A TRANSFER-FUNCTION APPROACH TO MODELING THE LEACHING OF SOLUTES TO SUBSURFACE DRAINS .1. NONREACTIVE SOLUTES

The leaching losses of indigenous chloride and autumn-applied solid sodium bromide were measured for a mole-pipe drainage system in a silt loam soil under pasture. About half the applied bromide was recovered in the 230 mm of drainage that occurred in the year following application. Soil sampling to 1 m depth in the following spring showed that 12% of the applied bromide had moved below the mole drain depth (450 mm) into the relatively impermeable subsoil. Chloride concentration in the drainage was about 50 g m-3 in autumn, decreasing to about 10 g m-3 by the end of the drainage season in spring. Assuming pasture uptake and animal returns of chloride were in approximate balance, chloride was treated as a conservative solute. The leaching of resident chloride in two consecutive years was successfully simulated using the simple net-applied-water form of the transfer function equation, with an exponential function for the probability density function (pdf) describing solute travel. To account for plant uptake of bromide, which was not balanced by animal returns, first order decay was incorporated into the transfer function equation for bromide. But the same travel pdf could only be used for bromide and chloride if it was assumed that the bromide entered the soil's transport volume gradually, rather than as an impulse at the time of application. When the transfer function model '''was tested against published data for the leaching of native and surface-applied chloride from the same site in two earlier years, it did not perform as well as expected. This was apparently due to exceptional hydrological events in each of those years, an unusually large and heavy rain storm in one, and an unusually long dry period during winter in the other. These events apparently affected leaching in a complex way, through either the travel pdf or the way solute entered the transport volume.