Neutron-Gamma Analysis of Soil for Digital Agriculture

*Galina Yakubova, Aleksandr Kavetskiy, Nikolay Sargsyan, Stephen A. Prior and Henry Allen Torbert*

### **Abstract**

This chapter describes technical aspects of neutron stimulated gamma ray analysis of soil carbon. The introduction covers general principles, different modifications of neutron gamma analysis, measurement system configurations, and advantages of this method for soil carbon analysis. Problems with neutron-gamma technology for soil carbon analysis and investigation methods including Monte-Carlo simulation of neutron interaction with soil elements are discussed. Based on investigation results, a method to extract the "soil carbon net peak" from raw acquired data was developed. A direct proportional dependency between the carbon net peak area and average carbon weight percent in the upper 10 cm soil layer for any carbon depth profile was demonstrated. Calibration of the measurement system using sand-carbon pits and field measurements of soil carbon are described. Compared to traditional chemical analysis (dry combustion) data, measurement results demonstrated good agreement between methods. Thus, neutron stimulated gamma ray analysis can be used for *in situ* determination of near surface soil carbon content and is applicable for precision geospatial mapping of soil carbon.

**Keywords:** neutron-gamma analysis, PFTNA, soil carbon storage, soil carbon distribution maps, scanning technology, ArcGIS, IGOR

### **1. Introduction**

Agricultural operations play important roles in productivity and profitability of soil resources, and can influence aspects of climate change and the ability of soil to sequester carbon which has relevance to when quantifying carbon storage for credits. Development of sustainable land use practices requires understanding and evaluating impacts of these practices on soil resources. Exact knowledge of soil chemical composition can improve modern precision agricultural practices. For these reasons, largescale measuring and mapping of soil elements (primarily carbon) on agricultural lands has become important today.

The current "gold standard" method of "dry combustion" for soil carbon determinations is based on extensive analysis of laboratory processed field samples. This method is labor intensive and time consuming. Other techniques (i.e., laser-induced breakdown spectroscopy, near- and mid-infrared spectroscopy, diffuse reflectance infrared Fourier transform spectroscopy, and pyrolysis molecular beam mass

spectrometry) yield carbon values for small soil volumes (0.01–10 cm3 ) near the soil surface [1], which may not be representative of large field areas.

Neutron-gamma analysis (NGA) can overcome these disadvantages and can be used to create soil carbon distribution maps of large field areas. Soil carbon content, expressed in average carbon weight percent in the upper 10 cm soil layer (*Cw%*), can be derived directly from *in-situ* neutron-gamma analysis results. Knowledge of soil density allows for calculating carbon content in mass units. The ratio of carbon in the 10–30 cm topsoil layer of various soil types can be used for estimating soil carbon to a 30 cm depth. This 0–30 topsoil layer is used for estimating carbon sequestered in soil [2].

Modified NGA, particularly Pulse Fast Thermal Neutron Analysis (PFTNA as classified by [3]), can be used for determining soil elemental (C, H, Si, etc.) contents. This can be accomplished by analyzing soil gamma spectra induced by 14.1 MeV neutron pulses. This includes gamma spectra acquired during neutron pulses (i.e., from gamma rays appearing due to inelastic neutron scattering, INS spectra) and between pulses (i.e., from gamma rays appearing due to thermal neutron capture, TNC spectra). Details concerning this methodology have been previously described [4–7].

A custom mobile PFTNA system was developed and constructed for measuring soil carbon in agricultural fields in the scanning regime [4, 6]. A GPS device and specially developed software were added to the mobile system for simultaneous acquisition of gamma signals and geographical positions. Maps of surface soil carbon distribution were constructed utilizing this system in conjunction with IGOR software [8] and ArcMap [9]. Technical aspects of neutron stimulated gamma ray analysis of soil carbon, developed algorithms, methodology and software for data acquisition, data processing, and mapping will be described in this chapter. In addition, factors that affect measurement error and required measurement times will be discussed.

