STOCHASTIC-ANALYSIS OF SOLUTE TRANSPORT IN PARTIALLY SATURATED HETEROGENEOUS SOIL .1. NUMERICAL EXPERIMENTS

Citation
D. Russo et al., STOCHASTIC-ANALYSIS OF SOLUTE TRANSPORT IN PARTIALLY SATURATED HETEROGENEOUS SOIL .1. NUMERICAL EXPERIMENTS, Water resources research, 30(3), 1994, pp. 769-779
Citations number
40
Categorie Soggetti
Limnology,"Environmental Sciences","Water Resources
Journal title
ISSN journal
00431397
Volume
30
Issue
3
Year of publication
1994
Pages
769 - 779
Database
ISI
SICI code
0043-1397(1994)30:3<769:SOSTIP>2.0.ZU;2-G
Abstract
The effect of water saturation on the transport of a conservative, non reactive solute pulse by quasi steady state and transient flows in a v ertical cross section of a hypothetical, yet realistic heterogeneous, partially saturated soil has been analyzed here by a series of numeric al experiments. Mean water saturation in the flow domain was controlle d by predetermined capillary pressure head at the entry zone, or by pr edetermined time interval between successive water applications. Analy ses of the simulation results show that under both quasi steady state and transient flow regimes, lower degree of water saturation enhances solute spreading, in qualitative agreement with the results of Lagrang ian-stochastic analyses of vadose-zone transport [Russo, 1993a, b]. Th e pattern of the velocity field, and, consequently, the spreading of t he solute body, were affected considerably by the boundary conditions imposed on the flow at the entry zone. In the case of the quasi steady state flows under predetermined capillary pressure head at the inlet zone, the flow was essentially unidirectional vertical, with slight lo cal derivations and, consequently, solute spreading took place mainly in the longitudinal direction. In the case of transient flows under pe riodic influx, the flow pattern was more complicated, essentially two dimensional, thereby restricting the longitudinal spreading of the sol ute plume and enhancing its transverse spreading. In the companion pap er (Russo et al., this issue), integrated measures of the simulated so lute plumes are compared with their predicted values, based on the fir st-order Lagrangian-stochastic analysis.