Modelling mean nitrate leaching from spatially variable fields using effective hydraulic parameters

When using simulation models for estimating the mean nitrate leaching on different soil types, the common approach is to interpret the field as a single equivalent soil column using effective hydraulic parameters, which are estimated from point measurements. The use of effective hydraulic parameters was evaluated on a coarse sandy soil and a sandy loam using the one-dimensional mechanistic model, DAISY. On each location, texture, soil water retention and hydraulic conductivity from 57 points were measured within an area of ca. 0.25 ha. The following approaches for estimation of effective hydraulic conductivity were examined: (1) geometric mean; (2) arithmetic mean; (3) estimated arithmetic mean from a lognormal distribution; and (4) mean estimated from a stochastic large-scale model for water flow, similar to the Richards equation in one dimension, but with large-scale effective parameters accounting for the local three-dimensional flow. The approach of interpreting the field as a number of non-interacting columns was examined by calculating the mean of the field as the mean of the 57 soil columns. The nitrate concentrations simulated by DAISY were compared with nitrate concentrations measured by ceramic suction cups at the 57 points at 25 cm and 80 cm depths during the winter period 1989/1990. At both locations, the nitrate concentrations simulated by the geometric mean, the stochastic approach and the mean of the 57 simulations matched the observed nitrate concentrations while the other approaches gave unreliable results on the coarse sand. Hence, to simplify the calculations the geometric mean can be used. (C) 1999 Elsevier Science B.V. All rights reserved.