This work introduces a Bayesian approach for the joint inversion of gravity and vertical gravity gradient data, designed to exploit the capabilities of the new quantum gravimeters developed in the FIQUgS Horizon Europe project. The method integrates absolute gravity, gradient observations, and geological prior information to generate coherent three-dimensional density models with well-defined structural boundaries. Continuous density values and discrete geological labels are estimated simultaneously through a Markov Random Field formulation and a Simulated Annealing–Gibbs sampling optimization scheme. The approach is demonstrated using data from the FIQUgS 2024 field campaign in Lisbon, where a Differential Quantum Gravimeter (DQG) was deployed to detect shallow archaeological voids in a challenging urban setting. The survey provided high-resolution measurements, with gradient data proving particularly valuable for enhancing sensitivity to near-surface features and mitigating the influence of regional mass effects. The joint inversion reconstructed a clear low-density anomaly consistent with a buried tunnel, achieving residuals comparable to observation noise. These results highlight the benefits of incorporating gradient information and show the potential of quantum-enabled gravimetry for high-resolution near-surface imaging.
Capponi et al. (Mon,) studied this question.