Author
 |
CHUNG, SUN-OK - UNIV OF MO |
|
Submitted to: University of Missouri, Columbia, Thesis
Publication Type: Other Publication Acceptance Date: 4/22/2004 Publication Date: 5/31/2004 Citation: Chung, S.O. 2004. On-the-go soil strength profile sensor [Ph. D. Dissertation]. Columbia, MO: University of Missouri, Columbia. 254 p. Interpretive Summary: Technical Abstract: Precision agriculture quantifies and manages within-field variability and requires denser spatial data than conventional agriculture that treats a field uniformly. In this study, a horizontally operating on-the-go soil strength profile sensor (SSPS) was developed so that mechanical resistance of soil could be measured at multiple soil depths simultaneously and continuously while the SSPS traveled across a field. Force divided by the base area of the SSPS sensing tip was defined as a prismatic soil strength index (PSSI), similar to the cone index (CI, penetration resistance divided by cone base area) of a vertically operating cone penetrometer, the standard tool currently used to quantify soil strength. The SSPS would provide CI-like measurements, but in a more efficient manner and with a higher spatial density than do cone penetrometers. Important SSPS design parameters were determined through preliminary analysis of cone index and tillage draft data, and mathematical modeling of soil failure caused by the SSPS and a cone penetrometer. The SSPS used 5 prismatic tips, each with a 60 degree cutting angle and a base area of 361 mm**2, and the design maximum operating depth, the upper limit and resolution of soil strength were 0.5 m, 19.4 MPa, and 0.14 MPa, respectively. Tip extension from the main blade and spacing between adjacent tips were optimized as 5.1 cm and 10 cm, respectively. The SSPS was tested in a soil bin, and in two field sites with variations in soil water content, bulk density, and texture. In the soil bin tests, increases in PSSI with speed were less than 15% up to a 3.0 m/s operating speed and 1.5 m/s was selected as a maximum field data collection speed, below which speed effects on PSSI would be negligible. Linear relationships between PSSI and CI were found for a given operating speed and depth. In field tests, PSSI data collected in adjacent, parallel transects were linearly related (r2=0.93), with slopes not statistically different from 1, indicating that the PSSI measurements were repeatable and stable. In general, PSSIs were higher at locations with lower apparent soil electrical conductivity (ECa, a surrogate for soil texture) and water content, and higher bulk density. The most accurate PSSI prediction models were obtained when interactions among the affecting factors were included as independent variables and when data were grouped into subsets by depth and/or ECa level. Coefficients of determination for estimating PSSI were 0.61 and 0.52 at sites 1 and 2, respectively. When CI was predicted as a function of PSSI and other affecting factors, coefficients of determination were 0.42 and 0.52 for the two test sites. Overall, the prototype SSPS performed well, providing repeatable and stable measurements of soil strength in various soil and operating conditions. With its ability to acquire PSSI data at a high spatial and vertical resolution, the SSPS would provide useful information for application in precision agriculture. Future study to improve the prototype SSPS (e.g. incorporation of a depth sensor) and to provide more information for interpretation of PSSI would be useful. |
