Recent Trends on the use of Infrared Spectroscopy for Soil Assessment

Infrared spectroscopy has emerged as a powerful tool to assess soil properties for both environmental science and agriculture. Here, we explore its recent trends and developments for soil assessment. This technique is an alternative that counters the limitations of traditional laboratory methods, offering a cost-effective and non-destructive approach. Here, the latest trends in the innovation landscape of infrared spectroscopy for soil assessment are explored, providing insights on its broad range of applications and into the future trajectory of this technology. Firstly, we delve into its applications in agriculture, highlighting its potential for prediction of many soil attributes. Next, we explore soil carbon assessment, emphasizing the importance of estimating soil organic carbon and soil carbon stock for soil quality. Soil pollution and elemental contents are addressed, focusing on the prediction of potentially toxic elements concentrations in soil, strongly relevant for environmental monitoring. Infrared spectroscopy emerges as a valuable tool for rapid and non-hazardous elemental content assessment. Soil physical properties prediction, traditionally limited to soil texture analysis, is extended through the application of novel approaches, shedding light on the broader potential of this technology for soil quality assessment. The ongoing developments in statistical modeling and technological innovation are also showcased, mainly focused on machine learning methods. Lastly, the importance of soil spectral libraries is emphasized, such as the Global Soil Spectral Calibration Library and Estimation Service, and Brazilian Soil Spectral Library. In conclusion, infrared spectroscopy has become an important tool in soil assessment, offering a multitude of applications across environmental and agricultural contexts. This review underscores the growing potential of this technology in advancing the standardization and reproducibility of sustainable soil assessment procedures, ensuring a brighter future for soil science.

Infrared spectroscopy has become increasingly popular for the evaluation of soil properties in recent times. By analyzing the response of molecular vibrations to infrared radiation [1], this analytical technique can reveal intricate details about the composition, structure, and functional groups present in soil samples [2]. Certain issues associated with conventional laboratory analyses impose limitations on the accurate assessment of soil properties, notably the time-intensive nature and the considerable expense of conducting such analyses. The use of infrared spectroscopy can lighten such problems as it is a fast and much less expensive technique than traditional analyses. It consists in a non-destructive method which presents a capacity to unveil information about various attributes from a soil sample with a single measure.

The main approaches currently driving the development of infrared spectroscopy in the context of soil assessment consists in the following subjects;

1. Applications for agriculture enhancement and innovation [3].
2. Carbon assessment [4-6].
3. Pollution evaluation and elemental contents prediction [7-10].
4. Assessment of soil physical properties [11].
5. Innovation in the context of statistical modeling and machine learning methods [12].
6. The development of soil spectral libraries [13].

This article explores the latest trends and progress in infrared spectroscopy for soil assessment, illustrating its growing role in environmental science and agriculture technology through recent studies (2022-2023). As innovative technologies are increasingly used and explored, the broad range of studies may make it difficult to visualize the direction in which innovation is headed. Therefore, this review aims to shed light on the evolving landscape of infrared spectroscopy in soil assessment, emphasizing not only its current use cases but also providing insights into the future trajectory of this technology.

The potential applications of soil infrared spectroscopy in agriculture are myriad. Recent studies include the prediction of soil available ions after the input of rock powders, with spectra in the Range of Visible (Vis), Near-Infrared (NIR), and Short-Wave Infrared (SWIR), achieving reasonable results [14]. It has also been used to predict nutrients and other attributes in agricultural soils, with successful results for Ca, CEC, TN, Mg, pH and Fe [15]. In the Vis-NIR range, it has achieved poor results for in-field prediction of pH and reasonable results for liming requirements [16]. However, in the range of Mid-Infrared (MIR), it has presented good results for pH (R² > 0.92) in a research upon the measurement of lime movement and dissolution in acidic soils [17]. The MIR range has also been tested for prediction of a broad range of attributes from forest soils, achieving good results for TC, TN, available Al, available Ca, CEC, and base saturation [18]. In a different approach, the range of Vis-NIR-MIR was successfully applied for the prediction of salinity and alkalinity indicators [19], which are important attributes for soil agricultural quality. These recent findings point to a future in which the use of infrared spectroscopy for agricultural purposes can become a routine practice.

Possibly the most studied subject in the field of infrared spectroscopy applied to soil assessment, soil carbon is an important attribute related to soil quality, soil microbiology, soil fertility, soil physical structure, and climate change. The estimation of Soil Organic Carbon (SOC) is a common approach, recently conducted by comparing the near and mid-infrared ranges on the prediction of SOC fractions in agricultural soils, allowing good accuracy for SOC distribution in particle-size fractions [20]. In a different approach for prediction of SOC, Vis-NIR and MIR were fused by means of different algorithms, achieving reliable results [21]. The use of hyper spectral imaging spectroscopy on the Vis-NIR range obtained from soil profiles’ measurements were also tested for the prediction of SOC, demonstrating the potential use of this technology for predicting SOC in undisturbed soil profiles [22]. Soil Organic Matter (SOM) is also an important attribute derived from soil carbon which has been used as target for prediction via FTIR spectroscopy, allowing for the comparison of SOM composition in different soil types [23].

