Analysis of surface moisture variations within large‐field sites

In the past several years, 1974–1977, NASA has conducted several research studies to develop an extensive collection of ground truth soil moisture data. As a result of these experiments, moisture data were available from 58 ‘large‐field sites,’ each being 400 × 400 m (40 acre). The field locations were 1 field in Phoenix, Arizona (sampled four times), 28 fields in Jefferson County, Kansas, 23 fields in Finney County, Kansas, and 5 fields in Hand County, South Dakota. At the first three locations, samples were taken in specific vertical increments or horizons (i.e., 0–1; 1–2, 2–5, 5–9, and 9–15 cm). In the South Dakota study, moisture samples were taken in increments from the surface ( i.e., 0–2.5, 0–5, and 0–10 cm). A detailed statistical analysis was made to define the general relationship and ranges of values of the field moisture relative to both the variance (standard deviation) and coefficient of variation ( CV ) for a given test site and depth increment. On the basis of the results of the variability study it was concluded that (1) moisture variations within any given‘large‐field’ area are inherent and are normally distributed about the mean, (2) neither a single (constant) value of the standard deviation nor coefficient of variation uniquely define the variability over the complete range of mean field moisture contents examined, and (3) using an upper bound standard deviation parameter clearly defines the maximum range of anticipated moisture variability. It was found that 87% of all large‐ moisture content standard deviations were less than 3%, while about 96% of all the computed values had an upper bound of s = 4% for these intensively sampled fields. Using these upper bound magnitudes as an estimate of the population standard deviation and a preselected confidence level, limit of error curves of mean soil moisture measurements for large‐field sites relative to the required number of samples were determined.