Estimation of Functional Soil Organic Carbon Pools with Diffuse Reflectance Spectroscopy.

There are increasing demands to provide accurate soil carbon data to assess ecosystem services and emerging carbon markets. To characterize the heterogeneity of soil carbon properties and processes it is critical to map them at an appropriate spatial scale, which translates into high density sampling. However, investments in labor and wet chemistry analysis needed to match the scale of carbon processes may offset the value of ecosystem services and carbon credits to soil managers. Visible/near infrared diffuse reflectance spectroscopy (VNIR-DRS) has recently become popular as an analytical method to rapidly and efficiently measure soil carbon. Extending the high throughput capability of VNIR-DRS to include rapid estimation of functional soil carbon pools could dramatically improve our ability to measure, model, and track soil carbon change. A large set of surface soil samples (n=~1000) were gathered from across Florida (0-20 cm). These samples are analyzed for total carbon by combustion analyzer, recalcitrant carbon by acid hydrolysis, and labile carbon by hot water extraction. Samples are scanned with a VNIR spectrometer and various modeling techniques are used to calibrate the sensor to the measured carbon functional pools, validated by an independent dataset. Our approach shows promise to provide relatively accurate predictions of various carbon pools at reasonable cost to be utilized for carbon tracking/monitoring across large landscapes.