Mr. Tom McCraw September 22, 1976 Page 2. whether In-Situ or soil sample data are more indicative of future health risk of inhabitants. Is long term health a function of an average (integrated) measure of exposure such as obtained by an In-Situ device, or is it more a function of peak plutonium concentrations from soil aliquots? Also, the use of In-Situ measurements in the cleanup determination implies we need to determine the relationship of these readings to soil concentrations. This can only be done under field conditions. I would suggest a number of In-Situ measurements be made at different locations. At each location the total soil scanned by the device should be carefully collected and mixed and a number of aliquots be analyzed for plutonium. In this way a calibration equation relating In-Situ and soil sample concentrations can be estimated. This will no doubt need to be repeated for different islands or portions of islands since the calibration relationship may not be the same for all areas. If the decision to cleanup is made primarily on the basis of In-Situ measure- ments then the calibration information is necessary in order to relate to the cleanup criteria which, presumably, will be stated in terms of plutonium concentrations in soil samples. This reflects a basic decision needed before a sampling plan is selected. If clean-up decisions are to be based on wet-chemistry determinationson soil samples then the In-Situ device may serve only as a means of reducing the number of analyses needed. In any case, we suppose some chemical determination wil] be required for calibration of the device. Let us assume that the In-Situ measurements are related linearly to the average Pu concentration in the surface soil scanned by the In-Situ device. For example, if the Pu/Am ratio is constant then the data should Took something like a linear "average relationship" through the origin as indicated in the plot below. Y ° Upper 95% Confidence Line 239-240, (pCi/g) U (Wet Chemistry . ie Observations) Y 0 v* AE data P points = Clean-up Level Average Relationship IN-SITU DATA