In summary, the initial portion of three component local P waves can be used to constrain subsurface shear-velocity structure in a similar way to teleseismic P-receiver functions, only at much higher temporal resolution and smaller spatial scale. More detailed structure can be achieved via modeling of local P and S waves involving whole waveform inversion or backprojection methods (Chiu and Langston, 2009, 2011). Instead, the algorithm presented in this study is straightforward. The shear velocity just beneath the free surface can be readily estimated from the ratio UR/UZ of the P wave after ray parameter p is determined (equation 3). Of course, the depth extent of subsurface shear velocity determined this way depends on the source duration of vertical P. For a local earthquake with duration of 0.1 s (M 3 earthquake), the shear velocity estimated from UR/UZ of P is actually the average velocity β from free surface to depthh = 0:1 s × β. For NEHRP site class C that corresponds to a depth extent of 50 m. With additional constraints from noise H/V ratio, the subsurface shear-velocity structure can be resolved better.