Water Flux in Moist Soil: Thermal Versus Suction Gradients

That thermal gradients cause moisture transport in soil has been well known for at least 50 years. It is, however, surprising that so little attention has been paid to this phenomenon, since soil in its natural environment is continuously subject to changing temperatures. In 1957 Philip and de Vries (14) published a theoretical article in which they attempted to reconcile the few existing experimental data. Since then, alternative approaches have been suggested by Derjaguin and Melnikova (5), Matthes and Bowen (12), and Taylor and Cary (16). The experimental observations available are, however, insufficient to make a fair test of any of the proposed theories. As a general statement, about all that can be said is: A thermal gradient in soil will cause water to move from a warm to a cooler area in both the liquid and vapor phases, and the rate of transfer is greater than can be predicted with Lick's law and the diffusion coefficient for water vapor into air. The experiment reported in this article was designed to yield data defining the relative importance of thermal gradients in transporting soil moisture and to probe the mechanisms of transfer