Determination of total soil organic C and hot water-extractable C from VIS-NIR soil reflectance with partial least squares regression and spectral feature selection techniques
Identifieur interne : 000404 ( PascalFrancis/Corpus ); précédent : 000403; suivant : 000405Determination of total soil organic C and hot water-extractable C from VIS-NIR soil reflectance with partial least squares regression and spectral feature selection techniques
Auteurs : M. Vohland ; C. EmmerlingSource :
- European journal of soil science [ 1351-0754 ] ; 2011.
Descripteurs français
- Pascal (Inist)
English descriptors
- KwdEn :
Abstract
The calibration of soil organic C (SOC) and hot water-extractable C (HWE-C) from visible and near-infrared soil reflectance spectra is hindered by the complex spectral interaction of soil chromophores that usually varies from one soil or soil type to another. The exploitation of spectral variables from spectroradiometer data is further affected by multicollinearity and noise. In this study, a set of soil samples (Fluvisols, Podzols, Cambisols and Chernozems; n = 48) representing a wide range of properties was analysed. Spectral readings with a fibre-optics visible to near-infrared instrument were used to estimate SOC and HWE-C contents by partial least squares regression (PLS). In addition to full-spectrum PLS, spectral feature selection techniques were applied with PLS (uninformative variable elimination, UVE-PLS, and a genetic algorithm, GA-PLS). On the basis of normalized spectra (mean centring + vector normalization), the order of prediction accuracy was GA-PLS » UVE-PLS > PLS for SOC; for HWE-C, it was GA-PLS > UVE-PLS, PLS. With GA-PLS, acceptable cross-validated (cv) prediction accuracies were obtained for the complete dataset (SOC, R2cv = 0.83, RPDcv = 2.42; HWE-Ccv, R2cv = 0.78, RPDcv = 2.13). Splitting the soil data into two groups with different basic properties (Podzols compared with Fluvisols/Cambisols; n = 21 and n = 23, respectively) improved SOC predictions with GA-PLS distinctly (Podzols, R2cv = 0.89, RPDcv = 3.14; Fluvisols/Cambisols, R2cv = 0.92, RPDcv = 3.64). This demonstrates the importance of using stratified models for successful quantitative approaches after an initial rough screening. GA selection frequencies suggest that the spectral region over 1900 nm, and in particular the hydroxyl band at 2200 nm are of great importance for the spectral prediction of both SOC and HWE-C.
Notice en format standard (ISO 2709)
Pour connaître la documentation sur le format Inist Standard.
| pA |
|
|---|
Format Inist (serveur)
| NO : | PASCAL 11-0335983 INIST |
|---|---|
| ET : | Determination of total soil organic C and hot water-extractable C from VIS-NIR soil reflectance with partial least squares regression and spectral feature selection techniques |
| AU : | VOHLAND (M.); EMMERLING (C.) |
| AF : | Faculty of Geography and Geosciences, Remote Sensing Department, University of Trier/54286 Trier/Allemagne (1 aut.); Faculty of Geography and Geosciences, Soil Science Department, University of Trier/54286 Trier/Allemagne (2 aut.) |
| DT : | Publication en série; Niveau analytique |
| SO : | European journal of soil science; ISSN 1351-0754; Royaume-Uni; Da. 2011; Vol. 62; No. 4; Pp. 598-606; Bibl. 3/4 p. |
| LA : | Anglais |
