Abstract:
Two interpolation methods were used to interpolate the 2m air temperature merged grid analysis product in 2018 to 2380 national assessment stations.The product was evaluated by correlation coefficient,average error,average absolute error,root mean square error and accuracy.The results show that the assessment results obtained by using proximity interpolation method are slightly better than using bilinear interpolation method,and the assessment results of merged grid analysis product has certain diurnal and monthly changes.Overall,the hourly merged grid analysis product is basically the same as station observations,and has a high reference value,the correlation coefficient between them is above 0.99,the root mean square error is below 1℃,the accuracy within 2℃ is above 98%,and the accuracy within 1℃ is above 95%.The results of provincial and sub-elevation assessment show that the results of the assessment are worse with the increase of elevation,so caution should be taken when applying this product in areas with higher elevation,more complex terrain or sparse meteorological stations.The daily maximum and minimum air temperatures obtained by the hourly merged grid analysis product is also highly indicative;and the product also has good monitoring ability for high temperature process.