Vis-NIR Spectroscopy and PLS Regression with Waveband Selection for Estimating the Total C and N of Paddy Soils in Madagascar

Kawamura, Kensuke; Tsujimoto, Yasuhiro; Rabenarivo, Michel; Asai, Hidetoshi; Andriamananjara, Andry; Rakotoson, Tovohery

doi:10.3390/rs9101081

Cited by 83 publications

(56 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These results also suggested that over 92% of the waveband information from the soil reflectance spectrum was redundant and did not contribute to or disturb the prediction. These findings support previous findings that the performance of PLS models can be improved through waveband selection, and the most useful information in the Vis-NIR region (400-2400 nm) predicted less than 20% of the forage [30,70], water [71] and soil parameters [18]. Moreover, the spectral data efficiency is also expected to improve by the optimization of the waveband subset using the GA-PLS model [30].…”

Section: Waveband Selection With Cross-validated Calibration Resultssupporting

confidence: 88%

“…Although partial least squares (PLS) regression is the most commonly used approach for soil spectral analyses, waveband selection can refine the performance of a PLS analysis [16][17][18]. The PLS regression method combines the most useful information from hundreds of wavebands into the first several PLS factors (or latent variables), whereas the less important factors might include background effects [19,20].…”

Section: Introductionmentioning

confidence: 99%

“…Thus, many techniques for selecting wavebands or wavelength regions have been developed, such as iterative stepwise elimination-PLS (ISE-PLS) regression [21], uninformative variable elimination-PLS (UVE-PLS) regression [22], competitive adaptive reweighted sampling (CARS) regression [23], interval PLS (iPLS) regression [24], moving window-PLS (MW-PLS) regression [25], and genetic algorithm-PLS (GA-PLS) regression [26]. In our previous study [18], removal of the redundant wavebands by ISE-PLS regression greatly improved the estimation of total carbon (TC) and total nitrogen (TN) in paddy soils. Among the waveband selection methods, GA-PLS regression has been used as a suitable method in chemometrics [27].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Laboratory Visible and Near-Infrared Spectroscopy with Genetic Algorithm-Based Partial Least Squares Regression for Assessing the Soil Phosphorus Content of Upland and Lowland Rice Fields in Madagascar

et al. 2019

Self Cite

View full text Add to dashboard Cite

As a laboratory proximal sensing technique, the capability of visible and near-infrared (Vis-NIR) diffused reflectance spectroscopy with partial least squares (PLS) regression to determine soil properties has previously been demonstrated. However, the evaluation of the soil phosphorus (P) content—a major nutrient constraint for crop production in the tropics—is still a challenging task. PLS regression with waveband selection can improve the predictive ability of a calibration model, and a genetic algorithm (GA) has been widely applied as a suitable method for selecting wavebands in laboratory calibrations. To develop a laboratory-based proximal sensing method, this study investigated the potential to use GA-PLS regression analyses to estimate oxalate-extractable P in upland and lowland soils from laboratory Vis-NIR reflectance data. In terms of predictive ability, GA-PLS regression was compared with iterative stepwise elimination PLS (ISE-PLS) regression and standard full-spectrum PLS (FS-PLS) regression using soil samples collected in 2015 and 2016 from the surface of upland and lowland rice fields in Madagascar (n = 103). Overall, the GA-PLS model using first derivative reflectance (FDR) had the best predictive accuracy (R2 = 0.796) with a good prediction ability (residual predictive deviation (RPD) = 2.211). Selected wavebands in the GA-PLS model did not perfectly match wavelengths of previously known absorption features of soil nutrients, but in most cases, the selected wavebands were within 20 nm of previously known wavelength regions. Bootstrap procedures (N = 10,000 times) using selected wavebands also confirmed the improvements in accuracy and robustness of the GA-PLS model compared to those of the ISE-PLS and FS-PLS models. These results suggest that soil oxalate-extractable P can be predicted from Vis-NIR spectroscopy and that GA-PLS regression has the advantage of tuning optimum bands for PLS regression, contributing to a better predictive ability.

show abstract

Section: Waveband Selection With Cross-validated Calibration Resultssupporting

confidence: 88%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Laboratory Visible and Near-Infrared Spectroscopy with Genetic Algorithm-Based Partial Least Squares Regression for Assessing the Soil Phosphorus Content of Upland and Lowland Rice Fields in Madagascar

et al. 2019

Self Cite

View full text Add to dashboard Cite

show abstract

“…The RPD is defined as the ratio of the standard deviation of the measured TPH concentration to the RMSE computed between the predicted and measured concentrations [27]. The interpretation of RPD values strongly depends on the context [27,64]. RPD values above 2 and 3 are often considered as good and excellent predictions of TPH, respectively, and values below 2 as poor ones.…”

Section: Variability Of Leaf Pigment Contents and Tph Estimationmentioning

confidence: 99%

Application of PROSPECT for estimating total petroleum hydrocarbons in contaminated soils from leaf optical properties