| EA : | The calibration of soil organic C (SOC) and hot water-extractable C (HWE-C) from visible and near-infrared soil reflectance spectra is hindered by the complex spectral interaction of soil chromophores that usually varies from one soil or soil type to another. The exploitation of spectral variables from spectroradiometer data is further affected by multicollinearity and noise. In this study, a set of soil samples (Fluvisols, Podzols, Cambisols and Chernozems; n = 48) representing a wide range of properties was analysed. Spectral readings with a fibre-optics visible to near-infrared instrument were used to estimate SOC and HWE-C contents by partial least squares regression (PLS). In addition to full-spectrum PLS, spectral feature selection techniques were applied with PLS (uninformative variable elimination, UVE-PLS, and a genetic algorithm, GA-PLS). On the basis of normalized spectra (mean centring + vector normalization), the order of prediction accuracy was GA-PLS » UVE-PLS > PLS for SOC; for HWE-C, it was GA-PLS > UVE-PLS, PLS. With GA-PLS, acceptable cross-validated (cv) prediction accuracies were obtained for the complete dataset (SOC, R2cv = 0.83, RPDcv = 2.42; HWE-Ccv, R2cv = 0.78, RPDcv = 2.13). Splitting the soil data into two groups with different basic properties (Podzols compared with Fluvisols/Cambisols; n = 21 and n = 23, respectively) improved SOC predictions with GA-PLS distinctly (Podzols, R2cv = 0.89, RPDcv = 3.14; Fluvisols/Cambisols, R2cv = 0.92, RPDcv = 3.64). This demonstrates the importance of using stratified models for successful quantitative approaches after an initial rough screening. GA selection frequencies suggest that the spectral region over 1900 nm, and in particular the hydroxyl band at 2200 nm are of great importance for the spectral prediction of both SOC and HWE-C. |
| CC : | 001E01P03; 002A32B03B3; 226C03 |
| FD : | Détermination; Eau chaude; Procédé extraction; Spectrométrie visible; Spectrométrie IR proche; Spectrométrie réflexion; Facteur réflexion; Rayonnement IR proche; Régression PLS; Carbone organique; Sol; Technique; Science terre; Science du sol; Sélection forme |
| FG : | Méthode optique; Propriété optique; Onde électromagnétique; Régression statistique; Méthode physique; Propriété physique; Matière organique; Analyse statistique |
| ED : | Determination; Hot water; Extraction process; visible spectroscopy; near infrared spectroscopy; Reflection spectrometry; Reflectance; near infrared radiation; PLS regression; organic carbon; soils; techniques; Earth science; Soil science; Feature selection |
| EG : | optical methods; optical properties; electromagnetic waves; regression analysis; physical methods; physical properties; organic materials; statistical analysis |
| SD : | Determinación; Agua caliente; Procedimiento extracción; Espectrometría reflexión; Coeficiente reflexión; Regresión PLS; Carbono orgánico; Suelo; Ciencia tierra; Ciencia del suelo; Selección de formas |
| LO : | INIST-2402.354000509403950110 |
| ID : | 11-0335983 |
Links to Exploration step
Pascal:11-0335983Le document en format XML
<record><TEI><teiHeader><fileDesc><titleStmt><title xml:lang="en" level="a">Determination of total soil organic C and hot water-extractable C from VIS-NIR soil reflectance with partial least squares regression and spectral feature selection techniques</title><author><name sortKey="Vohland, M" sort="Vohland, M" uniqKey="Vohland M" first="M." last="Vohland">M. Vohland</name><affiliation><inist:fA14 i1="01"><s1>Faculty of Geography and Geosciences, Remote Sensing Department, University of Trier</s1><s2>54286 Trier</s2><s3>DEU</s3><sZ>1 aut.