Date of Award


Degree Type


Degree Name

Doctor of Philosophy


Biosystems Engineering

Major Professor

John B. Wilkerson

Committee Members

Daniel C. Yoder, H. Paul Denton, Bruce Ralston


Terrain attributes are among the most studied soil characteristics. Although important, in only a few areas are topographic features mapped at the field-scale level. These features play an important role in assessing the crop production potential and erosion susceptibility of agricultural fields. Therefore high-resolution mapping of terrain attributes are vital to a better management of production fields. Today, terrain attributes are derived from elevation measurement.
A more direct form of measurement was developed by Rowe and Spencer (1976), pitch and roll angles were used to derive slope gradient and vehicle attitude. Yang (1997) related vehicle attitude to slope aspect. The existing mathematical models are difficult to implement with today’s low-cost micro-controllers because of existing trigonometric functions. Research conducted at the Biosystems Department of The University of Tennessee focused on the simplification of existing models and on the development of an electronic system to test two sensing techniques in dual-axial rotational measurement of a roving vehicle: a clinometer and an accelerometer. Tests were conducted in a field with a widely varying topography located on the Blount Experiment Research Unit, by mounting the electronic monitoring system on an ATV. Elevation data measured with a RTK-GPS were used to generate an accurate elevation map. Terrain attributes were calculated in 3 spatial resolutions: 4, 16, and 100 m2.
Simplification of the mathematical models relating pitch and roll angles to terrain attributes is possible because of the existing limitations on slope gradient of arable lands. Results obtained during field tests show that slope measurement accuracy varied according to spatial resolutions. The density of points used in the calculation of the terrain attributes also contributes to measurement accuracy. In general, mean absolute error (MAE) were less than 1° for both sensors in all resolutions tested. Data collected from pitch and roll sensors can also be used to detect field elevation changes.
In conclusion, it is possible to rely on measurements of vehicle axial rotation for the computation of field-scale terrain attributes. The sensing techniques tested were successfully used in these measurements. The application of simplified models to derive terrain measurements is possible due to the existing slope gradient limitation of arable lands. It is possible to describe terrain attributes in a scale similar to order I soil maps using the proposed electronic system and models. The system can also be used to pinpoint locations of elevation differences in the field.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Included in

Engineering Commons