Interim Baseline Radiological Investigation Report
November 14, 2012
Bear Lodge Project, for Rare Element Resources, Ltd.
Rev.1
25
To derive the Project-wide gamma-soil radium correlation, three iterations using linear
correlation methods were employed. The first attempt used all data (n=23), followed by a
truncated data set. The third attempt excluded non-detected values.
The first case (n=23) yields a fairly good correlation coefficient (R
2
= 0.63) with poor residuals
(i.e., predicted minus observed Ra-226 concentrations). Average residuals were positively biased
(indicating a tendency to overestimate soil Ra-226 concentrations). The regression model
indicated that two observations (0.3 pCi/g/20,031 cpm [SS-27-SS-081311] and 3.3 pCi/g/53,388
cpm [SS-33-SS-081311]) represented a large residual and leverage, respectfully.
Removing the two observations yielded a relatively poor correlation coefficient (R
2
=0.45) with
poor residuals. Average residuals exhibited a negative bias (indicating a tendency to
underestimate soil Ra-226 concentrations).
Returning to the original data set and removing the four, non-detected values in the data set (now
n=19) yields a relatively strong correlation coefficient R
2
= 0.71, with improved residuals; i.e., 8
positive and 9 negative values. One observation (3.3 pCi/g/53,388 cpm [SS-33-SS-081311])
continued to represent a large leverage on the model. The linear regression model for this
iteration, which has the highest correlation coefficient of the three, and was used to estimate Ra-
226 from gamma survey data for the Project is:
Concentration
Ra-226
= (5*10
-5
) cpm + 0.3318
A stepwise approach was not taken because it was clear, on cursory evaluation of the correlation
plot of all observations, that the data were too few and variable from which to draw salient
trends.
It is important to acknowledge that discrepancies between measured soil Ra-226 concentrations
reported by the laboratory and corresponding Ra-226 concentrations estimated by gamma
surveys are inevitable in a characterization survey of this nature and magnitude, given the
heterogeneity of the site (at least in some areas, differing detector-source geometry at various
sample/survey locations, prominent gamma-ray emitters from the thorium decay series
contributing to the instrument response, and uncertainty in the laboratory analytical results,
especially at low concentrations.
Using this equation, soil Ra-226 concentrations were estimated for each discrete gamma survey
measurement datum. Table 4.5 presents the resulting summary statistics for the Ra-226
concentrations predicted by the linear regression model from the scanning gamma count rates
observed in the Project.