</sZ></inist:fA14></affiliation></author><author><name sortKey="Emmerling, C" sort="Emmerling, C" uniqKey="Emmerling C" first="C." last="Emmerling">C. Emmerling</name><affiliation><inist:fA14 i1="02"><s1>Faculty of Geography and Geosciences, Soil Science Department, University of Trier</s1><s2>54286 Trier</s2><s3>DEU</s3><sZ>2 aut.</sZ></inist:fA14></affiliation></author></titleStmt><publicationStmt><idno type="wicri:source">INIST</idno><idno type="inist">11-0335983</idno><date when="2011">2011</date><idno type="stanalyst">PASCAL 11-0335983 INIST</idno><idno type="RBID">Pascal:11-0335983</idno><idno type="wicri:Area/PascalFrancis/Corpus">000404</idno></publicationStmt><sourceDesc><biblStruct><analytic><title xml:lang="en" level="a">Determination of total soil organic C and hot water-extractable C from VIS-NIR soil reflectance with partial least squares regression and spectral feature selection techniques</title><author><name sortKey="Vohland, M" sort="Vohland, M" uniqKey="Vohland M" first="M." last="Vohland">M. Vohland</name><affiliation><inist:fA14 i1="01"><s1>Faculty of Geography and Geosciences, Remote Sensing Department, University of Trier</s1><s2>54286 Trier</s2><s3>DEU</s3><sZ>1 aut.</sZ></inist:fA14></affiliation></author><author><name sortKey="Emmerling, C" sort="Emmerling, C" uniqKey="Emmerling C" first="C." last="Emmerling">C. Emmerling</name><affiliation><inist:fA14 i1="02"><s1>Faculty of Geography and Geosciences, Soil Science Department, University of Trier</s1><s2>54286 Trier</s2><s3>DEU</s3><sZ>2 aut.</sZ></inist:fA14></affiliation></author></analytic><series><title level="j" type="main">European journal of soil science</title><title level="j" type="abbreviated">Eur. j. soil sci.</title><idno type="ISSN">1351-0754</idno><imprint><date when="2011">2011</date></imprint></series></biblStruct></sourceDesc><seriesStmt><title level="j" type="main">European journal of soil science</title><title level="j" type="abbreviated">Eur. j. soil sci.</title><idno type="ISSN">1351-0754</idno></seriesStmt></fileDesc><profileDesc><textClass><keywords scheme="KwdEn" xml:lang="en"><term>Determination</term><term>Earth science</term><term>Extraction process</term><term>Feature selection</term><term>Hot water</term><term>PLS regression</term><term>Reflectance</term><term>Reflection spectrometry</term><term>Soil science</term><term>near infrared radiation</term><term>near infrared spectroscopy</term><term>organic carbon</term><term>soils</term><term>techniques</term><term>visible spectroscopy</term></keywords><keywords scheme="Pascal" xml:lang="fr"><term>Détermination</term><term>Eau chaude</term><term>Procédé extraction</term><term>Spectrométrie visible</term><term>Spectrométrie IR proche</term><term>Spectrométrie réflexion</term><term>Facteur réflexion</term><term>Rayonnement IR proche</term><term>Régression PLS</term><term>Carbone organique</term><term>Sol</term><term>Technique</term><term>Science terre</term><term>Science du sol</term><term>Sélection forme</term></keywords></textClass></profileDesc></teiHeader><front><div type="abstract" xml:lang="en">The calibration of soil organic C (SOC) and hot water-extractable C (HWE-C) from visible and near-infrared soil reflectance spectra is hindered by the complex spectral interaction of soil chromophores that usually varies from one soil or soil type to another. The exploitation of spectral variables from spectroradiometer data is further affected by multicollinearity and noise. In this study, a set of soil samples (Fluvisols, Podzols, Cambisols and Chernozems; n = 48) representing a wide range of properties was