Carbon stock or storage in soil presents relevance for both environmental and agricultural contexts, and has been explored by means of portable MIR spectroscopy in vineyard soils, achieving reliable results for varying geopedological conditions [24]. Soil carbon stock was also mapped via near-infrared spectroscopy in a desertification region from northeastern Brazil, a work in which the authors indicated that this technology could be used by government agencies for environmental monitoring [25]. The estimation of Soil Inorganic Carbon (SIC) is also an interesting approach which has been undertaken in the Vis-NIR range, coupling variable selection algorithms and deep learning methods, detecting SIC content quickly and accurately [26]. Carbon (C) is often analyzed in combination with Nitrogen (N) in studies of this nature. C and N contents has been successfully predicted in salt-affected soils, suggesting a relationship between salinity and C and N prediction [27]. Carbon to Nitrogen ratio (C:N) is an attribute related to soil nutrient availability which has been tested for quantification and depth distribution in forest soils using national-scale spectral data, yielding good results for soils with low to moderate C:N and poor results for high C:N values [28]. The growing attention towards the importance of C points to a continuous investigation on the potential of infrared spectroscopy for the prediction of C properties in soil.

The presence of Potentially Toxic Elements (PTE) in soil is an environmental problem which has a direct impact on human health, especially when occurring in agricultural soils. Therefore, determining the elemental contents and levels of pollution in soils is of utmost importance. However, the determination of PTE in soil is expensive, time-consuming and demand hazardous reagents. Such reality has led to diverse approaches in using infrared spectroscopy for prediction of soil elemental concentrations. The concentrations of Cd, Cu, Pb, Ni, Cr, Zn, Mn, and Fe in cultivated soils have been monitored by means of NIR spectroscopy, achieving reliable results for Cu, Pb, Zn, Mn, and Fe [29]. A rural catchment in southern Brazil was studied for the prediction of several PTE in soils and sediments, in which was found that the models for soils and sediments must be calibrated separately for yielding reliable results [30].

In a large river basin located in a transition between semiarid and coastal humid tropical climate zones in northeastern Brazil, several PTE were predicted in soil and sediments via NIR spectroscopy combined with Random Forest models, with satisfactory results obtained for Al, Ti, Sc, and V [31]. In a different and innovative approach, heavy metal concentrations were estimated in Technosols, with results pointing out that predicting PTE with traditional modeling approaches in Technosols is challenging [32]. One of the main difficulties in the prediction of PTE is developing models for large areas. By using regional scale sampling in northern Belgium, reasonable results were achieved for the prediction of Ni, Co, Cd, and Pb [33]. The search for an alternative to the problematic traditional analyses for determining PTE concentrations in soil is an ongoing quest, and the testing of infrared spectroscopy in different environmental settings to fill this gap remains an open challenge.

The classic approach for investigating soil physical properties via infrared spectroscopy were mainly focused on soil texture, i.e. quantification of sand, silt, and clay contents in soil samples. However, novel approaches increase the range of possibilities that this technology can explore. Soil bulk density, which is a key parameter for soil quality, have been successfully predicted by using MIR spectral libraries from Irish soils [34]. The aggregation of soil particles is a powerful indicator of soil quality. Soil aggregate stability indices consist in calculations which take into account the measured values of Aggregate Stability (AS). Vis-NIR spectroscopy has been tested in Belgium soils for prediction of AS, achieving reliable results [35]. With the same goal, the MIR range has also been tested with samples from mainland France in a work which pointed out that common soil properties and MIR spectral data yielded similar results in estimating AS values [36]. The dynamics of soil aggregate formation is also a worthy research subject, which was explored by means of MIR spectroscopy and yielded interesting results [37]. These recent approaches indicate that distinct soil physical properties are within the range of infrared spectroscopy potential as an analytical tool.

As infrared spectroscopy finds widespread application in the study of various soil attributes, researches in statistical innovation for this purpose are also expanding. Novel approaches to modeling and technological improvements are on the rise. To account for the prediction of Soil Organic Carbon (SOC), a comparison of multiple deep learning methods and variable selection algorithms was undertaken, yielding the best results by combining Interval Random Frog variable selection algorithm and Long Short Term Memory recurrent neural network [11]. In another approach for the improvement of SOC prediction, the removal of external factors [e.g. soil moisture] from on-line Vis-NIR spectra was attempted to be done by means of different algorithms, with the best result achieved using orthogonal signal correction [38]. Also for SOC modeling, the fusion of Vis-NIR and MIR was conducted by means of a parallel input-convolutional neural network, resulting in an efficient tool for in-field monitoring [39].

