Nonintrusive mapping of subsurface features for site-specific agriculture : Cumberland Plateau and Loess Uplands physiographic regions of Tennessee

Evaluation studies were conducted to determine if ground-penetrating radar (GPR) could be used to noninvasively map subsurface phenomena that affect the variability of crop yields in two physiographic regions of Tennessee. GPR technology offers great potential to agricultural researchers for noninvasive mapping of the various subsurface features found within these geological regions.

Calibration data were collected at three survey locales containing soils similar to those found in the main research areas. GPR survey methodologies within each region were developed, and optimal system settings were obtained. Primary subsurface features of interest to the study were also mapped and methodologies of interpreting the features from GPR imagery were evaluated using a "blind test". Results of the "blind test" indicated a relatively high degree of accuracy and repeatability.

Geographic information systems (GIS) were used as a tool to geographically join the GPR data to crop yield values from four primary research plots. Statistical analyses were then performed to determine the correlation between the two types of data. Correlation coefficients indicated that interpretations from GPR imagery were capable of describing a great deal of the spatial variability observed in the crop yield trends. Furthermore, results from the Least Squares Means analysis revealed a yield potential pattern of the soils that each GPR interpretation represented.

Recommendations were made concerning survey procedures, equipment, and interpretation methodologies. Difficulties encountered are also discussed along with suggested solutions. Finally, the direction of future research in the area of GPR in site specific farming (SSF) is discussed, and recommendations for this research are presented.