No abstract

No abstract

No abstract

Laser-induced breakdown spectroscopy (LIBS) is an emerging elemental analysis technology with the potential to provide rapid, accurate, and precise analysis ofsoil constituents. We evaluated the accuracy of LIBS in measuring soil profile C for field-moist, intact soil cores by interrogating 78 intact soil cores from three Montana agricultural fields. Samples from each core were analyzed in the laboratory for total C (TC). inorganic C (IC), and soil organic C (SOC). Partial least squares 2 (PLS2) regression calibration models were derived using 58 cores (227 samples) and independendy validated at the field scale with the remaining 20 cores (79 samples). We obtained the best LIBS validation predictions for IC(r^= 0.66, standard error of prediction [SEP] =5.3 gkg~', ratio product differential [RPD]= 1.7) followed by TC (r^ = 0.63, SEP = 6.0 gkg~', RPD = 1.6) and SOC (r^ = 0.22, SEP = 3.2 gkg"'. RPD = 1.1).Although the SEP for SOC was less than for TC and IC, low SOC variance limited our ability to evaluate LIBS SOC prediction capabilities. Regression coefficients from LIBS PLS2 models suggested a reliance on stoichiometric relationships between C and other elements related to total and inorganic C in the soil matrix (e.g., Ca, Mg, and Si) to discriminare TC from IC. Results indicate that LIBS spectral data collected on intact soil cores can be calibrated to accurately estimate and difiïrentiate between soil total and inorganic C concentrations at the field scale.Abbreviations: IC. inorganic carbon; LIBS, la.ser-induced breakdown spectroscopy; MSD, mean squared deviation; NU, nonunity regression line; PLS1, partial least squares 1 ; PLS2, partial least squares 2; RPD, ratio product differential, SB, squared bias; SEL, standard error of laboratory measurement; SEP, standard error of prediction; SOC, soil organic carbon; TC, total carbon; VisNIR, visible and near infrared.

No abstract

Core Ideas First rigorous integration of VisNIR and LIBS spectroscopic techniques for soil profile C measurement. Field‐moist intact soil cores scanned with LIBS and VisNIR simulating in situ measurements. Combined VisNIR‐LIBS wavelengths selected by MRCE were split nearly equally between sensors, suggesting molecular and elemental information was successfully integrated for predicting TC, IC, and SOC. Stoichiometric relationships in LIBS elemental emissions are more strongly correlated to IC concentration than were carbonate absorptions in VisNIR. Lack of variation and consistent, definable chemical composition and stoichiometry likely contributed to poor LIBS and combined VisNIR – LIBS prediction accuracy for SOC. Soil organic carbon (SOC) measurement is critically important to quantify regional and global soil C stocks and better understand soil C biogeochemical processes. Recent studies employing laser‐induced breakdown spectroscopy (LIBS) and visible‐near infrared diffuse reflectance spectroscopy (vis–NIRS) indicate their potential for in situ soil C determination. Visible and near infrared diffuse reflectance spectroscopy and LIBS spectroscopy fundamentally differ and we hypothesize that their integration would provide improved soil C predictions. We report the first rigorous integration of vis–NIRS and LIBS, evaluating the precision of vis–NIRS, LIBS, and combined vis–NIRS‐LIBS spectra for simulated in situ soil profile total C (TC), inorganic C (IC) and SOC measurement. Three multivariate variable selection and regression approaches were evaluated for soil C prediction. The highest soil C prediction accuracies were observed using multivariate regression with covariance estimation (MRCE). Inorganic C was best predicted by LIBS, vis–NIRS provided better SOC predictions, and TC was best predicted using combined vis–NIRS‐LIBS data. Combined vis–NIRS‐LIBS did not consistently increase soil C prediction accuracy. Soil organic C was not well predicted, presumably due to challenges associated with scanning surfaces of intact soil cores, variable SOC chemistries, and low SOC variation in the dataset. Considering the challenging conditions under which combined vis–NIRS– LIBS was tested for soil C measurement, data integration and model calibrations had acceptable performance. Further testing under more controlled soil conditions with samples containing greater SOC diversity is necessary to determine the technical potential of combined vis–NIRS/LIBS for soil C determination.

Laser-induced breakdown spectroscopy (LIBS) provides a potential method for rapid, in situ soil C measurement. In previous research on the application of LIBS to intact soil cores, we hypothesized that ultraviolet (UV) spectrum LIBS (200-300 nm) might not provide sufficient elemental information to reliably discriminate between soil organic C (SOC) and inorganic C (IC). In this study, using a custom complete spectrum (245-925 nm) core-scanning LIBS instrument, we analyzed 60 intact soil cores from six wheat fields. Predictive multi-response partial least squares (PLS2) models using full and reduced spectrum LIBS were compared for directly determining soil total C (TC), IC, and SOC. Two regression shrinkage and variable selection approaches, the least absolute shrinkage and selection operator (LASSO) and sparse multivariate regression with covariance estimation (MRCE), were tested for soil C predictions and the identification of wavelengths important for soil C prediction. Using complete spectrum LIBS for PLS2 modeling reduced the calibration standard error of prediction (SEP) 15 and 19% for TC and IC, respectively, compared to UV spectrum LIBS. The LASSO and MRCE approaches provided significantly improved calibration accuracy and reduced SEP 32-55% over UV spectrum PLS2 models. We conclude that (1) complete spectrum LIBS is superior to UV spectrum LIBS for predicting soil C for intact soil cores without pretreatment; (2) LASSO and MRCE approaches provide improved calibration prediction accuracy over PLS2 but require additional testing with increased soil and target analyte diversity; and (3) measurement errors associated with analyzing intact cores (e.g., sample density and surface roughness) require further study and quantification.

No abstract