Bulletin of the Seismological Society of America, Vol. 87, No. 3, pp. 745-754, June 1997 Moho Dip and Crustal Anisotropy in Northwestern Nevada from Teleseismic Receiver Functions by Xiaohua Peng and Eugene D. Humphreys Abstract Receiver functions are derived from teleseismic waves recorded during the 1988 to 1989 PASSCAL Basin and Range passive-source seismic experiment in northwestern Nevada. A velocity model involving both a planar dipping Moho and crustal anisotropy is needed to explain the radial and tangential motions of the ob- served Ps conversions. An arrival-time difference often observed between radial and tangential Moho Ps conversions suggests an anisotropic crust. The Ps conversions are large and indicate a major discontinuity under the area. The particle motion directions for most of the tangential components change sign between South Ameri- can events and events from the other two source areas (Japan and Tonga regions), providing good evidence for a Moho dipping approximately to the north. Also, the Ps conversions from the Moho follow direct P arrivals by about 3.2 sec under the southern part of the array, 3.4 sec under the southern central part of the array, and 3.7 sec under the northern part of the array, indicating a Moho that varies in depth from about 26 to 31 kin. A velocity model with the Moho dipping -9 ° in a nearly northerly direction and an anisotropic crust with a split time of -0.25 sec and a fast axis of --130 ° best explain these observations. Introduction Analysis of teleseismic receiver functions has been par- ticularly useful in large-scale structural studies using body waves to determine average crustal thickness (e.g., Burdick and Langston, 1977; Langston and Isaacs, 1981; Hebert and Langston, 1985) because the timing and amplitude of Ps conversions provide good constraint on the locations and velocity contrasts of major crustal and upper mantle discon- tinuities. In addition, both dipping interface (Langston, 1977; Zhang and Langston, 1995) and anisotropic effects (McNamara and Owens, 1993) may be seen with receiver- function analysis. Wave polarity of the tangential component from several backazimuths may be used to separate the ef- fects of planar dipping interface from anisotropy, allowing the interface dip direction and magnitude to be estimated. In this article, a crustal model that includes a dipping Moho and anisotropic crust is developed to explain the timing, am- plitudes, and polarization of radial and tangential motions of Ps conversions from the Moho. The 1988/1989 PASSCAL passive-source experiment was chosen to coincide with the center of the 1986 PASSCAL active-source Basin and Range experiment (Fig. 1). Most previous investigations of this area resulted in 1D interpre- tations of the crustal structure (Klemperer et aL, 1986; All- mendinger et al., 1987; Benz et al., 1990; Hawman et al., 1990; Randall and Owens, 1994), though Holbrook (1990) inferred laterally varying structure including a dipping Moho. We use the teleseismic data to examine the crust in the context of the previous work. Compared to the active- source experiments, the information provided by teleseismic Ps conversions is more sensitive to the Moho. We find that the model of Holbrook (1990) accounts well for the major features seen in the teleseismic waveforms, though in order to fit the timing and large Ps arrivals observed in the data, the Moho velocity contrast and the dip angle need to be increased. By Moho we refer to the interface where P-wave velocity increases from 6.5 to about 7.9 km/sec; an interface in this area at a depth of about 36 krn, with a velocity contrast of 7.9 to 8.3 km/sec, also has been called the Moho (e.g., Thompson et al., 1989). The difference is largely semantic-- most workers agree that the layer between the two possible Mohos was constructed by igneous activity in that it has been involved with reconstructing the crust. Assuming that the deeper of the two possible Mohos is roughly flat and hori- zontal (Thompson et al., 1989), our findings indicate that thickness of the intervening layer varies as a function of location, as generally suggested by Priestly et al. (1982) for this area. McNamara and Owens (1993) present a model of azi- muthal shear-wave velocity anisotropy based on the tele- seismic arrivals to a portion of the area covered by Hol- brook's model. Their focus is on shear-wave splitting of the Ps conversion from the crust-mantle boundary that indicates 745