The Global Navigation Satellite System (GNSS) provides high spatiotemporal resolution data for monitoring precipitable water vapor (PWV), a critical parameter for weather forecasting, climate research, and astronomical site selection. This study evaluates the accuracy of ERA5 and NASA reanalysis datasets by comparing them with GNSS-derived PWV data at the NSRT site. Statistical methods, including root mean square error (RMSE), mean absolute error (MAE), and mean bias error (MBE), were employed to assess their performance across seasonal and monthly scales. The results show that ERA5 and NASA datasets exhibit high consistency with GNSS-derived PWV data, effectively capturing seasonal variations. ERA5 demonstrates superior accuracy during winter with lower RMSE values (e.g., 0.56 mm in January), while NASA performs better in summer months under high water vapor conditions. Systematic biases are observed in both datasets, emphasizing the need for localized bias correction to enhance their reliability.This study highlights the complementary strengths of ERA5 and NASA datasets and underscores the importance of GNSS data in validating and improving reanalysis models. The findings provide valuable insights for water vapor monitoring, atmospheric studies, and the optimization of astronomical site selection.