Introducing a new chirp-rate-modulated generalized Stockwell transform with enhanced resolution

ABSTRACT The Stockwell transform (ST) is a widely used algorithm for time-frequency (TF) analysis of seismic signals, frequently applied in reservoir characterization and hydrocarbon detection. However, the traditional ST suffers from inherent limitations, including poor temporal resolution at low frequencies, reduced frequency resolution at high frequencies, and an inability to effectively capture transient components. These challenges arise from the fixed time window used in the ST, which imposes a single time-frequency resolution and limits energy concentration for rapidly varying signal components. To address these shortcomings, this study introduces a generalized S-transform (GST) with a dynamic, chirp-rate-dependent windowing mechanism. Our GST incorporates a novel generalized Gaussian window function with four tunable parameters, enabling enhanced adaptability to varying signal characteristics. The optimization of these parameters is guided by a dual-objective approach that combines concentration measurement and the Gini index, resulting in a significantly improved TF representation with better energy concentration and resolution. The effectiveness of the GST is demonstrated through applications to synthetic and real field data. Synthetic examples highlight the superior TF resolution of GST compared with the traditional ST and the adaptive generalized S-transform, showcasing its ability to resolve transient signal components with greater clarity. Real field data experiments further validate the GST’s capability, particularly in enhancing the visualization of low-frequency shadows, a critical indicator in hydrocarbon exploration.