Modelling soil carbon fractions with visible near-infrared (VNIR) and mid-infrared (MIR) spectroscopy

Knox, Nichola M.; Grunwald, Sabine; McDowell, M. L.; Bruland, Gregory L.; Myers, D. Brenton; Harris, Willie G.

doi:10.1016/j.geoderma.2014.10.019

Cited by 140 publications

(96 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The practical evaluation of overall soil quality, however, still remains a challenging issue (Gholizadeh et al, 2013). Because infrared spectroscopic techniques are sensitive to specific molecular vibrations of both organic and mineral soil components and can also measure accurately several physical, chemical and biological properties at the same time, they are widely recognized as suitable tools for assessing and monitoring soil (Viscarra Rossel et al, 2006;Knox et al, 2015;Ludwig et al, 2015). Although inherent differences in the near-infrared (NIR) and mid-infrared (MIR) spectral regions occur (for example in the NIR region overlapping combinations and overtone peaks may appear, and in the MIR region many bands correspond to simple compounds) (Bellon-Maurel & McBratney, 2011), both approaches have been used widely for qualitative and quantitative analysis of soil (Reeves, 2010;Akroume et al, 2016).…”

Section: Introductionmentioning

confidence: 99%

Tree genotype and seasonal effects on soil properties and biogeochemical functioning in Mediterranean pine forests

Pérez‐Izquierdo

Saint‐André²,

Santenoise³

et al. 2018

European J Soil Science

View full text Add to dashboard Cite

Summary In forests, intraspecific genetic variation in trees can affect the entire ecosystem, which in turn depends on the different processes occurring through space and time in soil. We hypothesized that, in addition to the effect of the local site, tree genotype and season would have an effect on the properties and functions of the edaphic environment. We studied soils beneath different genotypes of Pinus pinaster Ait. (Atlantic, Mediterranean and African) in 45‐year‐old common gardens in spring and autumn. The pH, organic matter, nutrients and infrared spectroscopy, together with enzyme activities, were determined and used to evaluate the soil properties and biogeochemical functioning. In addition to strong site effects, tree genotype and seasonal effects were detected on soil properties and functions. Both were major controlling factors of microbially mediated functioning, especially processes related to the carbon (C) and nitrogen (N) cycles. In general, the soil environment beneath the Atlantic trees was different from that under the Mediterranean and African genotypes, with differences in raw infrared spectra and increased activities of enzymes involved in hemicellulose degradation and N mobilization. Regarding the season, the largest soil humidity (RH), electric conductivity (EC), and N and potassium (K) concentrations, coupled with the smallest phosphorus (P) concentration and C:N ratios, were detected in autumn. Degradation of C peaked in autumn, whereas P and N mobilization usually peaked in spring. Our results showed that, beyond local site effects, there were detectable effects of tree genotype and season in Mediterranean forest soils, which governed microbially mediated processes that might have relevant functional consequences at the ecosystem level. Highlights Tree genotype and season shape the features and functioning of Mediterranean forest soils. Tree genotype, season and site effects on soil are identified separately under field conditions. Intraspecific genetic variation of trees defines soil properties and functioning. Tree genotype and season are key regulators of the edaphic environment in Mediterranean forests.

show abstract

Section: Introductionmentioning

confidence: 99%

Tree genotype and seasonal effects on soil properties and biogeochemical functioning in Mediterranean pine forests

Pérez‐Izquierdo

Saint‐André²,

Santenoise³

et al. 2018

European J Soil Science

View full text Add to dashboard Cite

show abstract

“…Its benefits are rapidity and energy efficiency. Soil properties can be assessed with infrared spectroscopy by analysing spectral curves according to spectral scans, and it is possible to process a large number of soil samples with this rapid and non‐polluting technique (Knox et al ., ).…”

Section: Introductionmentioning

confidence: 97%

“…Mid‐infrared spectroscopy (4000–400 cm −1 ) has been used frequently to predict soil properties such as soil organic carbon (SOC), total carbon (TC), soil organic matter (SOM), cation exchange capacity (CEC), clay, phosphorus (P), potassium (K) and others (Knox et al ., ; Terra et al ., ). Moreover, many of the functional groups that are typically characterized by distinct mid‐infrared regions (e.g.…”

