Abstract: The use of multi-source fusion grid real-time analysis products to interpolate missing station observation data can maintain the continuity of precipitation data.Selecting precipitation observation data from 39 faulty stations in Huludao from 05:00 on August 20,to 08:00 on August 24,2024 as the testing source,and using correlation analysis and other methods to evaluate and test the interpolation station data of multi-source fusion precipitation grid real-time products.The results showed that the proportion of real-time interpolation station data with precipitation observation errors within ±2 mm · h^-1 was 83.12%.The multi-source fusion of real-time interpolation station data can accurately reflect the precipitation intensity of Huludao,achieving the expected goal.Among them,the interpolation results of Jianchang Bashhan Station are relatively the best,while the interpolation results of Jianchang Tangshenmiao Station are relatively the worst; The grading test results show that the interpolation results for precipitation intensity below 20 mm · h^-1 are good,with 86.34% of the results having an error within ±2 mm · h^-1.As the hourly precipitation intensity increases,the actual interpolation results for hourly precipitation deteriorate,and the interpolation results for extreme precipitation above 50 mm ·h^-1 are relatively small.