基于协同克里金空气湿度空间插值研究

Air Humidity Based on CoKriging Interpolation

以东北三省(吉林、黑龙江、辽宁)38.90°-52.97°N,119.70°-132.97°E区域为试验区,利用东北三省1970-2009年气象站的空气相对湿度观测数据的40年平均值,采用普通克里金和协同克里金空间内插方法,估计试验区的空气湿度分布情况。结果表明,空气相对湿度和降水量具有显著相关关系。当选取其中46个观测值时,引入了降水量作为协变量的协同克里金的插值方法比采用46个观测值的普通克里金的插值方法的均方根误差(RMSE值)降低了7.66%;当采用同样的协变量(77个降水量站点数据),46个湿度观测值的协同克里金比77个湿度观测值的协同克里金的精度更高,均方根误差降低了20.58%。因此,引入了降水量作为协变量的协同克里金插值方法提高了插值精度,并且使用协同克里金插值方法可以适度减少观测站点数量。

Abstract： The three Provinces of Northeastern China located between 38.90°-52.97°N and 119.70°-132.97°E were taken as the studying area. Forty-year mean relative air humidity observed in meteorological stations from 1970 to 2009 was used. The ordinary Kriging and CoKriging interpolation methods were used for interpolation. The results were compared with each other in terms of its accuracies. Results showed that the air relative humidity and precipitation was significantly correlated. CoKriging had more accurate results than that of ordinary Kriging. It is indicated that the root-mean-square error produced based on CoKriging decreased by 7.66% in comparison with the ordinary Kriging, using the 46 observations.CoKriging with 46 observations got a higher accuracy than that with 77 observations using the same secondary variable （77） with the relative reduction of RMSE of 20.58%.Therefore, CoKriging interpolation method with precipitation as a covariate can improve accuracy.