“…Alternatively, gravity and free‐board modeling are used to model global and regional patterns of density variations in the upper mantle (Molnar et al, ; Mooney & Vidale, ), and compositional variations in mantle density can be calculated if thermal structure of the lithospheric mantle (LM) is known. One approach involves subtracting gravitational effect of crustal heterogeneity from the observed gravity field (Artemjev et al, ; Artemieva et al, ; Herceg et al, ; Kaban & Mooney, ; Kaban et al, , ; Mooney & Kaban, ; Yegorova & Starostenko, ), and the other includes mantle density calculation from lithosphere buoyancy in regions that are close to isostatic equilibrium (Artemieva, ; Artemieva & Vinnik, ; Cherepanova & Artemieva, ; Mooney & Vidale, ). In both cases the results are critically dependent on the quality and resolution of the crustal model (Herceg et al, ; Kaban & Schwintzer, ), which necessitates the development of high‐quality regional seismic models for the crust (Artemieva & Thybo, ; Cherepanova et al, ; Grad et al, ; Kennett et al, ; Tesauro et al, ; Xia et al, ).…”