MISSING LINKS: UNIFYING CONCEPTS IN SOIL TRANSPORT PROCESSES

Predictions of soil hydrology are limited by soil’s variability and extent. Soil survey results give only ballpark values of soil properties, and the equations using those properties have little physical basis. For example, the most popular models of the soil water retention function and the unsaturated hydraulic conductivity function are explicitly based upon capillary bundle models. These models are easy to visualize and calculate with, but they are physically unrealistic: used consistently, they give results that are quite incorrect. Furthermore, they set us up to have deeply flawed intuitions about how flow and transport work.

An attractive alternative to conceptual models based upon capillary bundles is to think of soil as being a pore network. Networks are inherently more complex, but fortunately the mathematics of connections and networks is mature enough that results exist that fit our needs. Aside from mathematical and conceptual consistency, an advantage of using more realistic conceptual models is that new patterns of transport properties can become evident. At a more practical level, these new models give some hope that reasonable estimates of soil hydraulic properties will be derivable from soil survey information, allowing improved predictions of landscape-scale hydrology.