Comparison of multivariate methods for inferential modeling of soil carbon using visible/near-infrared spectra

Vasques, Gustavo M.; Grunwald, Sabine; Sickman, J.O.

doi:10.1016/j.geoderma.2008.04.007

Cited by 337 publications

(248 citation statements)

References 39 publications

(67 reference statements)

Supporting

Mentioning

217

Contrasting

Unclassified

Order By: Relevance

“…The PLSR models were derived using the 'pls' package in R. The maximum number of latent variables was set at 20 and the type of model used was the classical orthogonal scores algorithm [28]. Ten-fold cross-validation was used to determine the optimal number of latent variables to be included as predictors in the models based on the RMSE of calibration.…”

Section: Discussionmentioning

confidence: 99%

Prediction of Soil Physical and Chemical Properties by Visible and Near-Infrared Diffuse Reflectance Spectroscopy in the Central Amazon

et al. 2017

View full text Add to dashboard Cite

Visible and near-infrared diffuse reflectance spectroscopy (VIS-NIR) has shown levels of accuracy comparable to conventional laboratory methods for estimating soil properties. Soil chemical and physical properties have been predicted by reflectance spectroscopy successfully on subtropical and temperate soils, whereas soils from tropical agro-forest regions have received less attention, especially those from tropical rainforests. A spectral characterization provides a proficient pathway for soil characterization. The first step in this process is to develop a comprehensive VIS-NIR soil library of multiple key soil properties to be used in future soil surveys. This paper presents the first VIS-NIR soil library for a remote region in the Central Amazon. We evaluated the performance of VIS-NIR for the prediction of soil properties in the Central Amazon, Brazil. Soil properties measured and predicted were: pH, Ca, Mg, Al, H, H+Al, P, organic C (SOC), sum of bases, cation exchange capacity (CEC), percentage of base saturation (V), Al saturation (m), clay, sand, silt, silt/clay (S/C), and degree of flocculation. Soil samples were scanned in the laboratory in the VIS-NIR range (350-2500 nm), and forty-one pre-processing methods were tested to improve predictions. Clay content was predicted with the highest accuracy, followed by SOC. Sand, S/C, H, Al, H+Al, CEC, m and V predictions were reasonably good. The other soil properties were poorly predicted. Among the soil properties predicted well, SOC is one of the critical soil indicators in the global carbon cycle. Besides the soil property of interest, the landscape position, soil order and depth influenced in the model performance. For silt content, pH and S/C, the model performed better in well-drained soils, whereas for SOC best predictions were obtained in poorly drained soils. The association of VIS-NIR spectral data to landforms, vegetation classes, and soil types demonstrate potential for soil characterization.

show abstract

Section: Discussionmentioning

confidence: 99%

Prediction of Soil Physical and Chemical Properties by Visible and Near-Infrared Diffuse Reflectance Spectroscopy in the Central Amazon

et al. 2017

View full text Add to dashboard Cite

show abstract

“…Esses grupos de dados diferenciam-se quanto ao número de amostras e de covariáveis envolvidas. Os três primeiros grupos contam com 73 amostras e 32 covariáveis (todas as possíveis), ou 12 covariáveis selecionadas pela regressão linear stepwise backward, conforme adotado por Vasques et al (2008), Poggio et al (2013) e Samuel-Rosa et al (2015, entre outros, ou 13 covariáveis definidas pelo limiar menor ou igual a 0,05 para o valor p da correlação. Estes grupos foram validados pela validação cruzada "leave one out".…”

Section: Methodsunclassified

Regressão linear múltipla e modelo Random Forest para estimar a densidade do solo em áreas montanhosas

Carvalho

Filho

Chagas

et al. 2016

Pesq. agropec. bras.

