Abstract: Using hourly ground observation data from 910 stations in Heilongjiang Province from 00:00 on 1 January 2020 to 00:00 on 25 December 2024 as the benchmark,the national-level hourly 10 m wind gridded live analysis product with a 5km resolution was evaluated by utilizing tools such as the METEVA validation tool and bilinear interpolation methods.The results indicate that this product exhibits low mean error,absolute error,and root mean square error within Heilongjiang Province,maintaining good consistency.Overall,the product demonstrates good applicability in Heilongjiang Province.In terms of spatial error distribution,the accuracy is relatively lower in the western plains and southeastern mountainous regions,while in the other regions,the accuracy for errors within ±2 m·s^-1 exceeds 85%.This may be attributed to the complex topography in the southeastern mountains and the significant variations in underlying surface properties of the western plains.From the perspective of annual variations,wind speed (wind force grade) generally exhibits negative deviations,particularly pronounced from April to May,while it performs relatively better in winter,with wind vectors displaying similar characteristics.The wind direction tests show better results in spring compared to summer.In terms of diurnal variations,the negative deviation of wind speed is larger during the daytime (08:00-18:00) and smaller at night.The smallest error is observed for winds of grade 0-1,while the negative deviation increase with winds for grade ≥2,and the diurnal variation amplitude becomes significantly larger with winds for grade ≥ 7.

Key words: Gridded wind field analysis products, Station observation, Hourly, Verification and evaluation