We investigated the effect of sample size in determining parameters in solute transport experiments. Potassium bromide was uniformly applied to an 8− x 12-m plot and then ponded with a 100-mm irrigation. Twenty sites were sampled on a 1.8− x 2.5-m grid 6 d after irrigation. A 103-mm-diameter sample tube was inserted into the soil at 15-cm intervals. A 22-mm concentric subsample was obtained from the larger sample. The variances of the Br- concentration, water content, and total Br- recovery were not significantly different between the two sample sizes at the 0.05 significance level. The mean difference between the large and small samples was significantly different at the 0.01 level for Br- concentration and Br- recovery. The slope of the regression line was significantly different from 1 for Br- concentration only. Geostatistical structure analysis indicated that the variance of solute velocity using the large sample would only be about 3% smaller than that of the small sample. Although there was a significant bias toward the larger sample size for Br- concentration and recovery, spatial variability effects are far greater than sample size effects, and the Br- transport was adequately represented by the data collected using the small sample. The representative elementary volume was equaled or exceeded with the 22-mm-diameter sample.
No takes yet. Share an insight, caveat, or question.
Rice et al. (1988) studied this question.