analysed. Spectral readings with a fibre-optics visible to near-infrared instrument were used to estimate SOC and HWE-C contents by partial least squares regression (PLS). In addition to full-spectrum PLS, spectral feature selection techniques were applied with PLS (uninformative variable elimination, UVE-PLS, and a genetic algorithm, GA-PLS). On the basis of normalized spectra (mean centring + vector normalization), the order of prediction accuracy was GA-PLS » UVE-PLS > PLS for SOC; for HWE-C, it was GA-PLS > UVE-PLS, PLS. With GA-PLS, acceptable cross-validated (cv) prediction accuracies were obtained for the complete dataset (SOC, R<sup>2</sup><sub>cv </sub>= 0.83, RPD<sub>cv</sub> = 2.42; HWE-C<sub>cv</sub>, R<sup>2</sup><sub>cv</sub> = 0.78, RPD<sub>cv</sub> = 2.13). Splitting the soil data into two groups with different basic properties (Podzols compared with Fluvisols/Cambisols; n = 21 and n = 23, respectively) improved SOC predictions with GA-PLS distinctly (Podzols, R<sup>2</sup><sub>cv</sub> = 0.89, RPD<sub>cv</sub> = 3.14; Fluvisols/Cambisols, R<sup>2</sup><sub>cv</sub> = 0.92, RPD<sub>cv</sub> = 3.64). This demonstrates the importance of using stratified models for successful quantitative approaches after an initial rough screening. GA selection frequencies suggest that the spectral region over 1900 nm, and in particular the hydroxyl band at 2200 nm are of great importance for the spectral prediction of both SOC and HWE-C.</div></front></TEI><inist><standard h6="B"><pA><fA01 i1="01" i2="1"><s0>1351-0754</s0></fA01><fA03 i2="1"><s0>Eur. j. soil sci.</s0></fA03><fA05><s2>62</s2></fA05><fA06><s2>4</s2></fA06><fA08 i1="01" i2="1" l="ENG"><s1>Determination of total soil organic C and hot water-extractable C from VIS-NIR soil reflectance with partial least squares regression and spectral feature selection techniques</s1></fA08><fA11 i1="01" i2="1"><s1>VOHLAND (M.)</s1></fA11><fA11 i1="02" i2="1"><s1>EMMERLING (C.)</s1></fA11><fA14 i1="01"><s1>Faculty of Geography and Geosciences, Remote Sensing Department, University of Trier</s1><s2>54286 Trier</s2><s3>DEU</s3><sZ>1 aut.</sZ></fA14><fA14 i1="02"><s1>Faculty of Geography and Geosciences, Soil Science Department, University of Trier</s1><s2>54286 Trier</s2><s3>DEU</s3><sZ>2 aut.</sZ></fA14><fA20><s1>598-606</s1></fA20><fA21><s1>2011</s1></fA21><fA23 i1="01"><s0>ENG</s0></fA23><fA43 i1="01"><s1>INIST</s1><s2>2402</s2><s5>354000509403950110</s5></fA43><fA44><s0>0000</s0><s1>© 2011 INIST-CNRS. All rights reserved.</s1></fA44><fA45><s0>3/4 p.</s0></fA45><fA47 i1="01" i2="1"><s0>11-0335983</s0></fA47><fA60><s1>P</s1></fA60><fA61><s0>A</s0></fA61><fA64 i1="01" i2="1"><s0>European journal of soil science</s0></fA64><fA66 i1="01"><s0>GBR</s0></fA66><fC01 i1="01" l="ENG"><s0>The calibration of soil organic C (SOC) and hot water-extractable C (HWE-C) from visible and near-infrared soil reflectance spectra is hindered by the complex spectral interaction of soil chromophores that usually varies from one soil or soil type to another. The exploitation of spectral variables from spectroradiometer data is further affected by multicollinearity and noise. In this study, a set of soil samples (Fluvisols, Podzols, Cambisols and Chernozems; n = 48) representing a wide range of properties was analysed. Spectral readings with a fibre-optics visible to near-infrared instrument were used to estimate SOC and HWE-C contents by partial least squares regression (PLS). In addition to full-spectrum PLS, spectral feature selection techniques