Constraining the depth of the lithosphere-asthenosphere boundary in tectonically complex regions using locally adjusted lithological forward models and seismic velocities

Abstract. In this study, we employ 1D migration of S-to-P receiver functions to determine the lithosphere-asthenosphere boundary (LAB) depth beneath 41 seismological stations in the Pannonian Basin (Hungary). Our approach incorporates forward models of seismic velocity profiles tailored to local lithospheric structures, allowing for an improved constraint on LAB depths in comparison with global velocity models. We systematically evaluate the impact of crustal structure, mantle composition, temperature variations, and partial melt on seismic velocities. Global velocity models, while effective for large-scale studies, have significant limitations in resolving lithospheric-scale structures due to their coarse parameterization. Our forward models incorporate variations in sedimentary layer velocities, Moho depth, and LAB-associated velocity reductions, leading to LAB depth estimates that differ from those derived using the IASP91 model. Notably, 1 vol.% partial melt below the LAB results in a pronounced velocity decrease, whereas metasomatism and volatile-bearing phases may cause localised velocity reductions but have limited effect on the determination of LAB depths. LAB depths obtained using local versus global velocity models reveal a strong correlation, yet notable discrepancies exist at stations with complex lithospheric structures or thick sedimentary cover, as further supported by anomalies in surface heat flow data. Our results demonstrate that locally calibrated seismic velocity models offer a more accurate representation of the lithosphere-asthenosphere transition than global models, particularly in regions with complex tectonic and thermal histories. Integrating geophysical and petrological approaches is key when investigating lithospheric structure, as well as the combined interpretation of different factors that shape seismic LAB signatures.