View full text Add to dashboard Cite

Resumo -O objetivo deste trabalho foi o desenvolvimento de modelos com diferentes conjuntos de dados, para estimar a densidade de solos de regiões tropicais montanhosas, a partir de atributos de solos comumente encontrados nas análises de perfis de solos descritos nos levantamentos regionais. O conjunto total de dados compõe-se de 163 amostras e foi dividido em seis grupamentos, dos quais três com 73 amostras, com o máximo de 32 covariáveis, e três com 163 amostras, com o máximo de 18 covariáveis. Testaram-se modelos de regressão linear múltipla (RLM) e randomForest (RF). A menor incerteza entre os modelos foi alcançada pelo RLM2, com R 2 de 0,56, 13 covariáveis e 73 amostras. Nos grupamentos com 163 amostras, os melhores modelos foram os RF, com R 2 médio de 0,48. A raiz quadrada da média do erro ao quadrado variou entre 0,09 e 0,14. As covariáveis mais importantes no modelo RF foram: carbono orgânico, hidrogênio, areia fina e grossa, saturação por bases e capacidade de troca catiônica. Pelo método "stepwise regression", as variáveis mais importantes foram: a relação silte/argila; areia grossa e fina; carbono orgânico; saturação por bases; e potássio.Termos para indexação: estoque de carbono, funções de pedotransferência, modelos dirigidos pelos dados, stepwise. Multiple linear regression and Random Forest model to estimate soil bulk density in mountainous regionsAbstract -The objective of this work was the development of models with different sets of data for estimating soil bulk density in tropical mountainous regions, from soil attributes commonly found in the analyses of soil profiles described in regional surveys. The complete dataset is composed of 163 samples and it was divided into six groups, of which three groups have 73 samples and the maximum of 32 covariables, and three have 163 samples and the maximum of 18 covariables. The linear regression (RLM) and randomForest (RF) models were tested. The lowest uncertainty between the models was achieved by RLM2, with R 2 of 0.56, 13 covariables, and 73 samples. Considering the groups with 163 samples, the best models were the RFs with mean R 2 of 0.48. The root mean squared error ranged between 0.09 and 0.14. The most important covariables in the RF model were: organic carbon, hydrogen, fine and coarse sand, base saturation, and cation exchange capacity. By the stepwise backward regression, the main covariables were: silt and clay relation; fine and coarse sand; organic carbon; base saturation; and potassium.

show abstract

“…The PLSR, which has turned into a popular algorithm in chemometrics, decreases the data, noise and calculation time with minor loss of the information contained in the original variables [37] and its arithmetic can be found in Wold et al [38]. It is strongly related to principal component regression (PCR), in that both methods use statistical rotations to defeat the problem of high dimensionality and multicollinearity [39,40].…”

Section: Plsr Modellingmentioning

confidence: 99%

Agricultural Soil Spectral Response and Properties Assessment: Effects of Measurement Protocol and Data Mining Technique

et al. 2017

View full text Add to dashboard Cite

Abstract:Soil spectroscopy has shown to be a fast, cost-effective, environmentally friendly, non-destructive, reproducible and repeatable analytical technique. Soil components, as well as types of instruments, protocols, sampling methods, sample preparation, spectral acquisition techniques and analytical algorithms have a combined influence on the final performance. Therefore, it is important to characterize these differences and to introduce an effective approach in order to minimize the technical factors that alter reflectance spectra and consequent prediction. To quantify this alteration, a joint project between Czech University of Life Sciences Prague (CULS) and Tel-Aviv University (TAU) was conducted to estimate Cox, pH-H 2 O, pH-KCl and selected forms of Fe and Mn. Two different soil spectral measurement protocols and two data mining techniques were used to examine seventy-eight soil samples from five agricultural areas in different parts of the Czech Republic. Spectral measurements at both laboratories were made using different ASD spectroradiometers. The CULS protocol was based on employing a contact probe (CP) spectral measurement scheme, while the TAU protocol was carried out using a CP measurement method, accompanied with the internal soil standard (ISS) procedure. Two spectral datasets, acquired from different protocols, were both analyzed using partial least square regression (PLSR) technique as well as the PARACUDA II ® , a new data mining engine for optimizing PLSR models. The results showed that spectra based on the CULS setup (non-ISS) demonstrated significantly higher albedo intensity and reflectance values relative to the TAU setup with ISS. However, the majority of statistics using the TAU protocol was not noticeably better than the CULS spectra. The paper also highlighted that under both measurement protocols, the PARACUDA II ® engine proved to be a powerful tool for providing better results than PLSR. Such initiative is not only a way to unlock current limitations of soil spectroscopy, but also offers considerable efficiency and cost-and time-saving possibilities, which lead to further improvements in prediction performance of spectral models.

show abstract

Comparison of multivariate methods for inferential modeling of soil carbon using visible/near-infrared spectra

Cited by 337 publications

References 39 publications

Prediction of Soil Physical and Chemical Properties by Visible and Near-Infrared Diffuse Reflectance Spectroscopy in the Central Amazon

Prediction of Soil Physical and Chemical Properties by Visible and Near-Infrared Diffuse Reflectance Spectroscopy in the Central Amazon

Regressão linear múltipla e modelo Random Forest para estimar a densidade do solo em áreas montanhosas

Agricultural Soil Spectral Response and Properties Assessment: Effects of Measurement Protocol and Data Mining Technique

Contact Info

Product

Resources

About