TABLE IV SOIL SAMPLING AND ANALYTICAL METHODOLOGY AT VARIOUS STUDY SITES 2 Sampling Area (cm ) Sampling Technique Location Amount Analyzed (2) Sample Prertreatmenct Los Alamces Core 4.5 None Whole sample ~ 25 Trinity Core 4.5 None Whole sample ~ 25 Eniwetok Core Thule, Greenland Excavation Rocky Flats Excavation 10 Ball-milled 100-127 Core-excavation Nevada Test Site 10-50 Bail-milled ~ 100 Remove > 0.6 cm 100 5 samples composited 5-10 Remove > 0.5 cm TABLE V 238 SPATIAL RELATIONSHIPS OF n Within Plot 160 to 2550 9 to 160 Within Strata 160 to 2560 CG to 160 Overall Coefficient of variation Pu CONCENTRATION VARIABILITY IN MORTANDAD CANYON SOILS 1.2 0.79 0.94 27 5 12 relate variability to any specific factor. The dif- ficulties in comparing soil plutonium data between 0.85 (0.60) 0.75 (0.31) 4 5 The variability was considerably reduced by examining the data from smaller spatial units as sites can be appreciated by examining some of the achieved through stratificarion. potential components of variability at the various CV was calculated as 1.2 when all the data from che sites. contaminated portion of the canyon were used, where- Soil sampling and analytical methodology The concentration as the data from two segments of scream channel employed at several sites are summarized in Table IV reduced the CV to 0.79 and 0.94 for the respective to emphasize the potential for these factors as strata. components of the total variability. Although chem- samples from each sampling location within che 2560- Wichin-plot variability based on triplicate ical procedures for plutonium analysis between sites m segment averaged 0.75 and 0.85 for the two strata, are similar in thac they involve HNO,-HF digestion, indicating that closely spaced replicate samples followed by column separation and alpha speccrometry, did not greatly reduce the variabilicy as measured considerable differences exist in sampling tech- over much larger areas as represented by the two niques, strata. 3 sample preparation, and in the size of soil The relationships between soil pluconium con- aliquot analyzed for plucoaium, One of che sources of tield sampling variability for soil plutonium is the heterogeneous distribution of the element within the study areas. An example of the effect of spatial heterogeneicy on variabil- centration variabiliry, soil particle size fraction, and soil depth were investigated at those sites where data were available Recky Flats) (i.e., Los Alamos and to examine further the components of ity is given in Table V for Mortandad Canyon at Los plutonium concentration variability. Alamos where a plutonium concentration gradient Table VI illustrate che relationship of tne CY exists with distance down the canyon. with soil particle size fraction and soil depth for Table V were calculated for all samples The CVs in (n = 27) The data in soils from Mortandad Canyon at Los Alamos. There collected within a 2560-m segment of stream channel, were no significant relationships between CV and for two strata (0 to 160 and 160 to 2560 a) within soil particle size fractions that segment, as inferred by analysis of variance. and for triplicate sampies taken at each of 9 plots within that segment. in either study area However, in all cases, there was a significant change (9 < 0.05) in CV with soil depth. 63