### **2. Materials and methods**

#### **2.1 Physical basis of PFTNA**

NGA is based on registration of gamma rays that appear in soil under neutron irradiation. A neutron generator is used as a neutron flux source. Each soil element issues gamma rays with predefined energies during certain nuclear reactions of that element with neutrons. Detectors register gamma rays as a spectrum that is the dependency of the registered gamma rays vs. their energy. In general, this gamma spectrum consists of many gamma peaks produced by various elements due to different processes of neutron-nuclear interactions and the continuous background. Since some peaks overlap, extraction of gamma peaks that correspond to particular soil elements is difficult. PFTNA can be used to overcome this problem. This method uses the difference in duration of INS (pico- and femto-second intervals) and TNC (dozens microsecond intervals) nuclear reactions to divide the spectra that appear due to these processes. With PFTNA, the neutron generator works in the pulse regime, and the single spectrum acquired is divided into two spectra in two separate memory locations. The INS spectrum, which is the gamma ray spectrum that appears due to inelastic neutron scattering of soil nuclei, is acquired during neutron pulses. The TNC spectrum, which appears due to thermal neutron capture, is acquired between pulses. Examples of these spectra are shown in **Figure 1**.

In the INS spectrum, the gamma peak used for determining soil carbon (centroid at 4.44 MeV) is still a complicated peak. This peak consists of the gamma response

**Figure 1.** *Example of INS and TNC gamma spectra showing some peaks of interest.*

from soil carbon, carbon in measurement system construction materials, and the cascade transition peak of silicon-28. A system background measurement should be conducted to define the gamma response corresponding to carbon in system components. This measurement is conducted under conditions where the effect of soil on the spectra is negligible (i.e., system is lifted to a height of more than 4–6 m above the ground). The silicon-28 cascade transition peak can be defined from determining values of the silicon-28 peak (centroid at 1.78 MeV in the spectra) and the cascade transition coefficient. The net carbon peak area can be computed by removing the background and silicon portions from the carbon peak. The net carbon peak area is directly associated with the average carbon content in the upper 10 cm soil layer expressed in weight percent [5]. This is true for any soil type regardless of carbon distribution shape. To relate net carbon peak area with average soil carbon content, corresponding calibration measurements (PFTNA measurements of model soil samples with well-known carbon contents) should be performed. Such measurements are needed to develop an equation for calculating soil carbon content from measured net carbon peak areas.

#### **2.2 Carbon content returned by PFTNA**

Since carbon distribution in soil is not uniform or known, several carbon content parameters can directly affect PFTNA measurements. Since soil carbon can sometimes be characterized as carbon surface density in the 30 cm soil layer, it was assumed that the PFTNA system acquired gammas from 30 cm soil layer from irradiation by 14 MeV neutrons [1]. However, Monte-Carlo simulations did not confirm this assumption for unpredictable soil carbon distributions and soil densities [5].

Carbon content can be expressed as the average carbon weight percent in a given soil layer. Previous work showed [5] that soil carbon (expressed in weight percent in 10 cm upper soil layer) can be directly estimated from PFTNA gamma spectra measurements and corresponding peak area calculations. This can be done by applying previously defined calibration dependency using homogeneous reference samples. Workability of this expression for any type of soil with any soil carbon distribution shape with depth was confirmed using Monte-Carlo (Geant4, [10]) simulations. In addition, experimental measurements in agricultural fields were confirmed by comparison to traditional soil chemical analysis results.

#### **2.3 PFTNA system design**

To conduct soil carbon field measurements, a mobile PFTNA system was constructed on a platform (75 cm 23 cm 95 cm; 300 kg) for towing by all-terrain vehicles over agricultural fields. The PFTNA system consisted of a MP320 pulsed neutron generator (NG; Thermo Fisher Scientific, Colorado Springs, CO), three 12.7 cm 12.7 cm 15.2 cm NaI(Tl) scintillation detectors (Scionix USA, Orlando, FL) with corresponding electronics (XIA LLC, Hayward, CA), a R2D-410 neutron detector (Bridgeport Instruments, LLC, Austin, TX), a power system (four 12 V 105 Ah DC105–12 batteries; a CGL 600 W-series DC-AC Inverter, Nova Electric, Bergenfield, NJ; and a PS4Quad Pro Charger, Pro Charging Systems, LLC, LaVergne, TN), a GPS device, an operational laptop, and an Android tablet. Iron and boric acid shielding is placed between the NG and gamma detectors to reduce irradiation of gamma detectors by fast neutrons (**Figure 2**).

The power system supplies all electronic equipment with 110 V AC voltage. Uninterrupted working time is 20 h.

The neutron generator produces a pulsed output of 10<sup>7</sup> –10<sup>8</sup> n s<sup>1</sup> depending on parameter settings; neutron energy is 14 MeV.

**Figure 2.** *Scheme of the PFTNA system.*

Specially developed software allows spectral acquisition and defines the time interval for saving spectra to the laptop hard drive. This software also reads and saves GPS coordinates of the PFTNA system during scanning.

The tablet is mounted in the towing vehicle and is used for GPS set up and tracking system movement.
