ISOTOPES AND RADIATION NUCLEAR TECHNIQUES IN GEOLOGIC, LUNAR, AND PLANETARY STUDIES I 2. Effects of Neutron Source Type on Soil

A number of radioisotopes have recently become commercially available as alternatives to radium-225 in moisture gauging devices using alpha-neutron sources for determining soil moisture, for well logging, and for other industrial applications in which hydrogenous materials are measured. To date, little work has been reported on 21 the comparability of data obtained with different sources when used in soil moisture meters of the same design. MacGillivray et al. 1 developed a tentative hypothesis that differences in "sphere of influence" due to different types of sources could account for anomalies they observed in correlating soil moisture data. This study investigates that hypothesis. Measurements with six moisture meters having sim ilar geometry but different neutron sources [ 226 RaBe, 239 PuBe, and 241 AmBe (α n)] were made in noninfinite volumes of soil at three moisture contents (1, 43, and 48% by volume), and in an effectively infinite volume of water. Detectors were interchanged between source housings to eliminate differences attributable to detection systems. Thus, measurements were made using each of the six sources with each of the six detection systems in each of the four media, yielding 144 measurement combinations. If differences in neutron energy between various sources influence the volume measured by the meter, then it would be expected that: 1) measurements made in the effectively infinite water medium would be unaffected; 2) there would be a minimum effect in the driest soil, since indicated moisture was near zero; and 3) the greatest difference would be observed in the mid-moisture range in noninfinite size containers. A straight line connecting the count rates obtained from the driest soil and from the water media for each meter would therefore not be affected by variability due to neutron energy differences. The deviation from that straight line of count rates obtained in the two intermediate moisture media, expressed as a percentage of the predicted count rate, is shown in Table I. a Significantly different from predicted count rate at both 95 and 99% level of. confidence. The two Ra-Be sources were not significantly different from one another, and gave values about 3% lower than Am-Be 2 and the two Pu-Be sources, and 4.6% lower than Am-Be 1. The actual values of these differences are a function of the physical characteristics of our test media, and would not be applicable to any other environment. They do, however, confirm the hypothesis that the type of neutron …