Integrated ground penetrating radar - electromagnetic induction techniques to support precision agriculture by assessing the effects of agricultural practices

Abstract

Precision agriculture (PA) is an emerging approach to increase agricultural productivity while enhancing resource use efficiency, reducing costs, and minimizing negative environmental impacts. Understanding the spatiotemporal variability of soil properties and state variables within the agricultural landscape is crucial for informed decision-making in PA. This thesis explored integrating two geophysical techniques—Ground-penetrating Radar (GPR) and Electromagnetic Induction (EMI)—to investigate soil properties and state variables in a boreal podzolic soil site in western Newfoundland, Canada. These techniques offer a high-resolution, non-destructive, fast and cost-effective alternative to traditional soil sampling. Proxies derived from GPR (dielectric constant and amplitude), and EMI (apparent electrical conductivity) provide valuable insights into subsurface features, soil water content (SWC), compaction, and salinity, all of which are crucial for agricultural and environmental assessments. This research aims to assess the effectiveness of integrated GPR-EMI for characterizing soil profiles and properties, particularly focusing on their ability to assess spatial variations in soil stratification, SWC, bulk density, and salinity. Implementation of the integrated GPR-EMI technique overcomes the intrinsic limitations specific to either technique. Field experiments were conducted using both GPR and EMI surveys across varying soil conditions accompanied by soil sampling for ground truthing. This study developed predictive models, tested their accuracy, and assessed the ability of the integrated GPR-EMI approach to predict soil properties and state variables. As a result, management zones were delineated based on the integrated technique. The findings demonstrated a strong correlation between GPR and EMI outputs with soil sampling, confirming their complementary strengths in soil assessments. GPR was particularly effective in detecting subsurface stratifications and SWC variations, while EMI provided reliable estimates of electrical conductivity, essential for mapping soil salinity. The integrated GPR-EMI technique proved valuable in improving the understanding of soil properties and state variables and their spatial variability. This research demonstrates the potential of GPR and EMI to be applied effectively in boreal podzolic soils to predict key soil properties and state variables with significant implications for soil management and decision-making in PA. Further research is encouraged to refine the models and expand their applicability across diverse landscapes, conditions and soil types.