were applied with PLS (uninformative variable elimination, UVE-PLS, and a genetic algorithm, GA-PLS). On the basis of normalized spectra (mean centring + vector normalization), the order of prediction accuracy was GA-PLS » UVE-PLS > PLS for SOC; for HWE-C, it was GA-PLS > UVE-PLS, PLS. With GA-PLS, acceptable cross-validated (cv) prediction accuracies were obtained for the complete dataset (SOC, R<sup>2</sup><sub>cv </sub>= 0.83, RPD<sub>cv</sub> = 2.42; HWE-C<sub>cv</sub>, R<sup>2</sup><sub>cv</sub> = 0.78, RPD<sub>cv</sub> = 2.13). Splitting the soil data into two groups with different basic properties (Podzols compared with Fluvisols/Cambisols; n = 21 and n = 23, respectively) improved SOC predictions with GA-PLS distinctly (Podzols, R<sup>2</sup><sub>cv</sub> = 0.89, RPD<sub>cv</sub> = 3.14; Fluvisols/Cambisols, R<sup>2</sup><sub>cv</sub> = 0.92, RPD<sub>cv</sub> = 3.64). This demonstrates the importance of using stratified models for successful quantitative approaches after an initial rough screening. GA selection frequencies suggest that the spectral region over 1900 nm, and in particular the hydroxyl band at 2200 nm are of great importance for the spectral prediction of both SOC and HWE-C.</s0></fC01><fC02 i1="01" i2="2"><s0>001E01P03</s0></fC02><fC02 i1="02" i2="X"><s0>002A32B03B3</s0></fC02><fC02 i1="03" i2="2"><s0>226C03</s0></fC02><fC03 i1="01" i2="X" l="FRE"><s0>Détermination</s0><s5>01</s5></fC03><fC03 i1="01" i2="X" l="ENG"><s0>Determination</s0><s5>01</s5></fC03><fC03 i1="01" i2="X" l="SPA"><s0>Determinación</s0><s5>01</s5></fC03><fC03 i1="02" i2="X" l="FRE"><s0>Eau chaude</s0><s5>02</s5></fC03><fC03 i1="02" i2="X" l="ENG"><s0>Hot water</s0><s5>02</s5></fC03><fC03 i1="02" i2="X" l="SPA"><s0>Agua caliente</s0><s5>02</s5></fC03><fC03 i1="03" i2="X" l="FRE"><s0>Procédé extraction</s0><s5>03</s5></fC03><fC03 i1="03" i2="X" l="ENG"><s0>Extraction process</s0><s5>03</s5></fC03><fC03 i1="03" i2="X" l="SPA"><s0>Procedimiento extracción</s0><s5>03</s5></fC03><fC03 i1="04" i2="2" l="FRE"><s0>Spectrométrie visible</s0><s5>04</s5></fC03><fC03 i1="04" i2="2" l="ENG"><s0>visible spectroscopy</s0><s5>04</s5></fC03><fC03 i1="05" i2="2" l="FRE"><s0>Spectrométrie IR proche</s0><s5>05</s5></fC03><fC03 i1="05" i2="2" l="ENG"><s0>near infrared spectroscopy</s0><s5>05</s5></fC03><fC03 i1="06" i2="X" l="FRE"><s0>Spectrométrie réflexion</s0><s5>06</s5></fC03><fC03 i1="06" i2="X" l="ENG"><s0>Reflection spectrometry</s0><s5>06</s5></fC03><fC03 i1="06" i2="X" l="SPA"><s0>Espectrometría reflexión</s0><s5>06</s5></fC03><fC03 i1="07" i2="X" l="FRE"><s0>Facteur réflexion</s0><s5>07</s5></fC03><fC03 i1="07" i2="X" l="ENG"><s0>Reflectance</s0><s5>07</s5></fC03><fC03 i1="07" i2="X" l="SPA"><s0>Coeficiente reflexión</s0><s5>07</s5></fC03><fC03 i1="08" i2="2" l="FRE"><s0>Rayonnement IR proche</s0><s5>08</s5></fC03><fC03 i1="08" i2="2" l="ENG"><s0>near infrared radiation</s0><s5>08</s5></fC03><fC03 i1="09" i2="X" l="FRE"><s0>Régression PLS</s0><s5>09</s5></fC03><fC03 i1="09" i2="X" l="ENG"><s0>PLS regression</s0><s5>09</s5></fC03><fC03 i1="09" i2="X" l="SPA"><s0>Regresión PLS</s0><s5>09</s5></fC03><fC03 i1="10" i2="2" l="FRE"><s0>Carbone organique</s0><s5>15</s5></fC03><fC03 i1="10" i2="2" l="ENG"><s0>organic carbon</s0><s5>15</s5></fC03><fC03 i1="10" i2="2" l="SPA"><s0>Carbono orgánico</s0><s5>15</s5></fC03><fC03 i1="11" i2="2" l="FRE"><s0>Sol</s0><s2>NT</s2><s5>24</s5></fC03><fC03 i1="11" i2="2" l="ENG"><s0>soils</s0><s2>NT</s2><s5>24</s5></fC03><fC03 