This study, based on Boyle’s law and Torricelli’s principle, designs a self-fabricated water-column system containing air to investigate its pressure equilibrium and dynamic behavior. The experiments vary the tube length and diameter, leak-hole shape and aperture, and initial water-column height as operating parameters to simulate and measure changes in water-column height when internal and external pressures reach equilibrium. In Experiment (I), the apparatus is used to measure atmospheric pressure, and its accuracy is validated through field measurements at Alishan, with deviations of less than 3% from the actual value. Experiment (II) investigates the influence of external vibrations on system stability, revealing that pressure equilibrium is disrupted only at specific frequencies. Fourier transform analysis confirms that this behavior results from a resonance effect. Based on Bernoulli’s principle, the relationship between water-level rise and airflow velocity is derived and subsequently applied in Experiment (III) for wind-speed measurement. Finally, an artificial intelligence–based machine learning model is introduced to predict the leakage rate, further enhancing analytical accuracy. Grounded in fundamental physical principles, this study integrates theoretical derivation with experimental verification to reveal the dynamic behaviors hidden within the water-column system. The findings not only demonstrate intriguing physical phenomena but also highlight the system’s potential for practical environmental-measurement applications.