|Reeves iii, James|
Submitted to: World Congress of Soil Science
Publication Type: Abstract only
Publication Acceptance Date: 8/14/2002
Publication Date: N/A
Citation: Interpretive Summary: Summary.
Technical Abstract: The objective was to evaluate the ability of spectroscopy to quantify soil carbon. Samples were scanned in the NIR on a NIRSystems 6500 scanning monochromator, and in the mid-IR using a DigiLab FTS-60 FTIR. All samples were scanned using diffuse reflectance and non-KBr diluted samples. In the NIR, various samples were scanned using a variety of sample devices including: a rotating sample cup, large sample transport device (in polyethylene bags), and fiber optic probe. In addition, samples were also scanned as is (natural field moisture content), dried but not ground, and dried and ground. In the mid-IR, dried, ground samples were scanned using a custom made sample transport device designed to increase the sample area scanned. Results demonstrated that: 1) Both inorganic- and organic-C can be accurately determined using either spectral range, but mid-IR calibrations are almost always the more accurate. 2) Mid-IR calibrations are generally more robust (less affected by sample diversity) than are NIR calibrations. 3) While accurate NIR calibrations can be developed using fiber optic probes, there is a decrease in calibration accuracy. 4. The use of polyethylene bags as sample holders appears to degrade NIR calibrations for dried, but not moist, soils. Finally, preliminary results indicate that accurate mid-IR calibrations covering a wide range of soil types are feasible. In conclusion, mid-IR and NIR spectroscopy offer the possibility of rapid, inexpensive, and accurate methods for the determination of soil carbon.