Section: Introductionmentioning

confidence: 99%

Investigation of soil properties using different techniques of mid‐infrared spectroscopy

Zhang

Shen

2018

European J Soil Science

View full text Add to dashboard Cite

Summary Diffuse reflectance spectroscopy (DRF), attenuated total reflectance spectroscopy (ATR) and Fourier transform infrared photoacoustic spectroscopy (PAS) are useful techniques that are commonly used to characterize soil properties. Here, these techniques were used to evaluate individual properties of 1456 paddy soil samples from Nanjing, China, to explore their features in the mid‐infrared range. The diffuse reflectance spectroscopy showed broad and shouldered peaks that ranged from 4000 to 500 cm−1. Attenuated total reflectance spectroscopy produced main peaks around 1000 cm−1. Fourier transform infrared photoacoustic spectroscopy spectra showed several moderate intensity peaks in the regions of 4000–2600 and 2500–500 cm−1. The effective variables (wavenumbers) were selected using an uninformative variable elimination (UVE) algorithm to predict soil pH, and soil organic matter (SOM), total nitrogen (TN) and available phosphorus (AP) contents; it used a self‐adaptive partial least squares model (SAM–PLS). The results showed that selected wavenumbers improved the accuracy of prediction for pH, and SOM, TN and AP contents. For the prediction of pH, ATR spectra showed marginally better performance than the other spectra. The DRF spectroscopy was the optimal method for SOM and TN prediction. Available phosphorus contents were predicted poorly by all three spectra. Furthermore, mid‐infrared incident light paths, absorption modes, sensing styles, as well as peak interference and the prediction model, were the main factors that affected the assessment of soil properties in the three types of spectroscopy. Highlights Soil spectra data were measured with DRF, ATR and PAS spectrometers over the range 4000–500 cm−1. A UVE algorithm was used to select effective wavenumbers from the full spectrum. The pH, SOM, TN and AP were estimated using three techniques based on a SAM–PLS model. Mid‐infrared incident light paths and absorption modes affected prediction performance of the three techniques.

show abstract

“…Fontán et al [9] found higher correlation values for inorganic C (R 2 = 0.76) than organic C (R 2 = 0.67) in a vertisol to a depth of 90 cm. In addition, successful predictions of soil carbon fractions have been made by VIS-NIR spectroscopy [10,11], suggesting this technique as a reliable alternative for assessing the impact of land use change on soil carbon pools. Other related studies show the performance of different multivariate methods in calibrating models [12][13][14].…”

Section: Introductionmentioning

confidence: 99%

Potential Model Overfitting in Predicting Soil Carbon Content by Visible and Near-Infrared Spectroscopy

et al. 2017

View full text Add to dashboard Cite

Soil spectroscopy is known as a rapid and cost-effective method for predicting soil properties from spectral data. The objective of this work was to build a statistical model to predict soil carbon content from spectral data by partial least squares regression using a limited number of soil samples. Soil samples were collected from two soil orders (Andisol and Ultisol), where the dominant land cover is native Nothofagus forest. Total carbon was analyzed in the laboratory and samples were scanned using a spectroradiometer. We found evidence that the reflectance was influenced by soil carbon content, which is consistent with the literature. However, the reflectance was not useful for building an appropriate regression model. Thus, we report here intriguing results obtained in the calibration process that can be confusing and misinterpreted. For instance, using the Savitzky-Golay filter for pre-processing spectral data, we obtained R 2 = 0.82 and root-mean-squared error (RMSE) = 0.61% in model calibration. However, despite these values being comparable with those of other similar studies, in the cross-validation procedure, the data showed an unusual behavior that leads to the conclusion that the model overfits the data. This indicates that the model should not be used on unobserved data.

show abstract

Modelling soil carbon fractions with visible near-infrared (VNIR) and mid-infrared (MIR) spectroscopy

Cited by 140 publications

References 43 publications

Tree genotype and seasonal effects on soil properties and biogeochemical functioning in Mediterranean pine forests

Tree genotype and seasonal effects on soil properties and biogeochemical functioning in Mediterranean pine forests

Investigation of soil properties using different techniques of mid‐infrared spectroscopy

Potential Model Overfitting in Predicting Soil Carbon Content by Visible and Near-Infrared Spectroscopy

Contact Info

Product

Resources

About