Comparative Analysis of Model-Based Acoustic Seismic Inversion Methods

This study presents a comparative analysis of two model-based seismic inversion methods: Maximum Likelihood (ML) and Sparse Spike (SS). A synthetic seismic cube was created using a Point-Spread Function (PSF) and the Butterworth wavelet to evaluate the performance of these techniques under controlled conditions. Metrics such as Mean Absolute Percentage Error (MAPE) and Pearson correlation coefficient were applied to assess the results. The findings revealed that the ML method consistently achieved better accuracy, showing lower errors and higher correlation with the reference model, while SS exhibited limitations in handling high-frequency variations and produced noisier results. These results emphasize the suitability of ML for acoustic impedance inversion, especially in scenarios characterized by structural complexity and low-frequency seismic data. Future research should focus on optimizing the SS algorithm to broaden its applicability in geophysical contexts.