan. Liquefaction-induced ground failure is a major source of levee damage in this case, so groundwater elevation is expected to be a critical factor affecting damage locations. What we seek is the water level in or beneath the levees themselves along the full length of the study region, which encompasses approximately 110km of levees. The available data includes river water levels (generally below the levee toe) and groundwater levels within levees measured from boreholes. Our approach, which is applied along the full length of the study region, is first to establish the river water elevation (RWE) both at the time of the earthquake and at the time of subsurface exploration in the levees, and second to establish the differential between levee ground water elevation (LGWE) and RWE at the exploration time. Each step presents challenges due to sparse and incomplete data. Once the above relations are established, the LGWE is taken as the sum of RWE on the earthquake date and the differential. Comparing the result to damage reports indicates that levee segments with LGWE higher than the levee base elevation have approximately nine times higher damage rates than those with deeper ground water.

A fragility model for seismic deformations of levees was developed in a separate study using case history data from the Shinano River region of Japan (SRJ). In that model, levee fragility was shown to be principally related to ground motion intensity, geomorphology, and ground water level relative to the levee base. Our objective in this manuscript is to demonstrate the applicability of the developed fragility models for geotechnical conditions along urban levees in the Central Valley region of California (CVC). For this purpose, we compare SPT penetration resistance data (in the form of energy- and overburden-corrected blow counts) between regions for common soil types conditional on geology and topography. Among the geologic categories considered, arguably the most important is Holocene flood plain deposits, which comprise 38% of investigated sites in CVC and 97% in the SRJ. Within this geological unit, we find penetration resistance data for coarse-grained soils in the SRJ and CVC study regions to be similar, whereas for fine-grained soils the CVC sediments are stiffer. For two other geological units (Holocene basin and Pleistocene), both coarse- and fine-grained deposits in the CVC are stiffer than Holocene floodplain deposits. We also considered topographical conditions (elevation, ground slope and river gradient) as alternative means for sorting the data, with the general conclusion that such indicators are less capable than geology of describing variations of penetration resistance within the respective regions. The results provide insight into the relative vulnerability of levees in the two regions for given levels of ground motion amplitude.