STOCHASTIC MODELING OF TILLAGE-INDUCED AGGREGATE BREAKAGE

A Markov chain-based, two-parameter model was developed to predict til lage-induced influences on soil aggregates. Model parameters for sever al tillage operations were identified on two soil types with a downhil l, simplex, multidimensional, minimization approach. Simulation result s suggest that the crushing model can predict tillage-induced aggregat e crushing and that average prediction errors are within 3% for the li mited cases of verification. This study indicates that the stochastic simulation is better than the conventional deterministic method in est imating the tillage effects on soil aggregate size distribution due to apparent randomness of variability in the field data.