Optimizing Soil Sampling with Information Entropy at Heavy-Metal Sites

Knowledgeof the spatial distribution of heavy metals is indispensablefor successful risk analysis of contaminated sites. The common practiceis to obtain soil samples for spatial interpolation through site investigation,which generally involves preliminary and detailed surveys. In thisstudy, we propose an information entropy-based site investigation(IESI) method in which an optimal design step is implemented to guidesoil sampling at the detailed survey stage. Two types of informationentropy (i.e., relative entropy and Shannon entropy) are used to designthe optimal sampling strategy. The results show that, within the IESImethod, relative entropy is superior to Shannon entropy in guidingsoil sampling. Combined with ordinary kriging, the IESI method outperformsconventional surveys for hypothetical and actual heavy metal-contaminatedsites as it can identify new polluted and clean areas. For quantitativecomparisons, the IESI method coupled with ordinary kriging, logarithmicordinary kriging, and universal kriging with linear and quadratictrends can improve the interpolation accuracy by 16-43% atthe actual heavy metal-contaminated site. Upon further examinationof the IESI method, informative sampling points are mainly distributedaround the polluted areas identified by the preliminary survey withsoil pollution probabilities between 0.75 and 0.95. This work providesan effective tool for delineating the spatial distribution and valuableinsights into identifying encryption areas at heavy-metal sites.