Soil organic matter and cation-exchange capacity sensing with on-the-go electrical conductivity and optical sensors

Abstract

An on-the-go optical soil sensor with 660nm red and 940nm near-infrared wavelengths with an electrical conductivity (EC) sensing unit were tested to estimate soil organic matter (SOM) and cation-exchange capacity (CEC) on 551ha on 15 fields in 6 U.S. states. For calibration between sensed data and lab-analyzed values, a multivariate linear regression (MLR) with leave-one-out cross validation was performed on fields with more than 10 lab samples and a single variable linear regression was performed on fields with less than 10 samples. From the SOM calibration results, 12 of 15 fields had good results with R2 of 0.80 or higher and RPD (Ratio of Prediction to Deviation=standard deviation / root mean square error of prediction) of 2.33 or greater. For CEC calibrations, six of nine fields had good results with R2 of 0.86 or higher and RPD of 2.78 or greater. The best calibration model was applied to each field and the estimated SOM and CEC maps exhibited strong spatial structure and high correlation to lab-analyzed SOM in all fields. EC and optical data in each field was normalized and combined together by state and tested with MLR. Combining fields in this manner showed good results with R2 of 0.80 or higher and RPD of 2.30 or greater for SOM in four of five states, and combined fields in two of three states showed good correlations to lab data with R2 of 0.86 or higher and RPD of 2.69 or greater for CEC. From these results, SOM and CEC mapping with soil EC and optical sensors seems to be a promising approach. Future research will be implemented to estimate SOM and CEC more precisely by developing a reliable universal calibration model using soil EC, optical data, soil moisture contents and topographic attributes for global areas.