Accurate estimation of precipitable water vapor (PWV) is critical for evaluating the feasibility of millimeter- and submillimeter-wave astronomical observations. We assess ERA5 and MERRA-2 reanalysis datasets at the Nanshan 26 m Radio Telescope (NSRT) and propose a Digital Elevation Model (DEM)-weighted Inverse Distance Weighting (IDW) interpolation method that accounts for both horizontal distance and elevation difference. Compared with nearest-neighbor and bilinear interpolation, the DEM-weighted approach substantially improves PWV estimation, with MERRA-2 achieving over 26% lower RMSE than its bilinear interpolation. Seasonal evaluation identifies MERRA-2 with DEM-weighted interpolation (M2-DEM) as the most accurate, with slightly higher summer bias than ERA5 DEM-weighted IDW (E5-DEM). A 30 yr PWV climatology (1995–2024) shows a pronounced annual cycle, stable interannual variability, and no significant long-term trend. Winter minima (typically below 3 mm) define the primary high-frequency observing window, while summer maxima arise from elevated temperatures and monsoonal moisture transport. Comparisons with QTT, GBT, and Effelsberg, along with atmospheric transmission modeling, confirm NSRT’s favorable K- and Q-band transparency and partial 3 mm accessibility under optimal conditions. These results establish M2-DEM as the preferred dataset-method combination for site-specific PWV evaluation and provide a strong basis for expanding NSRT’s high-frequency astronomical capabilities.