|KAVETSKIY, ALEXANDER - Auburn University|
|Prior, Stephen - Steve|
|Torbert, Henry - Allen|
Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 9/3/2014
Publication Date: 9/3/2014
Citation: Kavetskiy, A., Yakubova, G.N., Prior, S.A., Torbert III, H.A. 2014. Monte-Carlo simulation of soil carbon measurements by inelastic neutron scattering [abstract]. Complex Soil Systems Conference: A Path to Improved Understanding of Complex Soil Systems. p. 43.
Technical Abstract: Measuring soil carbon is critical for assessing the potential impact of different land management practices on carbon sequestration. The inelastic neutron scattering (INS) of fast neutrons (with energy around 14 MeV) on carbon-12 nuclei produces gamma rays with energy of 4.43 MeV; this gamma flux can be directly used for determining soil carbon. INS has several advantages over current well known methods of carbon determination (i.e., dry combustion, laser induced breakdown spectroscopy, mid and near infrared reflectance spectroscopy) such as: the method is non-destructive, no sample preparation is required, and a large soil volume is analyzed in a single measurement. While the theory and methodology of INS has been developed (Wielopolski et al., Brookhaven National Laboratory, 2000-2012), the development and testing of experimental equipment for directly measuring field soil carbon is currently being done at the USDA-ARS National Soil Dynamics Laboratory in Auburn, AL. To increase the accuracy and reliability of this method, a Monte-Carlo simulation of neutron transport and interaction with soil nuclei can be used. The Geant-4 tool kit (S. Agostinelli et al. 2003) makes it possible to develop software for conducting such simulations. In this work, the Geant4 based simulation software for neutron transport and interaction with soil nuclei combined with the acquisition of gamma rays by a system similar (i.e., neutron energy, geometry size, detector type and size) to the National Soil Dynamics Laboratory INS system was developed. Using this software, neutron induced gamma spectra were simulated for soil samples of different sizes, with differing amounts of carbon, and with different amounts of other soil components. Simulation results demonstrated good agreement with experimental data in the dependence of the carbon 4.43 MeV peak areas to the carbon content in soil samples for any sample thickness, or the silicon weight percentage of the soil. This agreement verifies the usefulness of this simulation method and offers the possibility of utilizing this method for the analysis of other factors influencing soil carbon measurement by the INS method. For example, the effects of soil moisture, density, and carbon distribution with depth were simulated. The results of these simulations will be discussed.