The evaluation of pressure distribution and bulk density models for infield agriculture and forestry traffic.

Abstract

There is evidence that soil compaction, through the use of mechanised equipment, causes detrimental effects to soil quality and reduces long-term productivity of soils. For economic reasons, farmers need to purchase larger, heavier machinery in order to cultivate larger areas under crops, resulting in larger forces on the soil. The severity of soil compaction is governed by various soil and vehicle properties and normally causes an increase in the soil's bulk density and a decrease in the air filled porosity. These changes in soil properties have negative effects on crop production and environmental sustainability. The aim of this study was to investigate and develop a model based decision support system for soil compaction management and research. Soil compaction occurs during the transfer of stresses from the tyre interface into the soil. Numerically, it has been modelled using both mechanistic and empirical models, which attempt to simulate the stress propagation and also sometimes the consequent damage to the soil. The SOCOMO soil compaction model is described and this model computes the stress at a point in the soil for any given horizontal and vertical stress distribution at the soil / tyre interface. It has been successfully used in the Netherlands and in Sweden to map the impact on the soil. The SOCOMO model was tested and verified at a forestry site in Richmond, KwaZulu- Natal. Relationships to determine bulk density were also tested and verified. The SOCOMO model performs satisfactory (RMSE = 47.9 kPa), although it tends to overestimate the pressures within the soil. This could be as a result of the high organic carbon content in the particular soil. Models predicting bulk density also performed satisfactory (RMSE = 69.9 kg.m" ), but resultant densities in the soil are generally underestimated. Future research is needed to find better relationships to estimate changes in dry bulk density and to test the model on a wider range of soils. If the model performs satisfactory it could provide a useful tool to determine the impact of soil compaction on crop yield.