Publication Date
2002
Keywords
radon, uranium, groundwater, bedrock, GIS, Kriging, IDW
Abstract
This study describes approaches to create surface maps of radon in groundwater based on measurements of radon (222Rn) in drilled bedrock wells at unevenly distributed sites and uranium bedrock maps from the South East of Sweden, the Östergotland county (N 58°14’ – N 58°56’ and E 14°53’ – E 16°06’), see figure 1. Geostatistical techniques of inverse distance weighted (IDW), kriging and cokriging were compared in terms of their interpolation power and correlation between the produced radon in the water layer and the bedrock uranium layer. The goal of these analyses and calculations is to improve our understanding concerning the factors influencing the transport of radon. Therefore, these interpolation techniques were investigated by optimizing parameters that are used in the specific interpolation. Using the IDW interpolator method at fixed radius enabled us to determine the linkage or search distances for auto correlation, and linkage between radon in water and bedrock. This method showed good agreement with the cokriging method when using uranium concentration as a secondary variable. Good interpolation layers (with least root mean square errors RMSE=232) were obtained by kriging. However, the kriged radon surface showed poor correlation with bedrock uranium layers. The best radon in water layer that match with uranium in bedrock layer was produced using IDW interpolator (RMSE=377, using all points). The correlation coefficient (R2) is 0.5 while for the kriging method the best correlation is R2= 0.1. A compromise between the two approaches is demonstrated.
BYU ScholarsArchive Citation
(2002)
"Spatial correlation between radon (222Rn) in groundwater and bedrock uranium (238U): GIS and geostatistical analyses,"
Journal of Spatial Hydrology: Vol. 2:
No.
2, Article 5.
Available at:
https://scholarsarchive.byu.edu/josh/vol2/iss2/5