Comparison of two spectrometers for profile soil carbon sensing

Authors: HODGE, ANNA - University Of Missouri; Sudduth, Kenneth - Ken

Submitted to: Proceedings of the American Society of Agricultural and Biological Engineers International (ASABE)
Publication Type: Proceedings
Publication Acceptance Date: 6/22/2012
Publication Date: 7/29/2012
Citation: Hodge, A.M., Sudduth, K.A. 2012. Comparison of two spectrometers for profile soil carbon sensing. In: American Society of Agricultural and Biological Engineers International Meeting Technical Papers. July 29 – August 1, 2012, Dallas, Texas. Paper No. 121338240.

Interpretive Summary: Rapid estimation of soil properties is needed for determining and mapping soil variability in site-specific management. One technology that can fulfill this need is diffuse reflectance spectroscopy, which measures light reflected from the soil in the visible and near infrared wavelength bands. Reflectance spectroscopy has potential to replace traditional laboratory testing, which is often slow and expensive. Laboratory reflectance spectroscopy, generally using dried soil samples, is well-established. Recently, ruggedized instruments that can take spectroscopy to the field without the need for sample collection have been developed. The goal of this study was to compare the accuracy of two spectrometers for estimating soil carbon. Both were operated in the laboratory under several configurations and one was also used as an in-situ field sensor. We found that both spectrometers performed very similarly in the laboratory and determined that a complete range of reflectance data was not needed for accurate carbon measurements. Although in-situ results were of acceptable accuracy for many uses, they were not as accurate as laboratory results. The findings of this study will be useful to potential users of in-situ soil spectrometers who need to know how results with those instruments compare to laboratory spectrometer results.

Technical Abstract: Visible and near-infrared reflectance spectroscopy satisfies the need for speed and precision in estimation of soil carbon and other soil properties. Previous work has established accuracy of the method, with much of the reported research done using bench spectrometers from a single manufacturer. However, other equipment is now available, capable of both bench and in-situ field operation. Therefore, the objective of this research was to compare two spectrometers, with emphasis on the consistency of soil carbon detection across instruments and determination of the impact wavelength range has on carbon detection. Additionally, the impact of bench versus in-situ probe data acquisition was examined. The spectrometers used were the ASD FieldSpec Pro FR operated as a bench model and the Veris P4000 VIS-NIR-EC-Force Probe operated as both a bench instrument and as a profiling device used in the field. At field sites in Missouri, Nebraska and Pennsylvania, probe spectrometer data were collected, followed by the removal of soil cores at each probe scanning site. The cores were then scanned with each bench top spectrometer. Partial least squares regression was used for statistical analysis and model development. Results showed that C estimation accuracy was very similar for the two spectrometers in bench mode, and that good C estimates were maintained with in-situ operation of the Veris instrument. Wavelength analysis confirmed that the visible spectral range had a negligible effect on C estimation accuracy. These results further validate the effectiveness of reflectance spectroscopy, including in-situ scanning with probe-style devices, for estimation of soil carbon.