We explore here the benefits of using constraints from seismic tomography in gravity data inversion and how inverted density distributions can be improved by doing so. The methodology is applied to a real field case in which we reconstruct the density structure of the Pyrenees along a southwest-northeast transect going from the Ebro basin in Spain to the Arzacq basin in France. We recover the distribution of densities by inverting gravity anomalies under constraints coming from seismic tomography. We initiate the inversion from a prior density model obtained by scaling a pre-existing compressional seismic velocity Vp model using a Nafe-Drake relationship : the Vp model resulting from a full-waveform inversion of teleseismic data. Gravity data inversions enforce structural similarities between Vp and density by minimizing the norm of the cross-gradient between the density and Vp models. We also compare models obtained from 2.5D and 3D inversions. Our results demonstrate that structural constraints allow us to better recover the density contrasts close to the surface and at depth, without degrading the gravity data misfit. The final density model provides valuable information on the geological structures and on the thermal state and composition of the western region of the Pyrenean lithosphere.

中文翻译：

---

使用压缩地震速度模型作为结构相似性约束的比利牛斯山脉三维重力异常反演

我们在这里探讨了在重力数据反演中使用地震层析成像约束的好处，以及如何通过这样做来改进反演密度分布。该方法应用于一个真实的现场案例，在该案例中，我们沿着从西班牙的埃布罗盆地到法国的阿尔扎克盆地的西南-东北横断面重建了比利牛斯山脉的密度结构。我们通过在地震层析成像的约束下反演重力异常来恢复密度分布。我们从通过使用 Nafe-Drake 关系缩放预先存在的压缩地震速度 Vp 模型获得的先验密度模型开始反演：Vp 模型由远震数据的全波形反演产生。重力数据反演通过最小化密度和 Vp 模型之间的交叉梯度范数来加强 Vp 和密度之间的结构相似性。我们还比较了从 2.5D 和 3D 反演获得的模型。我们的结果表明，结构约束使我们能够更好地恢复接近表面和深度的密度对比，而不会降低重力数据的失配。最终的密度模型提供了关于比利牛斯山脉岩石圈西部地区的地质结构以及热状态和成分的宝贵信息。