Key points• A 3D P-wave velocity model of the Earth's crust and upper mantle of the Caspian region was constructed.• Deep structural and velocity heterogeneities controlling regional seismicity were identified.• Low velocity upper mantle zones show a spatial correlation with moderate and strong earthquake epicenters.• The results provide a geophysical basis for improving seismic hazard assessment and seismic zoning in the Caspian region.

Rapid and reliable seismic monitoring requires accurate waveform inference, together with robustness to noise, incomplete sensing, and unstable predictions. This study investigates a quality-controlled, waveform-centered, AI-assisted framework for seismic event detection, P- and S-phase picking, graph-aware inter-station refinement, and rapid hazard-related characterization. The framework includes optional DAS, MEMS, and high-rate GNSS branches; however, the primary empirical validation is based on real waveform-centered IRIS records from the Almaty seismic region, not on a fully synchronized multimodal field deployment. The dataset includes seven seismic stations, HHZ waveforms sampled at 100 Hz, 219 seismic events, 1260 event traces, and 240 s P-centered windows from 1 January 2023 to 31 December 2024. Optional auxiliary branches are evaluated through controlled branch-availability, reduced-input, fallback, and stress-test scenarios. Under the standard-condition benchmark, the proposed framework achieved a precision of 0.941, recall of 0.932, F1 score of 0.936, false-alarm rate of 0.051, detection latency of 173 ms, and P- and S-pick mean absolute errors of 31 ms and 54 ms. Under controlled low-SNR testing, it retained an F1 score of 0.846. The findings support waveform-centered, quality-controlled monitoring, while broader cross-domain and fully synchronized multimodal validation remain necessary.

This study addresses the need for better constraints on the deep crustal and subcrustal structure of weakly seismic platform regions – areas that remain poorly understood in terms of deep geodynamics, yet are of high industrial and infrastructural importance. We present a geophysical framework for seismic hazard assessment in the Caspian region of Kazakhstan, based on integrated analysis of deep seismic sounding and seismometric data. The aim is to identify patterns in the distribution of structural and velocity heterogeneities within the crust and upper mantle and to assess their relationship with present-day seismicity. The dataset includes both retrospective and newly acquired deep seismic sounding (DSS) profiles, refraction seismic records, and tomographic sections. Two-dimensional P-wave velocity models were constructed by solving the inverse kinematic problem for refracted waves and subsequently integrated into a three-dimensional volumetric velocity model using the Kriging method. Digital structural maps of felsic and mafic layer thicknesses, Moho depth, and the thickness of the low-velocity subcrustal mantle were compiled. Orogenic domains show increased felsic layer thickness and pronounced Moho depth variations, while weakly seismic platform areas exhibit a more uniform thickness ratio between the major crustal units. The spatial distribution of low-velocity subcrustal mantle zones correlates with epicentral locations of moderate-to-strong earthquakes, supporting the use of this mantle parameter as a key indicator of present-day geodynamic instability. These results provide a geophysical foundation for seismic zoning of the Caspian region and for assessing seismic hazard at oil and gas facilities in Kazakhstan.