基于中红外与便携式X射线荧光光谱的传感器融合提升土壤属性预测精度

doi:10.1016/j.jia.2025.09.028

Journal of Integrative Agriculture ›› 2026, Vol. 25 ›› Issue (4): 1687-1699.DOI: 10.1016/j.jia.2025.09.028

基于中红外与便携式X射线荧光光谱的传感器融合提升土壤属性预测精度

收稿日期:2025-02-08 修回日期:2025-09-26 接受日期:2025-08-29 出版日期:2026-04-20 发布日期:2026-03-16

Improved selected soil properties predictions using MIR and pXRF sensor fusion

Junwei Wang^{1, 2}, Qi Zou^{1, 2}, Huimin Yuan^{1, 2#}

¹College of Resources and Environmental Sciences/National Academy of Agriculture Green Development/Key Laboratory of Plant–Soil Interactions, Ministry of Education/China Agricultural University, Beijing 100193, China
²National Observation and Research Station of Agriculture Green Development, Quzhou 057250, China

Received:2025-02-08 Revised:2025-09-26 Accepted:2025-08-29 Online:2026-04-20 Published:2026-03-16
About author:Junwei Wang, E-mail: wangjw272@cau.edu.cn; #Correspondence Huimin Yuan, E-mail: hmyuan@cau.edu.cn
Supported by:
This paper was supported by the National Key Research and Development Program of China (2023YFD1900104).

摘要/Abstract

摘要：

及时、准确地获取土壤养分信息对保障全球粮食安全和农业可持续发展至关重要。本研究利用中红外光谱（MIR）和便携式X射线荧光光谱（pXRF）两种近地传感技术，通过四种传感器融合策略：直接融合（DC）、基于稳定性竞争性自适应重加权采样（sCARS）和最小绝对收缩和选择算子（LASSO）算法的特征融合（sCARS-C、LASSO-C）、基于顺序正交偏最小二乘法（SO-PLS）算法的多模块融合和基于Granger-Ramanathan模型平均（GRA）算法的决策融合，以提升13种土壤属性的预测精度。研究结果表明：单一传感器模型（MIR或pXRF）可准确预测土壤有机质（SOM）、全氮（TN）、有效磷（AP）、钙（Ca）、铁（Fe）、锰（Mn）和pH（Rp²≥0.78, RPDp≥2.13），但对全钾（TK）、镁（Mg）、铜（Cu）、锌（Zn）、速效钾（AK）和全磷（TP）（Rp²≤0.75, RPDp≤1.99）的预测能力有限。DC模型显著提升了Mg（Rp²=0.76, RMSEp=358.76 mg kg^-1, RPDp=2.03）和TK（Rp²=0.75, RMSEp=775.96 mg kg^-1, RPDp=2.00）的预测精度。LASSO-C模型对AP、AK、TP、Zn、Mn和Cu的预测精度优于DC模型，其中AP（Rp²=0.89, RMSEp=21.37 mg kg^-1, RPDp=3.01）和Zn（Rp²=0.80, RMSEp=9.88 mg kg^-1, RPDp=2.32）为最优结果，这归因于LASSO算法有效地从MIR和pXRF全光谱数据中提取了特征变量，提高了预测精度。GRA模型对TP、pH、AK和Cu的预测精度最高，其Rp²值分别为0.80、0.82、0.82和0.65，RMSEp值分别为129.21 mg kg^-1、0.13、48.38 mg kg^-1和3.87 mg kg^-1，RPDp值分别为2.23、2.34、2.37和1.67。考虑到成本效益，推荐使用MIR光谱预测SOM、TN和Ca（Rp²≥0.88, RPDp≥2.87），使用pXRF光谱预测Ca、Fe和Mn（Rp²≥0.80, RPDp≥2.22）。因此，本研究证明了MIR与pXRF传感器融合在提升关键土壤养分预测精度的有效性，尤其对土壤有效养分和微量元素。

Abstract:

The timely and accurate assessment of soil nutrient information is essential for ensuring global food security and sustainable agricultural development. This study evaluated the individual and fusion performance of mid-infrared (MIR) and portable X-ray fluorescence (pXRF) spectroscopy for predicting selected soil properties. Four sensor fusion strategies were implemented: direct concatenation (DC), feature-level fusion using stability competitive adaptive reweighted sampling (sCARS) and least absolute shrinkage and selection operator (LASSO) algorithms (sCARS-C and LASSO-C), multi-block fusion via sequential orthogonal partial least squares (SO-PLS), and Granger-Ramanathan model averaging (GRA) method to enhance prediction accuracy for 13 soil properties. The findings revealed that single sensor models using either MIR or pXRF provided accurate estimations for soil organic matter (SOM), total nitrogen (TN), available phosphorus (AP), calcium (Ca), iron (Fe), manganese (Mn), and pH, but showed limitations for total potassium (TK), magnesium (Mg), copper (Cu), zinc (Zn), available potassium (AK), and total phosphorus (TP). The DC model significantly improved predictions for Mg (Rp²=0.76, RMSEp=358.76 mg kg^–1, RPDp=2.03) and TK (Rp²=0.75, RMSEp=775.96 mg kg^–1, RPDp=2.00). The LASSO-C model demonstrated superior prediction accuracy compared to the DC model for AP, AK, TP, Zn, Mn, and Cu, achieving optimal results for AP (Rp²=0.89, RMSEp=21.37 mg kg^–1, RPDp=3.01) and Zn (Rp²=0.80, RMSEp=9.88 mg kg^–1, RPDp=2.32). This enhancement is attributed to LASSO's effective selection of feature information from the complete MIR and pXRF spectra. The GRA models achieved the highest prediction accuracy for TP, pH, AK, and Cu, with Rp² values of 0.80, 0.82, 0.82, and 0.65, RMSEp values of 129.21 mg kg^–1, 0.13, 48.38 mg kg^–1, and 3.87 mg kg^–1, and RPDp values of 2.23, 2.34, 2.37, and 1.67, respectively. For single-sensor applications, MIR spectra are recommended for predicting SOM, TN, and Ca (Rp²≥0.88, RPDp≥2.87), while pXRF is more cost-effective for measuring Ca, Fe, and Mn (Rp²≥0.80, RPDp≥2.22). This research demonstrates the effectiveness of MIR and pXRF sensor fusion in enhancing soil nutrient assessment accuracy, particularly for available nutrients and micronutrients.

Key words: soil available nutrients , micronutrients , multi-block fusion , feature-level fusion , model averaging

Junwei Wang, Qi Zou, Huimin Yuan. 基于中红外与便携式X射线荧光光谱的传感器融合提升土壤属性预测精度[J]. Journal of Integrative Agriculture, 2026, 25(4): 1687-1699.

Junwei Wang, Qi Zou, Huimin Yuan. Improved selected soil properties predictions using MIR and pXRF sensor fusion[J]. Journal of Integrative Agriculture, 2026, 25(4): 1687-1699.

参考文献

Andrade R, Faria W M, Silva S H G, Chakraborty S, Weindorf D C, Mesquita L F, Guilherme L R G, Curi N. 2019. Prediction of soil fertility via portable X-ray fluorescence (pXRF) spectrometry and soil texture in the Brazilian Coastal Plains. Geoderma, 357, 113960.

Antonangelo J, Zhang H. 2024. Assessment of portable X-ray fluorescence (pXRF) for plant-available nutrient prediction in biochar-amended soils. Scientific Reports, 14, 20377.

Bao S D. 1981. Soil and Agricultural Chemistry Analysis. China Agricultural Press, Beijing. pp.129–159. (in Chinese)

Barra I, Haefele S M, Sakrabani R, Kebede F. 2021. Soil spectroscopy with the use of chemometrics, machine learning and pre-processing techniques in soil diagnosis: Recent advances - A review. TrAC Trends in Analytical Chemistry, 135, 116166.

Biancolillo A, Næs T. 2019. Chapter 6 - The sequential and orthogonalized PLS regression for multiblock regression: theory, examples, and extensions. in: Cocchi M, ed., Data Handling in Science and Technology, Data Fusion Methodology and Applications. Elsevier, Oxford, UK. pp. 157–177.

Briedis C, Baldock J, de Moraes Sá J C, Dos Santos J B, Milori D M B P. 2020. Strategies to improve the prediction of bulk soil and fraction organic carbon in Brazilian samples by using an Australian national mid-infrared spectral library. Geoderma, 373, 114401.

Chang C, Laird D A, Mausbach M J, Hurburgh C R. 2001. Near-infrared reflectance spectroscopy-principal components regression analyses of soil properties. Soil Science Society of America Journal, 65, 480–490.

Dangal S, Sanderman J, Wills S, Ramirez-Lopez L. 2019. Accurate and precise prediction of soil properties from a large mid-infrared spectral library. Soil Systems, 3, 11.

Gholizadeh A, Coblinski J A, Saberioon M, Ben-Dor E, Drábek O, Demattê J A M, Borůvka L, Němeček K, Chabrillat S, Dajčl J. 2021. Vis-NIR and XRF data fusion and feature selection to estimate potentially toxic elements in soil. Sensors (Basel), 21, 2386.

Granger C W, Ramanathan R. 1984. Improved methods of combining forecasts. Journal of Forecasting, 3, 197–204.

Greenberg I, Vohland M, Seidel M, Hutengs C, Bezard R, Ludwig B. 2023. Evaluation of mid-infrared and X-ray fluorescence data fusion approaches for prediction of soil properties at the field scale. Sensors (Basel), 23, 662.