i1="11" i2="2" l="SPA"><s0>Suelo</s0><s2>NT</s2><s5>24</s5></fC03><fC03 i1="12" i2="2" l="FRE"><s0>Technique</s0><s5>28</s5></fC03><fC03 i1="12" i2="2" l="ENG"><s0>techniques</s0><s5>28</s5></fC03><fC03 i1="13" i2="X" l="FRE"><s0>Science terre</s0><s5>29</s5></fC03><fC03 i1="13" i2="X" l="ENG"><s0>Earth science</s0><s5>29</s5></fC03><fC03 i1="13" i2="X" l="SPA"><s0>Ciencia tierra</s0><s5>29</s5></fC03><fC03 i1="14" i2="X" l="FRE"><s0>Science du sol</s0><s5>30</s5></fC03><fC03 i1="14" i2="X" l="ENG"><s0>Soil science</s0><s5>30</s5></fC03><fC03 i1="14" i2="X" l="SPA"><s0>Ciencia del suelo</s0><s5>30</s5></fC03><fC03 i1="15" i2="2" l="FRE"><s0>Sélection forme</s0><s4>CD</s4><s5>96</s5></fC03><fC03 i1="15" i2="2" l="ENG"><s0>Feature selection</s0><s4>CD</s4><s5>96</s5></fC03><fC03 i1="15" i2="2" l="SPA"><s0>Selección de formas</s0><s4>CD</s4><s5>96</s5></fC03><fC07 i1="01" i2="2" l="FRE"><s0>Méthode optique</s0><s5>33</s5></fC07><fC07 i1="01" i2="2" l="ENG"><s0>optical methods</s0><s5>33</s5></fC07><fC07 i1="02" i2="2" l="FRE"><s0>Propriété optique</s0><s5>34</s5></fC07><fC07 i1="02" i2="2" l="ENG"><s0>optical properties</s0><s5>34</s5></fC07><fC07 i1="02" i2="2" l="SPA"><s0>Propiedad óptica</s0><s5>34</s5></fC07><fC07 i1="03" i2="2" l="FRE"><s0>Onde électromagnétique</s0><s5>35</s5></fC07><fC07 i1="03" i2="2" l="ENG"><s0>electromagnetic waves</s0><s5>35</s5></fC07><fC07 i1="03" i2="2" l="SPA"><s0>Onda electromagnética</s0><s5>35</s5></fC07><fC07 i1="04" i2="2" l="FRE"><s0>Régression statistique</s0><s5>36</s5></fC07><fC07 i1="04" i2="2" l="ENG"><s0>regression analysis</s0><s5>36</s5></fC07><fC07 i1="04" i2="2" l="SPA"><s0>Regresión estadística</s0><s5>36</s5></fC07><fC07 i1="05" i2="2" l="FRE"><s0>Méthode physique</s0><s5>37</s5></fC07><fC07 i1="05" i2="2" l="ENG"><s0>physical methods</s0><s5>37</s5></fC07><fC07 i1="06" i2="2" l="FRE"><s0>Propriété physique</s0><s5>38</s5></fC07><fC07 i1="06" i2="2" l="ENG"><s0>physical properties</s0><s5>38</s5></fC07><fC07 i1="06" i2="2" l="SPA"><s0>Propiedad física</s0><s5>38</s5></fC07><fC07 i1="07" i2="2" l="FRE"><s0>Matière organique</s0><s5>50</s5></fC07><fC07 i1="07" i2="2" l="ENG"><s0>organic materials</s0><s5>50</s5></fC07><fC07 i1="07" i2="2" l="SPA"><s0>Materia orgánica</s0><s5>50</s5></fC07><fC07 i1="08" i2="2" l="FRE"><s0>Analyse statistique</s0><s5>64</s5></fC07><fC07 i1="08" i2="2" l="ENG"><s0>statistical analysis</s0><s5>64</s5></fC07><fN21><s1>227</s1></fN21><fN44 i1="01"><s1>OTO</s1></fN44><fN82><s1>OTO</s1></fN82></pA></standard><server><NO>PASCAL 11-0335983 INIST</NO><ET>Determination of total soil organic C and hot water-extractable C from VIS-NIR soil reflectance with partial least squares regression and spectral feature selection techniques</ET><AU>VOHLAND (M.); EMMERLING (C.)</AU><AF>Faculty of Geography and Geosciences, Remote Sensing Department, University of Trier/54286 Trier/Allemagne (1 aut.); Faculty of Geography and Geosciences, Soil Science Department, University of Trier/54286 Trier/Allemagne (2 aut.)</AF><DT>Publication en série; Niveau analytique</DT><SO>European journal of soil science; ISSN 1351-0754; Royaume-Uni; Da. 2011; Vol. 62; No. 4; Pp. 598-606; Bibl. 3/4 p.