Georeferenced tractor wheel slip data for prediction of spatial variability in soil physical properties

Abstract

The upcoming technological breakthrough in the cropping system will offer a more detailed insight into soil-to-plant, man-to-soil, and man-to-plant impacts, thus improving the forecasting and ensuring more efficient in-field management. This study presents various on-the-go sensing procedures which were conducted in order to evaluate the quality of spatial estimations of soil physical properties such as soil compaction, soil moisture content, bulk density and texture. Standard statistical tools showed high positive correlations between soil specific resistance and soil compaction (R2 = .75), soil electromagnetic conductivity and moisture content (R2 = .72) and tractor wheel slip and soil compaction (R2 = .64). Variogram modeling of spatial autocorrelation gave the highest prediction error for tillage resistance (9.85%), followed by cone index (4.49%), moisture content (3.7%), bulk density (1.39%), clay + silt content (.98%), soil electromagnetic conductivity (.95%) and the least error was obtained for tractor wheel slip (.58%). The Central Composite Design (CCD) analysis confirmed significant contribution of soil compaction in the modeling of the specific soil resistance and tractor wheel slip, while soil moisture content and fine particle (clay + silt) content had a major impact on soil electromagnetic conductivity measurement. Soil bulk density had considerable importance in CCD modeling of tractor wheel slip.