The high density of the teleseismic data and the well established thick sedimentary structure of the Bohai Bay Basin enable us to study in detail the crustal and upper mantle structure of the region. The imaged Moho exhibits a significant lateral undulation, from >40 km in the west Taihang Mountain area to ∼30 km in the interior of the east Bohai Bay Basin (Figure 2d). The depth uncertainty is confined to within 2 km from bootstrap testing based on various crustal velocity models, even without considering the thick sedimentary structure (Figure 2c). The thinner crust with small thickness variations in the basin area confirms the previous speculation that the crust might have been thinned at the same time the lithosphere was thinned in eastern China during late Mesozoic to Cenozoic [Gao et al., 1998; Zhang et al., 2004]. This result is consistent with the present-day high surface heat flow of the region [Hu et al., 2000]. Moreover, our receiver function imaging experiences indicate that our newly proposed wave equation-based depth migration technique is not only efficient in structural imaging using Ps converted phases, but also applicable to the topside reflected PpPs multiples. Compared with Ps phases, PpPs multiples are more useful in studying crustal structure, especially the Moho of thick-sediments-covered basin areas.
 Possibly due to the interference of sediment reverberations or because of the high complexity of the real structure, we cannot obtain a clear image at depths of ∼150 km–350 km. However, the structure of the upper mantle transition zone is well imaged (Figure 3c). The 410 is sharp and consistent, and displays a simple near horizontal feature around 422-km depth. The 660 also displays small undulations about the average depth of 683 km beneath the major part of the Bohai Bay Basin, although some local structural complexity and possible depressions are present at the northwestern basin-mountain boundary and the southeastern coastal area. For most parts of both discontinuities, depth errors of <5 km are obtained in a same way as that for the Moho but with sedimentary structure correction. The smooth topography of both discontinuities imaged here and observed previously [e.g., Ai and Zheng, 2003, Figure 3c] within the basin area may be either attributed to a relatively smooth temperature distribution or indicative of the insignificance of the combined influence of temperature and chemical components, such as water and/or iron [Smyth and Frost, 2002; Van der Meijde et al., 2003], on the topography of upper mantle discontinuities. Despite the significant variations at both edges of the profile, the thickness of the mantle transition zone also appears to have only small variations of ≤5 km in the basin's interior where the thickness is determined most reliably (marked by thick dashed lines in Figures 1 and 4) . The transition zone beneath the basin is on average ∼10–15 km thicker than the global average (Figure 4). According to mineral physics studies on phase transformations in the olivine system [Ito and Takahashi, 1989], this implies a generally cold mantle environment consistent with the cooling of a flat–lying slab at the bottom of the upper mantle under eastern China as imaged by tomography [Fukao et al., 2001; Zhao, 2004]. Mineral physics calculations also suggest that at low temperatures multiple phase transformations occur, due to the co-existence of non-olivine components such as garnet with olivine, at most distinct depths around the bottom of the upper mantle [Vacher et al., 1998]. This low-temperature regime of mineral phase changes has been invoked to interpret the presence of multiple 660-km discontinuities at subduction zones [e.g., Simmons and Gurrola, 2000; Ai et al., 2003], and is plausibly responsible for that observed near the southeastern coastal area of the study region. In contrast, geochemical data [Xu et al., 1998; Xu, 2001] and surface heat flow measurements [Hu et al., 2000] suggested a rather warm present-day upper mantle, likely associated with the Mesozoic-Cenozoic lithosphere reactivation and thinning in this region. Such a discrepancy indicate that the shallow and deep upper mantle in the study region probably have not reached a thermal equilibrium, and the Cenozoic lithospheric process and magmatism in the eastern China continent might have originated within the shallow upper mantle, possibly no deeper than the middle of the transition zone.