摘要
为了探索夏季空气负离子浓度的空间分布格局,以黑龙江省为研究区,选取了331个具有代表性的观测点进行空气负离子浓度观测研究,根据所测数据,采用3种不同插值方法(普通克里金法、反距离加权法、径向基函数法)对全省331个不同观测点的空气负离子浓度数据进行空间插值分析,采用平均误差、误差均方根、平均标准误差、标准平均值、标准均方根误差参数对插值结果进行对比,研究不同方法对该地区空气负离子浓度插值结果的影响。结果表明,黑龙江省空气负离子浓度的Moran’s I指数值为0.36,表明全省空气负离子浓度具有空间自相关性;通过对普通克里金法的4种变异函数模型(环形模型、球面模型、指数模型、高斯模型)对比发现,指数模型的准确度更高,更适合用于空气负离子浓度的插值;265个预测站点和66个验证站点的回归分析显示,3种不同插值方法的预测值和检验值均具有相关性,显著性水平检验R^2结果从大到小分别为为普通克里金法(0.840 33)、径向基函数法(0.803 96)、反距离加权法(0.610 16)。综合对比发现,采用变异函数为指数模型的普通克里金法对黑龙江省夏季空气负离子浓度插值结果最优,能够对全省的空气负离子浓度空间分布进行较好的预测。
To explore the spatial distribution pattern of negative air ion,the thesis used the Heilongjiang Province as the exploration area and chose 331 typical observation points to do the negative air ion observation.According to the datas we have,three interpolation methods were adopted(Ordinary Kriging method,inverse distance weighted method,radial basis function method)to do the spatial interpolation analysis of the negative air ion of 331 observation points in Heilongjiang Province.Correlation parameters were adopted,such as mean error,root mean square error,mean standard error,standard mean value,standard root mean square error,to compare the results of interpolation and study the effects of the results of negative air ion in this area that different methods brought to.The results showed that,the Moran’s I of negative air ion concentration in Heilongjiang Province was 0.36,indicated that the negative air ion had spatial autocorrelation in Heilongjiang Province.According to the comparision of four variograms models of Ordinary Kriging method(circulai model,spherical model,exponential model,Gaussian model),it was found that the accuracy of exponential model was higher and this model was more suitable for the interpolation of negative air ion concentration.The regression analysis of 265 forecst sites and 66 validation sites showed that the focast figures and examine figures of three kinds of different interpolation methods had relevance.Significance level test R2 results was Ordinary Kriging(0.840 33),radial basis function method(0.803 96)and inverse distance weighting method(0.610 16)from high to low.According to synthetic comparision,it was found that optimal semivariogram for the exponential model of ordinary Kriging interpolation results of negative air ion concentration in Heilongjiang Province were the best,and it could focast the spatial distribution of negative air ion in Heilongjiang Province well.
作者
王吉祥
杨江宁
张冬有
WANG Ji-xiang;YANG Jiang-ning;ZHANG Dong-you(Key Laboratory of Remote Sensing Monitoring of Geographic Environment,College of Geographical Science,Harbin Normal University,Harbin 150025,China)
出处
《湖北农业科学》
2018年第5期24-29,共6页
Hubei Agricultural Sciences
基金
国家自然科学基金项目(41171412)
黑龙江省自然科学基金项目(D201303)
哈尔滨师范大学博士后项目(13RBHZ03)
关键词
空气负离子浓度
变异函数
空间插值
黑龙江省
air negative ion concentration
variation function
spatial interpolation
Heilongjiang Province
