Skip to main content
ARS Home » Midwest Area » Columbia, Missouri » Cropping Systems and Water Quality Research » Research » Publications at this Location » Publication #334488

Research Project: Sustainable Intensification of Grain and Biomass Cropping Systems using a Landscape-Based GxExM Approach

Location: Cropping Systems and Water Quality Research

Title: Sensor data fusion for soil health assessment

Author
item Veum, Kristen
item Sudduth, Kenneth - Ken
item KREMER, ROBERT - University Of Missouri
item Kitchen, Newell

Submitted to: Geoderma
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/17/2017
Publication Date: 5/29/2017
Citation: Veum, K.S., Sudduth, K.A., Kremer, R.J., Kitchen, N.R. 2017. Sensor data fusion for soil health assessment. Geoderma. 305:53-61. doi: 10.1016/j.geoderma.2017.05.031.

Interpretive Summary: Soil health is linked to enhanced agricultural profitability and environmental protection. Assessment and monitoring of soil health is critical, especially in the Central Claypan Region of Missouri, where claypan soils have a high potential for runoff, soil erosion, and soil degradation. However, soil health is traditionally evaluated using a range of costly and time-consuming chemical, physical, and biological laboratory measurements. Estimating soil health using soil sensors such as visible and near-infrared (VNIR) reflectance spectroscopy, soil apparent electrical conductivity (ECa), and penetration resistance measured by cone penetrometer (i.e., cone index, CI) would save time and money for scientists and producers. In this study, we evaluated the benefit of combining ECa and CI information with VNIR data to estimate soil health for a range of perennial grassland and annual cropping systems. Soil health was quantified by using the Soil Management Assessment Framework (SMAF), which translates laboratory measurements into comprehensive scores related to crop productivity, environmental protection, and other important soil functions. Models to estimate soil health measurements and SMAF scores were developed. The model using VNIR in conjunction with ECa and CI successfully estimated 78% of the variability in the overall SMAF scores. The results of this study support the use of sensor fusion for in-field soil health assessment to drive management decisions and increase profitability. Overall, this study will benefit scientists and producers by demonstrating the potential for rapid, inexpensive, and high resolution soil health assessment in the field.

Technical Abstract: Assessment of soil health involves determining how well a soil is performing its biological, chemical, and physical functions relative to its inherent potential. Due to high cost, labor requirements, and soil disturbance, traditional laboratory analyses cannot provide high resolution soil health data. Therefore, sensor-based approaches are important to facilitate cost-effective, site-specific management for soil health. In the Central Claypan Region of Missouri, USA, visible and near-infrared (VNIR) diffuse reflectance spectroscopy has successfully been used to estimate biological components of soil health as well as Soil Management Assessment Framework (SMAF) scores. In contrast, estimation models for important chemical and physical aspects of soil health have been less successful with VNIR spectroscopy. The primary objective of this study was to apply a sensor fusion approach to estimate soil health indicators and SMAF scores using VNIR spectroscopy in conjunction with soil apparent electrical conductivity (ECa), and penetration resistance measured by cone penetrometer (i.e., cone index, CI). Soil samples were collected from two depths (0-5 and 5-15 cm) at 108 locations within a 10-ha research site encompassing different cropping systems and landscape positions. Soil health measurements and VNIR spectral data were obtained in the laboratory, while CI and ECa data were obtained in situ. Calibration models were developed with partial least squares (PLS) regression and model performance was evaluated using coefficient of determination (R2), root mean square error (RMSE), residual prediction deviation (RPD), and the ratio of prediction error to interquartile range (RPIQ). Models integrating ECa and CI with VNIR reflectance data improved estimates of the overall SMAF score (R2 = 0.78, RPD = 2.13, RPIQ = 3.66) relative to VNIR alone (R2 = 0.69, RPD = 1.82, RPIQ = 3.14), reducing RMSE by 14%. Improved models were also achieved for estimates of the individual chemical, biological, and physical soil health scores, demonstrating reductions in RMSE of 2.8, 5.4, and 10.0%, respectively. The results of this study illustrate the potential for rapid quantification of soil health by fusing VNIR sensors with auxiliary data obtained from complementary sensors.