Estimation of soil lead content based on GF-5 hyperspectral images, considering the influence of soil environmental factors

Ding, Shuai; Zhang, Xia; Sun, Wendong; Shang, Ke; Wang, Yibo

doi:10.1007/s11368-022-03169-0

Cited by 4 publications

(4 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Research has shown that iron oxides have an impact on the spectral quantitative retrieval of soil heavy metals [39]. With an increase in iron oxide content or a higher ratio of iron oxide content to organic matter content, the stability and predictive ability of the retrieval model weaken, leading to the formation of absorption bands around 500 nm and 950 nm due to iron oxides.…”

Section: Mechanism Analysis Of Soil Arsenic Concentration In Pingtan ...mentioning

confidence: 99%

“…The transition from ground-based hyperspectral retrieval to satellite-based retrieval poses different challenges and issues, such as the robustness of characteristic bands and the universal availability of optimal models. Therefore, it is crucial to consider the comprehensive use of both ground-based and satellite-based characteristic bands to address these challenges effectively [39].…”

mentioning

confidence: 99%

See 1 more Smart Citation

Ground-Based Hyperspectral Retrieval of Soil Arsenic Concentration in Pingtan Island, China

Zheng,

Luan,

Liu

et al. 2023

Remote Sensing

View full text Add to dashboard Cite

The optimal selection of characteristic bands and retrieval models for the hyperspectral retrieval of soil heavy metal concentrations poses a significant challenge. Additionally, satellite-based hyperspectral retrieval encounters several issues, including atmospheric effects, limitations in temporal and radiometric resolution, and data acquisition, among others. Given this, the retrieval performance of the soil arsenic (As) concentration in Pingtan Island, the largest island in Fujian Province and the fifth largest in China, is currently unclear. This study aimed to elucidate this issue by identifying optimal characteristic bands from the full spectrum from both statistical and physical perspectives. We tested three linear models, namely Multiple Linear Regression (MLR), Partial Least Squares Regression (PLSR) and Geographically Weighted Regression (GWR), as well as three nonlinear machine learning models, including Back Propagation Neural Network (BP), Support Vector Machine Regression (SVR) and Random Forest Regression (RFR). We then retrieved soil arsenic content using ground-based soil full spectrum data on Pingtan Island. Our results indicate that the RFR model consistently outperformed all others when using both original and optimal characteristic bands. This superior performance suggests a complex, nonlinear relationship between soil arsenic concentration and spectral variables, influenced by diverse landscape factors. The GWR model, which considers spatial non-stationarity and heterogeneity, outperformed traditional models such as BP and SVR. This finding underscores the potential of incorporating spatial characteristics to enhance traditional machine learning models in geospatial studies. When evaluating retrieval model accuracy based on optimal characteristic bands, the RFR model maintained its top performance, and linear models (MLR, PLSR and GWR) showed notable improvement. Specifically, the GWR model achieved the highest r value for the validation data, indicating that selecting optimal characteristic bands based on high Pearson’s correlation coefficients (e.g., abs(Pearson’s correlation coefficient) ≥0.45) and high sensitivity to soil active materials successfully mitigates uncertainties linked to characteristic band selection solely based on Pearson’s correlation coefficients. Consequently, two effective retrieval models were generated: the best-performing RFR model and the improved GWR model. Our study on Pingtan Island provides theoretical and technical support for monitoring and evaluating soil arsenic concentrations using satellite-based spectroscopy in densely populated, relatively independent island towns in China and worldwide.

show abstract

Section: Mechanism Analysis Of Soil Arsenic Concentration In Pingtan ...mentioning

confidence: 99%

mentioning

confidence: 99%

Ground-Based Hyperspectral Retrieval of Soil Arsenic Concentration in Pingtan Island, China

Zheng,

Luan,

Liu

et al. 2023

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…The selected Orbiter satellite OHS sensor is used for data acquisition (Steinberg et al, 2016;Ding et al, 2022). Push scan imaging mode is adopted, with single imaging of 150 km × 400 km, with a spatial resolution of 10 m, and a spectral resolution of 2.5 nm.…”

Section: Hyperspectral Data Acquisition and Processingmentioning

confidence: 99%

“…Remote sensing technology has technical advantages in the identification, inversion, supervision, and risk assessment of soil composition (Kusuma et al, 2012;Shi X. et al, 2014;Rathod et al, 2016). Accurate extraction of iron oxide (Liu and Sun, 2019), water content (Cudahy et al, 2010), and heavy metal (Ding et al, 2022) has been achieved through the process of characteristic bands selection (Pessoa et al, 2016), building remote sensing reflectance index (Shi X.-Z. et al, 2014), establishing calculation model (Mendes et al, 2022), conducting band combination (Sellitto et al, 2009), content prediction (Li, 2020), etc.…”

Section: Introductionmentioning

confidence: 99%

Response of soil oxides in complex terrain region to environment revealed by hyperspectral observations

Zhang

Tang

et al. 2023

Front. Environ. Sci.

View full text Add to dashboard Cite

Soil oxides are important diagnostic indicators for soil quality assessment, and their content has a positive effect on adsorption of heavy metals, remediation of pollution, and enhancement of organic matter enrichment. Moreover, soil oxides are affected by multiple environmental factors such as soil pH, geomorphological, and elevation, and their distribution is highly regional. In this study, 421 ground soil samples and 2 hyperspectral satellite data were collected. The content distribution of 8 main oxides in soil, including SiO2, Al2O3, Fe2O3, MgO, CaO, Na2O, K2O, and Corg. were obtained according to the process of data preprocessing, spectral transformation, feature extraction, band combination, and model establishment. The results indicated that the homogenization spectral transformation method is the most suitable; The calculation accuracy of complex spectral characteristic parameters is inferior to that of characteristic bands; The difference model is more suitable for the extraction of soil oxide content in this study area; The average content of SiO2 and CaO shows a decreasing trend with the elevation increasing; The average content of SiO2 and Fe2O3 shows an upward trend with the increase of slope; East is the main geomorphic direction of soil oxide enrichment. SiO2, Al2O3, and MgO are the three soil oxides with the highest correlation. The response relationship between soil oxides, topography and pH is discussed, and the distribution rule of soil oxides and environmental determinants are obtained, which provides a set of technical schemes for digital soil research. This study can extract the content of soil components in a large area, analyze the causes, and provide a new technical scheme for soil investigation and treatment.

show abstract

Estimating the heavy metal contents in farmland soil from hyperspectral images based on Stacked AdaBoost ensemble learning

Lin

Jiang

et al. 2022

Ecological Indicators

View full text Add to dashboard Cite

Estimation of soil lead content based on GF-5 hyperspectral images, considering the influence of soil environmental factors

Cited by 4 publications

References 33 publications

Ground-Based Hyperspectral Retrieval of Soil Arsenic Concentration in Pingtan Island, China

Ground-Based Hyperspectral Retrieval of Soil Arsenic Concentration in Pingtan Island, China

Response of soil oxides in complex terrain region to environment revealed by hyperspectral observations

Estimating the heavy metal contents in farmland soil from hyperspectral images based on Stacked AdaBoost ensemble learning

Contact Info

Product

Resources

About