The accurate estimation of subsurface reservoir properties from seismic data is a key task of geophysical exploration, yet it remains a challenge due to the inherent ill-posedness of the inverse problem and the complex, multi-scale nature of geological heterogeneity. Recent advances in deep learning and data-driven methodologies offer promising potential, providing powerful tools to extract valuable quantitative information from seismic observations.

This special issue aims to showcase cutting-edge research at the intersection of geophysics and machine learning, focusing on intelligent inversion and prediction techniques for reservoir characterization. We invite contributions that present innovative algorithms and frameworks to estimate critical properties such as impedance, porosity, lithology, and fluid content. Topics of interest include, but are not limited to:

(1) Physics-informed neural networks and hybrid inversion schemes;

(2) Deep learning for high‑resolution reservoir property modeling;

(3) Unsupervised and self-supervised learning for reservoir prediction;

(4) Generative models and uncertainty quantification in seismic inversion;

(5) Multi-modal data integration (seismic, well logs, geology) for enhanced modeling;

(6) Seismic inversion or prediction with global or probabilistic optimization algorithm;

(7) Seismic rock-physics modeling or inversion aided by machine learning;

(8) Advanced model-driven seismic inversion addressing ill-posedness.