Abstract:

Taking the concentrations of lead in the monitoring and analysis data of the road dust in Longyan city as an example, we made spatial interpolation study for pollutant concentration using three types of representative interpolation methods in geostatistics, inverse distance weighted (IDW), radial basis function (RBF) and ordinary Kriging (OK). By validating and appraising the resu
lt, we compared the three interpolation methods. Among the total 56 samples, 44 road dust samples were used for interpolation, and the rest 12 samples were used for validation. Compared with those of other interpolation methods, the ME of OK is -0.129 6, the RMSE of OK is 24.19. OK is apparently better than IDW and RBF for describing the error parameters. The results of comparison indicate that OK shows the best result in characterizing spatial distribution trend of pollutant concentration; IDW and RBF interpolate methods show less accuracy for the spatial distribution of lead elements.

Key words: spatial interpolation, environment monitoring data, statistical processing