Hong Y, Munnaf M A, Guerrero A, Chen S, Liu Y, Shi Z, Mouazen A M. 2022. Fusion of visible-to-near-infrared and mid-infrared spectroscopy to estimate soil organic carbon. Soil & Tillage Research, 217, 105284.

Javadi S H, Munnaf M A, Mouazen A M. 2021. Fusion of Vis-NIR and XRF spectra for estimation of key soil attributes. Geoderma, 385, 114851.

Ji W J, Adamchuk V I, Biswas A, Dhawale N M, Sudarsan B, Zhang Y K, Viscarra Rossel R A, Shi Z. 2016. Assessment of soil properties in situ using a prototype portable MIR spectrometer in two agricultural fields. Biosystems Engineering, 152, 14–27.

Johnson J M, Vandamme E, Senthilkumar K, Sila A, Shepherd, K D, Saito K. 2019. Near-infrared, mid-infrared or combined diffuse reflectance spectroscopy for assessing soil fertility in rice fields in sub-Saharan Africa. Geoderma, 354, 113840.

Kandpal L M, Munnaf M A, Cruz C, Mouazen A M. 2022. Spectra fusion of mid-infrared (MIR) and X-ray fluorescence (XRF) spectroscopy for estimation of selected soil fertility attributes. Sensors (Basel), 22, 3459.

Karami A, Moosavi A A, Pourghasemi H R, Ronaghi A, Ghasemi-Fasaei R, Vidal E, Lado M. 2024. Proximal sensing approach for characterization of calcareous soils using multiblock data analysis. Geoderma Regional, 36, e00752.

Li H, Wang J W, Zhang J X, Liu T Q, Acquah G E, Yuan H M. 2022. Combining variable selection and multiple linear regression for soil organic matter and total nitrogen estimation by DRIFT-MIR spectroscopy. Agronomy, 12, 638.

Li H D, Xu Q S, Liang Y Z. 2018. libPLS: An integrated library for partial least squares regression and linear discriminant analysis. Chemometrics and Intelligent Laboratory Systems, 176, 34–43.

Li X L, Pan W J, Li D C, Gao W, Zeng R, Zheng G H, Cai K, Zeng Y T, Jiang C Y. 2024. Can fusion of Vis-NIR and MIR spectra at three levels improve the prediction accuracy of soil nutrients? Geoderma, 441, 116754.

Li Y Q, Pan T H, Li H R, Zou X B. 2019. NIR spectral feature selection using lasso method and its application in the classification analysis. Spectroscopy and Spectral Analysis, 39, 3809–3815. (in Chinese)

Lobsey C R, Viscarra Rossel R A, Roudier P, Hedley C B. 2017. Rs-local data-mines information from spectral libraries to improve local calibrations. European Journal of Soil Science, 68, 840–852.

Madhavan D B, Kitching M, Mendham D S, Weston C J, Baker T G. 2016. Mid-infrared spectroscopy for rapid assessment of soil properties after land use change from pastures to Eucalyptus globulus plantations. Journal of Environmental Management, 175, 67–75.

Martín Peinado F, Morales Ruano S, Bagur González M G, Estepa Molina C. 2010. A rapid field procedure for screening trace elements in polluted soil using portable X-ray fluorescence (PXRF). Geoderma, 159, 76–82.

McBride M B. 2021. Estimating soil chemical properties by diffuse reflectance spectroscopy: Promise versus reality. European Journal of Soil Science, 73, e13192.

Mukherjee S. 2022. Soil fertility and nutrient management. In: Current Topics in Soil Science. Springer, Cham. pp. 241–248.

Ni L S, Fang N F, Shi Z H, Tan W F. 2019. Mid-infrared spectroscopy tracing of channel erosion in highly erosive catchments on the Chinese Loess Plateau. Science of the Total Environment, 687, 309–318.

Ning H W, Wang J W, Jiang H, Chen Q S. 2022. Quantitative detection of zearalenone in wheat grains based on near-infrared spectroscopy. Spectrochimica Acta (Part A: Molecular and Biomolecular Spectroscopy), 280, 121545.

O’Rourke S M, Minasny B, Holden N M, McBratney A B. 2016a. Synergistic use of Vis-NIR, MIR, and XRF spectroscopy for the determination of soil geochemistry. Soil Science Society of America Journal, 80, 888–899.

O’Rourke S M, Stockmann U, Holden N M, McBratney A B, Minasny B. 2016b. An assessment of model averaging to improve predictive power of portable Vis-NIR and XRF for the determination of agronomic soil properties. Geoderma, 279, 31–44.

Peltre C, Bruun S, Du C W, Thomsen I K, Jensen L S. 2014. Assessing soil constituents and labile soil organic carbon by mid-infrared photoacoustic spectroscopy. Soil Biology and Biochemistry, 77, 41–50.

