Xiao Wang, Jianli Ding, Lijing Han, Jiao Tan, Xiangyu Ge
{"title":"Enhancing soil particle content prediction accuracy: advanced hyperspectral analysis and machine learning models","authors":"Xiao Wang, Jianli Ding, Lijing Han, Jiao Tan, Xiangyu Ge","doi":"10.1007/s11368-024-03886-8","DOIUrl":null,"url":null,"abstract":"<h3 data-test=\"abstract-sub-heading\">Purpose</h3><p>Prediction of soil particle content is essential for soil texture classification, soil management and agricultural production. This study aimed to achieve high-accuracy predictions of soil particle content in the Ogan-Kucha River Oasis using hyperspectral data and environmental variables.</p><h3 data-test=\"abstract-sub-heading\">Materials and methods</h3><p>We collected 62 representative surface soil samples (depth: 0–10 cm), and conducting indoor soil particle content and spectral measurements. The relationship between environmental variables and soil particle content was analyzed using the Boruta algorithm, and seven three-band spectral indices (TBIs) were constructed using an optimal band algorithm. By integrating environmental covariates and TBIs, soil particle inversion models were developed using the extreme learning machine (ELM), backpropagation neural networks (BP), neural networks optimized with the sparrow search algorithm (SSA-BP), and neural networks optimized with the sparrow search algorithm enhanced by Sine chaos mapping (Sine-SSA-BP).</p><h3 data-test=\"abstract-sub-heading\">Results and discussion</h3><p>The results demonstrated that (1) the Boruta algorithm identified key environmental covariates that affect specific soil particle components; (2) there was significant variation in the correlation between different TBIs and soil particle content, with absolute correlation coefficients ranging from 0.225 to 0.852; (3) the estimation models established by the four machine learning algorithms performed well in predicting soil particle content, particularly for silt (<i>R</i><sup><i>2</i></sup>: 0.664–0.858, RMSE: 11.107–17.128) and clay (<i>R</i><sup><i>2</i></sup>: 0.444–0.857, RMSE: 0.550–1.405), for which higher accuracy was achieved; and (4) compared with the traditional ELM (<i>R</i><sup><i>2</i></sup>: 0.422–0.664), BP (<i>R</i><sup><i>2</i></sup>: 0.487–0.673) and SSA-BP models (<i>R</i><sup><i>2</i></sup>: 0.625–0.777), the Sine-SSA-BP model showed a significant improvement in prediction accuracy, with the highest <i>R</i><sup><i>2</i></sup> reaching 0.858.</p><h3 data-test=\"abstract-sub-heading\">Conclusion</h3><p>Compared to the traditional ELM, BP and SSA-BP models, the Sine-SSA-BP model significantly excelled in predicting soil particle content, offering innovative insights and robust support for soil texture classification and management.</p>","PeriodicalId":17139,"journal":{"name":"Journal of Soils and Sediments","volume":null,"pages":null},"PeriodicalIF":2.8000,"publicationDate":"2024-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Soils and Sediments","FirstCategoryId":"97","ListUrlMain":"https://doi.org/10.1007/s11368-024-03886-8","RegionNum":3,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"ENVIRONMENTAL SCIENCES","Score":null,"Total":0}
引用次数: 0
Abstract
Purpose
Prediction of soil particle content is essential for soil texture classification, soil management and agricultural production. This study aimed to achieve high-accuracy predictions of soil particle content in the Ogan-Kucha River Oasis using hyperspectral data and environmental variables.
Materials and methods
We collected 62 representative surface soil samples (depth: 0–10 cm), and conducting indoor soil particle content and spectral measurements. The relationship between environmental variables and soil particle content was analyzed using the Boruta algorithm, and seven three-band spectral indices (TBIs) were constructed using an optimal band algorithm. By integrating environmental covariates and TBIs, soil particle inversion models were developed using the extreme learning machine (ELM), backpropagation neural networks (BP), neural networks optimized with the sparrow search algorithm (SSA-BP), and neural networks optimized with the sparrow search algorithm enhanced by Sine chaos mapping (Sine-SSA-BP).
Results and discussion
The results demonstrated that (1) the Boruta algorithm identified key environmental covariates that affect specific soil particle components; (2) there was significant variation in the correlation between different TBIs and soil particle content, with absolute correlation coefficients ranging from 0.225 to 0.852; (3) the estimation models established by the four machine learning algorithms performed well in predicting soil particle content, particularly for silt (R2: 0.664–0.858, RMSE: 11.107–17.128) and clay (R2: 0.444–0.857, RMSE: 0.550–1.405), for which higher accuracy was achieved; and (4) compared with the traditional ELM (R2: 0.422–0.664), BP (R2: 0.487–0.673) and SSA-BP models (R2: 0.625–0.777), the Sine-SSA-BP model showed a significant improvement in prediction accuracy, with the highest R2 reaching 0.858.
Conclusion
Compared to the traditional ELM, BP and SSA-BP models, the Sine-SSA-BP model significantly excelled in predicting soil particle content, offering innovative insights and robust support for soil texture classification and management.
期刊介绍:
The Journal of Soils and Sediments (JSS) is devoted to soils and sediments; it deals with contaminated, intact and disturbed soils and sediments. JSS explores both the common aspects and the differences between these two environmental compartments. Inter-linkages at the catchment scale and with the Earth’s system (inter-compartment) are an important topic in JSS. The range of research coverage includes the effects of disturbances and contamination; research, strategies and technologies for prediction, prevention, and protection; identification and characterization; treatment, remediation and reuse; risk assessment and management; creation and implementation of quality standards; international regulation and legislation.