</SO><LA>Anglais</LA><EA>The calibration of soil organic C (SOC) and hot water-extractable C (HWE-C) from visible and near-infrared soil reflectance spectra is hindered by the complex spectral interaction of soil chromophores that usually varies from one soil or soil type to another. The exploitation of spectral variables from spectroradiometer data is further affected by multicollinearity and noise. In this study, a set of soil samples (Fluvisols, Podzols, Cambisols and Chernozems; n = 48) representing a wide range of properties was analysed. Spectral readings with a fibre-optics visible to near-infrared instrument were used to estimate SOC and HWE-C contents by partial least squares regression (PLS). In addition to full-spectrum PLS, spectral feature selection techniques were applied with PLS (uninformative variable elimination, UVE-PLS, and a genetic algorithm, GA-PLS). On the basis of normalized spectra (mean centring + vector normalization), the order of prediction accuracy was GA-PLS » UVE-PLS > PLS for SOC; for HWE-C, it was GA-PLS > UVE-PLS, PLS. With GA-PLS, acceptable cross-validated (cv) prediction accuracies were obtained for the complete dataset (SOC, R<sup>2</sup><sub>cv </sub>= 0.83, RPD<sub>cv</sub> = 2.42; HWE-C<sub>cv</sub>, R<sup>2</sup><sub>cv</sub> = 0.78, RPD<sub>cv</sub> = 2.13). Splitting the soil data into two groups with different basic properties (Podzols compared with Fluvisols/Cambisols; n = 21 and n = 23, respectively) improved SOC predictions with GA-PLS distinctly (Podzols, R<sup>2</sup><sub>cv</sub> = 0.89, RPD<sub>cv</sub> = 3.14; Fluvisols/Cambisols, R<sup>2</sup><sub>cv</sub> = 0.92, RPD<sub>cv</sub> = 3.64). This demonstrates the importance of using stratified models for successful quantitative approaches after an initial rough screening. GA selection frequencies suggest that the spectral region over 1900 nm, and in particular the hydroxyl band at 2200 nm are of great importance for the spectral prediction of both SOC and HWE-C.</EA><CC>001E01P03; 002A32B03B3; 226C03</CC><FD>Détermination; Eau chaude; Procédé extraction; Spectrométrie visible; Spectrométrie IR proche; Spectrométrie réflexion; Facteur réflexion; Rayonnement IR proche; Régression PLS; Carbone organique; Sol; Technique; Science terre; Science du sol; Sélection forme</FD><FG>Méthode optique; Propriété optique; Onde électromagnétique; Régression statistique; Méthode physique; Propriété physique; Matière organique; Analyse statistique</FG><ED>Determination; Hot water; Extraction process; visible spectroscopy; near infrared spectroscopy; Reflection spectrometry; Reflectance; near infrared radiation; PLS regression; organic carbon; soils; techniques; Earth science; Soil science; Feature selection</ED><EG>optical methods; optical properties; electromagnetic waves; regression analysis; physical methods; physical properties; organic materials; statistical analysis</EG><SD>Determinación; Agua caliente; Procedimiento extracción; Espectrometría reflexión; Coeficiente reflexión; Regresión PLS; Carbono orgánico; Suelo; Ciencia tierra; Ciencia del suelo; Selección de formas</SD><LO>INIST-2402.354000509403950110</LO><ID>11-0335983</ID></server></inist></record>Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Wicri/Rhénanie/explor/UnivTrevesV1/Data/PascalFrancis/Corpus
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 000404 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/PascalFrancis/Corpus/biblio.hfd -nk 000404 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien
|wiki= Wicri/Rhénanie
|area= UnivTrevesV1
|flux= PascalFrancis
|étape= Corpus
|type= RBID
|clé= Pascal:11-0335983
|texte= Determination of total soil organic C and hot water-extractable C from VIS-NIR soil reflectance with partial least squares regression and spectral feature selection techniques
}}
| This area was generated with Dilib version V0.6.31. |  |