Pedo-transfer functions with multiple linear regressions to predict solute-transport parameters

Abstract

Transport parameters of soluble chemicals through soils are needed to assess thepollution risks of soil and groundwater resources. But, it is time consuming, laborious,expensive and, practically, impossible to experimentally measure such parameters for awide range of solutes and soil types. So, indirect estimate of the parameters by pedotransfer function is becoming popular. The aim of this study was to develop and evaluatepedo-transfer functions (PTFs) for solute-transport parameters by multiple linearregression (MLR) analysis. For this, transport parameters of three heavy metal/metalloid compounds $ (NaAsO_2, Pb(NO_3)_2, Cd(NO_3)_2),$, a pesticide (carbendazim) and aninert salt $(CaCl_2)$ through 14 agricultural soils of Bangladesh were determined. Thetransport experiments were done in repacked soil columns under unsaturated steadystate water flow conditions. Breakthrough data of the solutes were measured with timedomain reflectometry (TDR), and velocity (V), dispersion coefficient (D) and retardationfactor (R) of the solutes were determined by analyzing the data by a transfer-functionmethod. Bulk density ($gamma$), organic carbon (OC) content, clay (C) content, pH, median graindiameter ($D_{50}$) and uniformity coefficient ($C_u$) of the soils were determined. Regressionmodels for V, D and R were developed with $gamma$, OC, C, pH, $D_{50}$ and $C_{50}$ as the input variables.Bulk density and clay content were found the most sensitive input variables to the MLRmodels. The MLR models fairly predicted V, D and R, and thus provide a way ofsignificantly enhancing prediction of reactive solute transport through agricultural soils.