Shi T R, Fu Z C, Miao X H, Lin F F, Ma J Y, Gu S Y, Li L, Wu C F, Luo Y M. 2024, Would it be better for partition prediction of heavy metal concentration in soils based on the fusion of XRF and Vis-NIR data? Science of the Total Environment, 908,168381.

Shi X Y, Song J H, Wang H J, Lv X, Zhu YQ, Zhang W X, Bu W Q, Zeng L Y. 2022. Improving soil organic matter estimation accuracy by combining optimal spectral preprocessing and feature selection methods based on pXRF and Vis-NIR data fusion. Geoderma, 430, 116301.

Shrestha G, Calvelo-Pereira R, Roudier P, Martin A P, Turnbull R E, Kereszturi G, Jeyakumar P, Anderson C W N. 2022. Quantification of multiple soil trace elements by combining portable X-ray fluorescence and reflectance spectroscopy. Geoderma, 409, 115649.

Song J, Shi X, Wang H, Lv X, Zhang W, Wang J, Li T, Li W. 2024. Combination of feature selection and geographical stratification increases the soil total nitrogen estimation accuracy based on Vis-NIR and pXRF spectral fusion. Computers and Electronics in Agriculture, 218, 108636.

Soriano-Disla J M, Janik L J, Viscarra Rossel R A, MacDonald L M, McLaughlin M J. 2014. The performance of visible, near-, and mid-infrared reflectance spectroscopy for prediction of soil physical, chemical, and biological properties. Applied Spectroscopy Reviews, 49, 139–186.

Tavares T R, Minasny B, McBratney A, Cherubin M R, Marques G T, Ragagnin M M, Alves E E N, Padarian J, Lavres J, de Carvalho H W P. 2023. Estimating plant-available nutrients with XRF sensors: Towards a versatile analysis tool for soil condition assessment. Geoderma, 439, 116701.

Tavares T R, Molin J P, Nunes L C, Wei M C F, Krug F J, de Carvalho H W P, Mouazen A M. 2021. Multi-sensor approach for tropical soil fertility analysis: Comparison of individual and combined performance of VNIR, XRF, and LIBS spectroscopies. Agronomy, 11, 1028.

Terhoeven-Urselmans T, Vagen T, Spaargaren O, Shepherd K D. 2010. Prediction of soil fertility properties from a globally distributed soil mid-infrared spectral library. Soil Science Society of America Journal, 74, 1792–1799.

Wang J W, Liu T Q, Zhang J X, Yuan H M, Acquah G E. 2022. Spectral variable selection for estimation of soil organic carbon content using mid-infrared spectroscopy. European Journal of Soil Science, 73, e13267.

Wold S, Sjöström M, Eriksson L. 2001. PLS-regression: A basic tool of chemometrics. Chemometrics and Intelligent Laboratory Systems, 58, 109–130.

Wu M H, Lin N, Li G J, Liu H L, Li D L. 2022. Hyperspectral estimation of petroleum hydrocarbon content in soil using ensemble learning method and LASSO feature extraction. Environmental Pollutants and Bioavailability, 34, 308–320.

Xu D D, Zhao R, Li S, Chen S C, Jiang Q, Zhou L, Shi Z. 2019. Multi-sensor fusion for the determination of several soil properties in the Yangtze River Delta, China. European Journal of Soil Science, 70, 162–173.

Xu D Y, Chen S C, Xu H Y, Wang N, Zhou Y, Shi Z. 2020. Data fusion for the measurement of potentially toxic elements in soil using portable spectrometers. Environmental Pollution, 263, 114649.

Xue J, Zhang X L, Chen S C, Lu R, Wang Z, Wang N, Hong Y S, Chen X Y, Xiao Y, Ma Y X, Shi Z. 2023. The validity domain of sensor fusion in sensing soil quality indicators. Geoderma, 438, 116657.

Zhang X L, Xue J, Chen S C, Zhuo Z Q, Wang Z, Chen X Y, Xiao Y, Shi Z. 2024. Improving model performance in mapping cropland soil organic matter using time-series remote sensing data. Journal of Integrative Agriculture, 23, 2820–2841.

Zheng K Y, Li Q Q, Wang J J, Geng J P, Cao P, Sui T, Wang X, Du Y P. 2012. Stability competitive adaptive reweighted sampling (sCARS) and its applications to multivariate calibration of NIR spectra. Chemometrics and Intelligent Laboratory Systems, 112, 48–54.

Zou Q, Wang J W, Yuan H M. 2025. The accurate estimation of soil available nutrients achieved by feature selection coupled with preprocessing based on MIR and pXRF fusion. European Journal of Agronomy, 168, 127633.

基于中红外与便携式X射线荧光光谱的传感器融合提升土壤属性预测精度

Improved selected soil properties predictions using MIR and pXRF sensor fusion

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 0

编辑推荐

Metrics