“…However, a comparison of different continental regions calculated by different groups is challenging, in part due to fundamentally different assumptions on thermal structure of the upper mantle and on the depth range to which mantle gravity anomalies are confined. In particular, one may use thermal models constrained by surface heat flux and xenolith geothermobarometry and resolved on a 1° × 1° spatial grid (Artemieva et al, ; Artemieva & Mooney, ) or may constrain mantle temperatures from seismic tomography (Goes et al, , ; Kaban et al, ), despite that geophysical studies indicate the presence of a strong nonthermal (compositional, melt, water, and grain size) component in seismic velocity variations (Afonso & Schutt, ; Artemieva, ; Faul & Jackson, ; Godey et al, ; Lee, ; Zhu et al, ), and lateral resolution of these models is significantly less than 1° × 1° (Foulger et al, ).…”