<p>An essential tool for the seismic monitoring and interpretation of CO₂ plumes at storage sites is the application of time-lapse/4D seismic inversion. This study proposes a solution that leverages the robust capabilities of the genetic algorithm (GA). Inspired by the principles of biological evolution, the GA is a population-based optimization method categorized within the broader framework of evolutionary techniques. This study provides a detailed methodology for employing the GA to address the direct 4D seismic inversion problem. Prior to its application, the optimization mechanics are thoroughly examined, and a systematic approach for tuning the hyperparameters necessary for effective implementation is delineated. Additionally, to reduce computational demands, a customized variant of the Gaussian representation (GR) has been developed, specifically designed for the 4D seismic inversion problem. The GR methodology presented in this study follows a straightforward yet effective workflow that diminishes computational effort while maintaining high-resolution outcomes. The optimized procedure is first applied to synthetic tests to ascertain the optimal hyperparameters. Following this, the framework of the GA is extended to include applications to a real seismic monitoring dataset obtained from the Sleipner field. The results demonstrate that the proposed workflow enhances the traditional deterministic inversion approaches. Moreover, the GA technique yields a generation of optimized solutions, thereby producing a statistical distribution of the results, which is critical for uncertainty quantification. The methodology introduced is versatile and can be implemented across a wide variety of seismic monitoring datasets.</p>