The next large earthquake on the southern San Andreas fault is expected to produce strong ground shaking hundreds of kilometers to the north, in Los Angeles, California. This will occur because the geologic structures beneath the Earth?s surface act as a wave guide channeling earthquake waves into downtown Los Angeles. It has been shown that the ground motions one can expect could be a factor of four larger than what earthquake hazard models predict. This difference may significantly increase damages resulting from earthquakes. It should also be accounted for when planning for earthquake resilience in the greater Los Angeles area. The inaccurate estimates for ground shaking result from incomplete Earth models which do not properly represent the geologic structures that channel seismic energy. Here, the researchers build a more detailed Earth model of the region. They use recordings from a new style of seismic field experiment carried out between 2017 and 2020; it consisted of laying out more than 700 modern instruments along 10 lines across the study area. This dense geometry allows the team to image better than ever before the subterranean structures channeling earthquake energy. The researchers carry out a full suite of analyses of the dense dataset. They refine and update models for regional seismic velocity. They simulate expected ground motions accounting for the new findings. The new Earth models are made available to the scientific community by updating the Southern California Earthquake Center community velocity models; this is done using a newly developed platform for embedding high-resolution models into larger-scale regional models. Results of the study improve seismic hazard assessment in the Los Angeles area. This collaborative project also supports a female early-career scientist from an underrepresented group, and graduate and undergraduate students at Louisiana State University and the California Institute of Technology. The researchers use receiver functions to construct profiles of the structure along each of the lines, using several teleseismic events that were recorded. The three-dimensional (3-D) shear wave velocities for the basins will be obtained from surface wave analysis based on ambient noise correlation between the nodes and broadband stations that were deployed as part of the survey, and the Southern California Seismic Network stations in the region. They combine the profiles and 3-D shear wave velocity estimates by using the gravity data across the basins. They produce attenuation (QS and QP) models from ambient noise and local earthquakes. The expected result will be a 3-D Earth model for the basins with a well-defined basement interface and shear wave velocities down through the sedimentary layers obtained at a higher resolution than existing models. As a further step, the team will simulate earthquake ground motions with the new model and compare strong motions to those predicted from noise correlations. The unique geometry of the survey allows the team to try more advanced methods including level-set inversion, autocorrelation methods, and possibly full waveform inversion. This advances the application of some of these methodologies for the first time to this particular type of data set. The full suite of methodologies developed here will be applicable other major cities around the world that are located on sedimentary basins such as San Francisco, Seattle, Mexico City and Lima. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.