Abstract: [Objective]With the increasing demand for hydrogen energy utilization, underground hydrogen storage (UHS) has become a prominent research topic in recent years. The diffusion coefficient of H2 in water under high-temperature and high-pressure conditions is critical for quantitatively characterizing H? migration behavior in reservoir pore spaces and estimating H2 leakage through caprocks. However, previous studies have primarily focused on H2 diffusion in water at ambient conditions. [Methods]In this study, the dissolution and diffusion processes of H2 in aqueous solutions were quantitatively observed in suit using micro-laser Raman spectroscopy within transparent high-pressure quartz capillaries. The diffusion coefficients of H2 in water were experimentally determined under conditions of 10~30MPa and 298.15~393.15K. [Results]The results indicate that the diffusion coefficient of H? in water increases with rising temperature. At 20MPa, when the temperature increased from 298.15K to 363.15K, the diffusion coefficient rose by approximately 211%. The relationship between the diffusion coefficient and temperature can be fitted using the Speedy-Angell power-law equation: D=23.572?10-9?[(T/213.54)-1]2.021. In contrast, the diffusion coefficient of H2 in water is less affected by pressure, showing a slight decreasing trend as pressure increases. At 363.15K, when the pressure increased from 10MPa to 30MPa, the diffusion coefficient decreased by approximately 4.8%. Using the measured diffusion coefficients combined with an empirical formula for effective diffusivity, the H2 leakage through caprocks was estimated. The results demonstrate that as caprock thickness increases, H2 leakage gradually decreases, the diffusion rate significantly slows, and the time required for H2 to diffuse out of the reservoir is prolonged. [Conclusion]Therefore, in practical UHS projects, deep geological formations with low temperatures and thicker caprocks should be prioritized. This study provides essential parameters for quantitatively characterizing H? migration behavior and diffusion flux in UHS, while also offering a reference for the design and optimization of UHS schemes.