Agencies and research groups engaged in studying measures for enhancing the resiliency of communities have recently placed emphasis on the need for extensive implementation of monitoring systems for rapid post-event assessment of structural integrity. Designing a monitoring system for a building requires a thorough knowledge of its potential nonlinear dynamic behavior with an associated localization of interstory drift. Extending this task across a regional scale becomes even more challenging because of the heterogeneity of the buildings inventory and the limited knowledge of the characteristics of the demand especially for sites located in the near-field of a major fault. The existing observational database of near-field ground motion records is in fact too limited to constitute a comprehensive basis for full understanding of the potential range of structural response variability at different locations near a major fault. In addition, current insight into monitoring system design typically relies on linear structural models and sensors deployed on a limited number of floors. In this context, this paper presents first results of a study that combines synthetic earthquake ground motions generated from a massively parallel regional-scale geophysics wave propagation model at frequencies of engineering interest (0-5 Hz) with nonlinear tall building models. The objective is to gain new insight into the potential impact of localization of nonlinearities in structures subjected to realistic near-field earthquakes and develop a methodology that optimizes the deployment of sensors at the building and site level. In addition to the large database of synthetic motions, available real records are also employed to compare and contrast with the trends observed using synthetic ground motions. Preliminary results confirm a tendency of the demand to localize in specific portions of the structure, especially when nonlinearities occur. The building analyses provide guidance for various sensor deployment configurations associated with different probability of error in measuring structural drifts.