The project was initially designed as a low density soil geochemical reconnaissance meant to define broad regional trends in geochemical baselines. Similar surveys based on 100 x 100 km grids have been carried out by the USGS (Severson and Tidball, 1979; Severson and Wilson, 1990). Principles of low density survey design were discussed by Garrett (1983).

An 80 x 80 km grid was selected for the prairie survey, in conformance with IGCP 259 recommendations on low-density geochemical surveys (Darnley et al., 1995). These were subdivided into 40 x 40 km, 20 x 20 km and 10 x 10 km subcells and 1 x 1 km target cells were selected using a randomizing process for sampling (Garrett, 1994). This sampling design permits the spatial variability of the data to be quantified using Analysis of Variance. Rules specified procedures for selection of an alternative 1 x 1 km cell if the target could not be occupied. In each 80 x 80 km or 1600 km2 cell, two sites were sampled for A horizon, C horizon, and till, if present at a nearby exposure such as a roadcut. Additional soil samples were selectively taken in order to characterize the spatial variability between more closely spaced samples.