Abstract: The extended spatial autocorrelation (ESPAC)-based microtremor exploration(natural-source surface wave exploration) technology has been extensively used in shallow formation exploration owing to its simplicity, efficiency, and accuracy.However, the imaging effect of dispersion energy extracted based on the ESPAC method is unsatisfactory in practical applications.In particular, different observation array arrangements influence the extraction of dispersion curves from collected data.By investigating the imaging principle of the ESPAC method, this study conducted the simulation experiment of natural-source microtremor recording through ambient noise simulation.It compared the differences in dispersion energy under various dominant frequency distributions of wavelets.Moreover, it quantitatively analyzed the influence of different station arrangements and acquisition durations on the imaging quality of dispersion energy.The comparative study reveals the imaging patterns of the ESPAC method in shallow surface exploration.The ESPAC method can maximize the imaging quality of dispersion energy in the fundamental mode while considering both efficiency and exploration costs.The results of this study were applied to engineering application cases to further verify the simulation results.