Depth Constraints on Crustal Density Variations by Comparing Isostatic Gravity and Velocity Models in the Los Angeles Region, Southern California

High-resolution velocity and gravity data were collected along the two transects of the Los Angeles Regional Seismic Experiment to determine crustal structure. Along LARSE I (Seal Beach to Barstow) and LARSE II (Santa Monica bay to the southern Sierra Mountains), Bouguer gravity data indicate two distinct geophysical terranes separated by the San Andreas fault, with the Mojave Desert characterized by gravity values 30-40 mGal lower than those over the Transverse Ranges. The gravity gradient separating the two terranes is centered over the San Andreas fault and is significantly steeper along LARSE I than along LARSE II. One model attributes the steep gravity gradient on LARSE I to lateral density variations in upper and middle crustal rocks, allowing for the presence of a crustal root beneath the San Gabriel Mountains [Godfrey et al., in press]. Another model attributes the gravity gradient to an abrupt step at the crust-mantle boundary (no crustal root; [Zhu, 2000]). To test whether a step at the Moho beneath the San Andreas fault is required by gravity data, we predict gravity from a three-dimensional crustal velocity model of the upper 10 km and compare it to a filtered version of the Bouguer gravity field (isostatic residual gravity field). In general this comparison shows that the isostatic residual gravity field and predicted gravity from the tomography are similar. The gravity predicted from the 3-D Vp model matches well with the isostatic gravity along both transects. There is disagreement across the San Andreas fault between the predicted gravity from the upper10 km of the 2-D high-resolution crustal velocity model and the isostatic gravity, which could be caused by anisotropy in the Pelona schist. Anisotropy would affect the LARSE results more than the gravity data or the 3-D Vp model. The maximum anisotropic change in Vp of 0.8 km/s, caused by foliation of the Pelona schist [McCaffree Pellerin and Christensen, 1998], is sufficient to explain the difference in the gravity signature, although more ray-tracing modeling is needed to determine if an anisotropic effect can be detected in the high-resolution LARSE data. Thus, most of the San Gabriel Mountains gravity high can be explained by higher-density upper crustal rocks, and a density step at Moho depths [Zhu, 2000] is not required by the gravity data.