THE appHcabihty of a non-uniform mixing model to describe the transfer to runoff of four surface-applied herbicides (atrazine, alachlor, propachlor, and cyanazine) and three nutrients (NH4, NO3, and PO4) was tested using field plot data for four levels of residue cover (0, 375, 750, and 1500 kg/ha). The model incorporates the varying degree of mixing with depth between rainwater and soil during the chemical transfer process, as well as the effects of infiltration on chemical movement into the soil before and after runoff begins. Starting from the time of runoff initiation, it is assumed that the degree of mixing between rainfall and soil solution, p, decreases exponentially with soil depth so that p = exp ( bz), where b is a constant and z is the soil depth. The adsorption-desorption process is represented by a proportional relationship, Q = aC, where C, is the concentration of chemical in the adsorbed phase on soil particles (g/g), C is the concentration in soil solution (g/mL) and a is a constant. Numerical computations are made in small intervals of soil depth and time. In all cases, model calculations compared well with measured chemical concentration-time curves. The structure of the model for the variable degree of mixing with depth, the partitioning equation for adsorbed and solution phases, and the downward displacement of the chemical by infiltration seems to reasonably represent reality. The a parameter, for a particular chemical, was generally constant among replicates of a given residue level. The b parameter varied some between replicates apparently due to spatial variations of surface conditions. Values of a and b, in general increased as residual level increased. The increase in a indicates that the residue cover on the soil surface shields some of the chemical underneath, which is, in effect, similar to increasing the adsorption of the chemical. The increasing values of b indicate that increased residue cover decreased the depth of soil mixing with rainwater.
No takes yet. Share an insight, caveat, or question.
Heathman et al. (1986) studied this question.