Influence of Soil Properties on the Statistical Characterization of Dry Aggregate Strength
- E. Perfect ,
- B. D. Kay and
- A. P. da Silva
Dry aggregate strength measurements, important in tillage and erosion studies, are often analyzed statistically using the Weibull model. The parameters of this model, α and β, characterize the absolute strength and the spread of strengths, respectively. Information is lacking on the influence of soil properties on values of α and β. We measured dry aggregate strength as the specific rupture energy (E in J g−1) of 40 aggregates in two size classes (4–8 and 8–16 mm) from 36 locations along a 500-m transect across four soil types with a wide range of physical and chemical properties (e.g., 8.0–72.6% sand content). Estimates of α and β were obtained by fitting Weibull's model to these data using nonlinear regression. The α and β values were lognormally and normally distributed, respectively. The mean log(α) was -2.4 and the mean β was 1.3. Multiple regression analysis was used to relate log(α) and β to soil properties. The resulting regression equations explained 93% of the variation in log(α) and 60% of the variation in β. The log(α) increased with decreasing aggregate size and sand content and increasing organic matter content. The β increased (indicating a narrower range of strength values) with decreasing log(α) and aggregate size. The regression equations were validated by predicting the cumulative relative frequency distribution of E for aggregates from a different site where independent measurements of E were available for comparison. A good agreement (R2 ≥ 0.91) was obtained between predicted and observed values of E.Please view the pdf by using the Full Text (PDF) link under 'View' to the left.
Copyright © .