Assessing dataset equivalence and leveling data in geochemical mapping

•Two methods are proposed for assessing dataset equivalence in geochemical mapping.•The methods also include procedures for the leveling of bias between datasets.•Both methods are applied to a case study involving Fe, V and Y datasets.

Combining data originating from two or more geochemical surveys can be highly beneficial for establishing geochemical maps with increased resolution and/or coverage area. However this practice requires assessing the equivalence between datasets and, if needed, applying data leveling to correct possible biases between datasets. Here we propose two original methods for assessing equivalence and for leveling data when datasets contain records that are located within the same perimeter. The first method is designed for datasets that are similarly spatially distributed and is based on the Kolmogorov-Smirnov test and quantile regression. The second method does not require datasets to be similarly spatially distributed and is based on prior knowledge about the factors explaining the geochemical concentrations and on BLS (Bivariate Least Squares) regression. The scope of application, pros, cons and detailed practical recommendations are presented for each method. Both methods were applied to a case study involving Fe, V and Y datasets originating from two European geochemical mapping projects: the Geochemical Mapping of Agricultural Soils of Europe (GEMAS) and the Baltic Soil Survey (BSS). Both methods for assessing the equivalence and obtaining leveling equations yielded comparable results thereby illustrating their effectiveness and their feasibility.