| Abstrakt: |
The soil water characteristic (SWC) of sands is an important hydraulic parameter in designing golf courses and athletic fields. A modified version of the Arya-Paris model of the soil water characteristic was adapted to 14 golf course media that contained no to minor amounts of clay and silt. In this model, the particle-size distribution curve is divided into a number of fractions and the natural pore length, Li(n). is scaled using the diameter of spherical particles as the length unit. The scaled pore length is given by 2Riniαi, where ni is the number of spherical particles in the ith fraction, 2R, is the particle diameter, and αi, is the scaling parameter, which is calculated using the relationship log niαi = a + blogni. Although the model adapted well, there were concerns about the sensitivity of predicted SWCs to uncertainties in parameters a and b. Consequently, we developed and evaluated a procedure to predict Li(n) directly from straight pore lengths, Li(c) in counterpart cubic dose-packed assemblages of spherical particles, using the relationship logLi(n) = c + dlogLi(c). Predicted pressure heads using both procedures were similar with best-fit parameters. When uncertainties were imposed on Parameters a, b and c, d, however, SWCs using the latter procedure showed far less sensitivity, as measured by the root mean square residuals (RMSRs). In addition, for sand materials grouped together on the basis of similarity in particle-size distribution and bulk density, replacing individual best-fit parameters by the group mean parameters did not have significant effects on predicted pressure heads. [ABSTRACT FROM AUTHOR] |
Plný text