Enhancing Deep Crustal Structure Imaging: An Efficient Seismic Tomography Method Combining First Arrivals and Later Phases

Volcanoes play a crucial role in shaping the Earth's morphology through tectonic processes. Studying volcanic structure using imaging technologies, especially in the deep crust, is essential for gaining deep insights into the Earth's internal structure and geological history. However, conventional first-arrival-based seismic tomography often falls short in capturing fine deep crustal structural features, particularly in complex volcanic terrains, orogenic belts, and crust-mantle transition zones.

The limitations of primary P-waves and S-waves include the uneven spatial distribution of seismic stations, limited coverage of earthquake events, and, in particular, insufficient penetration depth of ray paths. These limitations hinder the reliable imaging of deep volcanic structures. Regional seismic later phases, such as Moho-reflected PmP waves, crustal reflected SmS waves, various crustal converted phases, and refracted waves, provide valuable information on the deep crustal structure, including the Moho discontinuity and uppermost mantle. These later phases undergo diverse medium response processes along different propagation paths, effectively compensating for the insufficient ray coverage and weak deep constraint capability of commonly used first-arrival data.

This study¹⁻⁵ presents an efficient integrated imaging method that combines primary seismic first arrivals and multi-type later phases extracted from dense regional seismic network observations and deep seismic sounding experiments. The approach involves:

1.Precise phase picking combining manual and automated techniques, and optimization of the traditional ray tracing algorithm for complex layered crustal media.
2.Hierarchical weight assignment rules formulated to assign weights to different seismic phases based on their reliability and sensitivity to subsurface structures.
3.Multi-parameter joint inversion constraints incorporating both velocity perturbation and seismic anisotropic characteristics into the inversion process.

Thus, our method effectively improves the identification capability of low-velocity anomalies in the mid-lower crust, depicts the spatial morphology and burial scale of deep magma storage chambers, and clearly traces the vertical extension trends of volcanic magma transport conduits. The inversion results reveal extensively distributed low-velocity anomalous zones that are genetically closely related to deep magmatic thermal activities and reveal directional crustal anisotropic characteristics formed under the joint control of long-term regional tectonic stress fields and continuous magma intrusion processes.

The later-phase data have been successfully applied in typical volcanic distribution areas and adjacent tectonic transition zones, such as those in Asia and Europe. This method can provide solid and reliable deep structural constraints for systematic research covering volcanic magmatic evolutionary history, regional continental dynamic evolution mechanisms, volcanic geological disaster early warning, and quantitative risk assessment.

The high-resolution crustal structure obtained through the later-phase joint imaging technique holds enormous interdisciplinary development potential. The quantitatively calculated crustal seismic wave velocity values, anisotropic fast velocity directions, and medium physical property parameters can be closely matched and systematically cross-analyzed with multiple types of geochemical research data as well. Thus, we can build an interdisciplinary research bridge connecting deep crustal geophysical exploration and surface geochemical mechanism analysis. Moreover, this bidirectional mutual verification and complementary interpretation mechanism can effectively help solve the multi-solution problem inherent in independent seismic structural interpretation, and also compensate for the inability of traditional geochemical research to constrain the deep spatial occurrence state of magmatic materials.

With stable technical performance and wide environmental adaptability, this improved method is expected to have broad cooperative application prospects in combination with volcanic geology, geochemistry, petrology, and other related disciplines.

References
1.Li, S., Sun, A., Li, T., Tong, P., Fang, L., An, Y., Zhang, Y., Zhao, P. & Yang, F. Constructing and training of a deep learning dataset for PmP waves in the southeastern Tibetan Plateau. Earth Sci. 51(3), 1169–1181 (2026). doi: 10.3799/dqkx.2025.128
2.Sun, A. & Zhao, D. Anisotropic tomography beneath Northeast Tibet: Evidence for regional crustal flow. Tectonics 39, e2020TC006161 (2020).
3.Sun, A., Zhao, D., Gao, Y., Tian, Q. & Liu, N. Crustal seismic imaging of Northeast Tibet using first and later phases of earthquakes and explosions. Geophys. J. Int. 217, 405–421 (2019).
4.Sun, A., Zhao, D., Ikeda, M., Chen, Y. & Chen, Q. Seismic imaging of southwest Japan using P and PmP data: Implications for arc magmatism and seismotectonics. Gondwana Res. 14, 535–542 (2008).
5.Zhao, D., Todo, S. & Lei, J. Local earthquake reflection tomography of the Landers aftershock area. Earth Planet. Sci. Lett. 235, 623–631 (2005).