Lassalle

Fabre

Crédoz

et al. 2019

Journal of Hazardous Materials

View full text Add to dashboard Cite

Recent advances in hyperspectral spectroscopy suggest making use of leaf optical properties for monitoring soil contamination in oil production regions by detecting pigment alterations induced by Total Petroleum Hydrocarbons (TPH). However, this provides no quantitative information about the level of contamination. To achieve this, we propose an approach based on the inversion of the PROSPECT model. 1620 leaves from five species were collected on a site contaminated by 16 to 77 g.kg-1 of TPH over a 14-month period. Their spectral signature was measured and used in PROSPECT model inversions to retrieve leaf biochemistry. The model performed well for simulating the spectral signatures (RMSE < 2%) and for estimating leaf pigment contents (RMSE ≤ 2.95 µg.cm-2 for chlorophylls). Four out of the five species exhibited alterations in pigment contents when exposed to TPH. A strong correlation was established between leaf chlorophyll content and soil TPH concentrations (R 2 ≥ 0.74) for three of them, allowing accurate predictions of TPH (RMSE = 3.20 g.kg-1 and RPD = 5.17). The accuracy of predictions varied by season and improved after the growing period. This study demonstrates the capacity of PROSPECT to estimate oil contamination and opens up promising perspectives for larger-scale applications.

show abstract

“…In this paper, R 2 c and R 2 p represent the coefficient of determination of the calibration set and the prediction set, respectively, and RMSE c and RMSE p represent the root mean square error of the calibration set and prediction set, respectively. Besides this, the RPD has been suggested to be at least 3 for agriculture applications; where 2 < RPD < 3 indicates a model with a good prediction ability; 1.4 < RPD < 2 is an intermediate model needing some improvement; and RPD <1.4 indicates that the model has a poor prediction ability [36].…”

Section: Model Evaluation Indexmentioning

confidence: 99%

The Effects of Drying Temperature on Nitrogen Concentration Detection in Calcium Soil Studied by NIR Spectroscopy

Nie

Dong

et al. 2018

Applied Sciences

View full text Add to dashboard Cite

Soil nitrogen is one of the crucial components for plant growth. An accurate diagnosis based on soil nitrogen information is the premise of scientific fertilization in precision agriculture. Soil nitrogen content acquisition based on near-infrared (NIR) spectroscopy shows the significant advantages of high accuracy, real-time analysis, and convenience. However, soil texture, soil moisture content, and drying temperature all affect soil nitrogen detection by NIR spectroscopy. In order to investigate the effects of drying temperature on calcium soil nitrogen detection and its characteristic bands, soil samples were detected at a 25 • C placement (ambient temperature) after 40 • C drying (medium temperature), 60 • C drying (medium-high temperature), 80 • C drying (high temperature), and 105 • C drying (extreme high temperature), respectively. Besides that, the original spectra were pretreated with five preprocessing methods, and the characteristic variables were selected by competitive adaptive reweighted squares (CARS) and backward interval partial least squares (BIPLS). The partial least squares (PLS) method was used for modeling and analysis. The predictive abilities were assessed using the coefficients of determination (R 2 ), the root mean squared error (RMSE), and the residual predictive deviation (RPD). As a result, the characteristic bands focus on 928-960 nm and 1638-1680 nm when soil was detected after 40 • C, 60 • C, and 80 • C drying. Calcium soil obtained the best prediction accuracy (R 2 p = 0.966, RMSE p = 0.128 g kg , RPD = 5.03) after 40 • C drying by the method of CARS-BIPLS-PLS. Meanwhile, the prediction model also performed well when soil was detected after 60 • C drying (R 2 p = 0.946, RMSE p = 0.172 g/kg, RPD = 4.53) and 80 • C drying (R 2 p = 0.961, RMSE p = 0.143 g kg , RPD = 4.98). However, the calcium soil obtained the worst detection result when soil was placed at 25 • C. In conclusion, a low or extremely high drying temperature had an adverse influence on the soil nitrogen detection, and the 40 • C drying temperature as well as the CARS-BIPLS-PLS method were optimal to enhance the soil nitrogen detection accuracy.

show abstract

Vis-NIR Spectroscopy and PLS Regression with Waveband Selection for Estimating the Total C and N of Paddy Soils in Madagascar

Cited by 83 publications

References 53 publications

Laboratory Visible and Near-Infrared Spectroscopy with Genetic Algorithm-Based Partial Least Squares Regression for Assessing the Soil Phosphorus Content of Upland and Lowland Rice Fields in Madagascar

Laboratory Visible and Near-Infrared Spectroscopy with Genetic Algorithm-Based Partial Least Squares Regression for Assessing the Soil Phosphorus Content of Upland and Lowland Rice Fields in Madagascar

Application of PROSPECT for estimating total petroleum hydrocarbons in contaminated soils from leaf optical properties

The Effects of Drying Temperature on Nitrogen Concentration Detection in Calcium Soil Studied by NIR Spectroscopy

Contact Info

Product

Resources

About