Premium
A simple predictive approach to solute transport in layered soils
Author(s) -
DYSON J. S.,
WHITE R. E.
Publication year - 1989
Publication title -
journal of soil science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.244
H-Index - 111
eISSN - 1365-2389
pISSN - 0022-4588
DOI - 10.1111/j.1365-2389.1989.tb01294.x
Subject(s) - effluent , drainage , soil science , dispersion (optics) , log normal distribution , chloride , nitrate , chemistry , mineralogy , mechanics , hydrology (agriculture) , geology , environmental science , geotechnical engineering , mathematics , environmental engineering , statistics , ecology , physics , organic chemistry , optics , biology
SUMMARY Solute transport through layered columns (repacked aggregates overlying sand) was studied under steady flow conditions. Predictions of transport were simplified by assuming that the distribution of solute travel times in one layer was not correlated with that in the other. The implications of this assumption were developed for the transfer function model (TFM) and the convection‐dispersion model (CDM) of solute transport. The parameter values in each model were obtained from experiments carried out on columns containing only aggregates or sand. The solutes used were nitrate (surface‐applied) and chloride (previously distributed); predictions of the chloride movement were made using the parameter values for the nitrate. The predictions were tested against experimental values of drainage effluent concentration and solute concentration with depths in the columns (measured at the end of the experiments). The TFM (with an assumed lognormal distribution of travel times) and the CDM did not differ significantly, mainly because the spatial scale of the experiments was small. Because the parameter values for the columns of aggregates or sand were determined from the drainage effluent data, they were average values for whole columns. These parameters were satisfactory for predicting drainage effluent concentration from the two‐layer columns. However, they were not satisfactory for predicting the depth distribution of solute, particularly in the sand, because the water content of the sand increased with depth, unlike that of the aggregates, which was approximately constant with depth. The overall results of this study on materials of differing transport characteristics suggest that the assumption of uncorrelated travel times between layers has a potentially wide application. The approach taken here needs to be tested on undisturbed layered soils.