Submitted to: Soil Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/20/1999
Publication Date: 10/1/2000
Citation: N/A Interpretive Summary: This research describes an improved method to estimate saturated hydraulic conductivity using soil porosity, an easily measured soil property, and an index of soil pore size distribution. The soil pore size distribution can be easily estimated from soil survey data. Understanding the process of water infiltration into soil is important to be able to predict soil erosion, runoff of water and chemicals from soil, water availability to plants, movement of chemicals to groundwater, a salt leaching, and groundwater recharge. An important soil parameter is the conductivity of soil to water, the soil hydraulic conductivity. Researchers and agricultural managers who need to be able to predict infiltration must have estimates of saturated hydraulic conductivity (KSAT) including the range of values they can expect. Unfortunately, soil hydraulic conductivity can be expensive to measure. A commonly used relationship to estimate KSAT relates saturated conductivity (KSAT) to the product of a coefficient (B) and porosity raised to a power (n). Our objective was to determine how and whether B and n are related to parameters which describe the release of water from soil, as suction is applied. We applied a neural network technique to explore the relationships among the parameters. We found useful relationships among the parameters which will allow us to better estimate saturated soil hydraulic conductivity from easily measured soil properties. This improved method of estimating soil hydraulic conductivity will make it easier for agricultural managers to obtain realistic values of hydraulic conductivity in order to predict infiltration.
Technical Abstract: The modified Kozeny-Carman equation relates saturated conductivity (Ks) to the product of a coefficient (B) and porosity raised to a power (n). Various studies have reported different values of B and n which are found in different data sets. Our objective was to determine whether and how B and n are related to Brooks-Corey's air entry pressure hb and pore distribution index. The Southern region soil hydro logic database of ca. had 500 samples explored. All soils had contents of both silt and clay less than 70%. Neural networks were used to relate B and n to bubbling pressure (hb) and lambda, and a genetic algorithm was applied to find weights in neural networks. Dependencies of B and n on hb and lambda had similar shapes. Values of B and n were almost constant for values of the lambda greater than one, and were close to 950 cm day-1 and 2.5, respectively. As the values of the lambda decreased from one to zero, values of B and n decreased. The larger the air-entry pressure, the steeper the decrease in B and n was. The results of this study will help improve the accuracy of KS estimation from porosity.