Preferential soil water movement was simulated using a modified version of
the MACRO model. The study contributes to the analysis of the held water re
gime and the solute balance of two tile-drained no-till agricultural soils
influenced by shallow groundwater tables. Previously measured soil matric p
otentials and groundwater levels as well as Bromide and dye tracer experime
nts indicated that flow and transport along preferential pathways might be
quantitatively important at the heavy clay as well as at the sandy loam sit
e. The objective of this study was to optimally describe the field measured
matric potentials and groundwater level fluctuations by using a one- and a
two-domain version of the MACRO model and by model calibration using avail
able data. The original MACRO model was modified with respect to simulating
effects of groundwater table fluctuations and tile drains in the one-dimen
sional flow model by including a sink term, based on the potential theory a
nd an empirical approach, that approximately relates the local to the held
water regime. We use an implicit finite difference discretization together
with a Newton-Raphson iterative scheme for the numerical solution of the mo
del. The coupled matrix and macropore flow equations are solved sequentiall
y. Model parameters were calibrated using laboratory-measured soil hydrauli
c parameters and field-measured time series of matric potentials and water
tables. Comparisons between the measured and simulated water table fluctuat
ions indicate that the consideration of preferential flow in macropores imp
roves the one-dimensional description of the water regime of the aggregated
clayey soil. For the sandy loam soil, neither the modified MACRO nor a one
-domain model could explain observed preferential flow patterns. The failur
e at the sandy loam site may possibly be caused by periodically occurring h
ydrophobic conditions at the soil surface or by the existence of fissures i
n a relatively low permeable subsoil horizon, effects which could not be co
nsidered in the model. Hydraulic parameters of the soil matrix and macropor
e system could satisfactorily be calibrated. However, the parameters of the
exchange term describing water transfer between the inter- and intraaggreg
ate pore system could not be clearly identified using the available field d
ata. (C) 1999 Elsevier Science B.V. All rights reserved.