Abstract:Based on the national multi-source analysis products by the National Meteorological Information Center of CMA and ground automatic stations observation data in 2023,the applicability of hourly temperature, relative humidity, wind and precipitation in Fujian were evaluated by the error statistical analysis,skill score and other methods.The result show that the quality of temperature product is good,the mean error is between -0.5 and 0.5℃,correlation coefficient is 0.999.The mean absolute error of relative humidity is 0.646%,correlation coefficient is 0.997,which means that it has a high accuracy.The wind product are consistent with automatic station observations,with the mean error is between -0.2 and 0 m/s,Inland areas perform better than coastal areas.The TS score of precipitation products for light rain is higher than those in other classes,and as the rainfall magnitude increases,the false alarm rates and missing report rate increase in general.For Typhoon Doksuri,the precipitation product are consistent with automatic station observations with respect to the precipitation intensity and the precipitation area,but there are deviations in the reproduction capabilities of precipitation peak .