Predicting Water and Chloride Transport in Drained Soils Derived from Glacial Till

Abstract

AbstractField‐scale predictions of nitrate (NO3) leaching from farmland are needed to enable zones vulnerable to ground or surface water pollution with NO3 to be identified. Use of simple water, solute transport, and N dynamics models is one way of making these predictions. In this paper, linked water and solute transport models (WATBAL/TRANSOL) are calibrated for two important arable soil types in Scotland: a clay loam (aeric haplaquept) and a sandy loam (aquic eutrochrept). Error in prediction of water and solute transport is evaluated using data from two drained plot experiments. WATBAL predicted annual deep percolation of 25 and 44% of the measured drainage flux at the clay loam and sandy loam sites, respectively. The RMS errors in prediction of annual drainflow were 14 and 13% at the clay loam and sandy loam sites, respectively. Prediction of the onset of autumn drainflow over 3 to 5 yr was good for the clay loam soil, but not the sandy loam soil. Chloride leaching was underpredicted by about 22% at the sandy loam site and by 15% at the clay loam site in the first year after chloride tracer application. This indicates the likely extent of under prediction of solute leaching to drains when these models are linked to other models, such as simulations of N dynamics.