For the prediction of soil organic matter from NIR spectroscopy, different pre-processing methods were combined with multivariate algorithms, yielding for the best combination the accuracy of 85% in comparison with the traditional analysis [40]. MIR spectra from a large soil spectral library was combined with a convolutional neural network for the prediction of exchangeable potassium, and it was found, by comparing with a baseline partial least squares regression, that this approach greatly improves the model performance [41]. By fusing NIR spectroscopy and image information, it was possible to build a vehicle-mounted total nitrogen content prediction system capable of achieving a reliable accuracy for farmland-scale [42]. The estimation of soil texture was also a target for recent innovations by the fusion of NIR spectroscopy and image data which were fed to a convolutional neural network, demonstrating the potential to accurately predict soil and clay fractions [43]. Such examples are but a fraction from the various enhancements being developed all over the world in the context of modeling innovations for infrared spectroscopy applied to soil assessment.

Data unavailability is probably the main obstacle for training and testing models, developing novel approaches, and further exploring the capabilities of soil infrared spectroscopy. Good quality data is the raw material which keeps the research in an upward direction. To account for such deficit, many research groups and scientific agencies around the world are engaged in the development of soil spectral libraries. The Global Soil Spectral Calibration Library and Estimation Service, backed by Global Soil Laboratory Network and the Soil Spectroscopy for Global Good network, offers a valuable global resource, particularly benefiting developing countries with limited soil data and resources for conventional soil analyses [44]. The Brazilian Soil Spectral Library, which already has NIR and laboratory data from multiple soil samples, was recently enhanced with 4309 samples in the MIR range, contributing to the data availability for researches in this context [45]. The successful development of both Vis-NIR and MIR soil spectral libraries in New Zealand highlighted its ability to rapidly and non-destructively predict a wide range of soil properties, with MIR outperforming Vis-NIR in modeling performance [46]. Some recent works are directly linked with the use of soil spectral libraries, such is the data mining of urban soil for estimating organic carbon [47], and the GLOBAL-LOCAL approach, which aims to use large global Vis-NIR soil libraries to develop local models [48]. It is possible to state that the development of soil spectral libraries is a stepping stone for the standardization and reproducibility of soil infrared spectroscopy procedures.

Infrared spectroscopy, as detailed in this comprehensive review, has proven itself as an important tool for soil assessment. Its applications in agriculture, carbon analysis, pollution detection, physical property evaluation, and the development of soil spectral libraries highlight its versatility and enduring significance in environmental science and agriculture. The innovations in regard of the modeling approaches and technological applications of this technique shed light into a promising future. As the demand for rapid, cost-effective, and sustainable solutions in soil science continues to grow, infrared spectroscopy stands at the forefront, aiming for standardized procedures and generalized approaches for accurate assessments, ultimately driving sustainable advancements in environmental and agricultural practices.

There are no conflicts of interest in this work.

1. Mohamed ES, Saleh AM, Belal AB, Gad A. Application of near-infrared reflectance for quantitative assessment of soil properties. The Egyptian Journal of Remote Sensing and Space Science. 2018;21(1):1-14. doi: 10.1016/j.ejrs.2017.02.001.
2. Ng W, Minasny B, Jeon SH, Mc Bratney A. Mid-infrared spectroscopy for accurate measurement of an extensive set of soil properties for assessing soil functions. Soil Security. 2022;6:100043. doi: 10.1016/j.soisec.2022.100043.
3. Ahmadi A, Emami M, Daccache A, He L. Soil properties prediction for precision agriculture using visible and near-infrared spectroscopy: A systematic review and meta-analysis. Agronomy. 2021;11(3):433. doi: 10.3390/agronomy11030433.
4. Reeves III JB. Near-versus mid-infrared diffuse reflectance spectroscopy for soil analysis emphasizing carbon and laboratory versus on-site analysis: where are we and what needs to be done? Geoderma. 2010;158(1-2):3-14. doi: 10.1016/j.geoderma.2009.04.005.
5. Bellon-Maurel V, McBratney A. Near-infrared (NIR) and mid-infrared (MIR) spectroscopic techniques for assessing the amount of carbon stock in soils–Critical review and research perspectives. Soil Biology and Biochemistry. 2011;43(7):1398-1410. doi:10.1016/j.soilbio.2011.02.019.
6. Angelopoulou T, Balafoutis A, Zalidis G, Bochtis D. From laboratory to proximal sensing spectroscopy for soil organic carbon estimation-A review. Sustainability. 2020;12(2):443. doi: 10.3390/su12020443.
7. Shi T, Chen Y, Liu Y, Wu G. Visible and near-infrared reflectance spectroscopy-an alternative for monitoring soil contamination by heavy metals. J Hazard Mater. 2014 Jan 30;265:166-76. doi: 10.1016/j.jhazmat.2013.11.059. Epub 2013 Dec 7. PMID: 24361494.
8. Horta A, Malone B, Stockmann U, Minasny B, Bishop TFA, McBratney AB, R Pallasser, Pozza L. Potential of integrated field spectroscopy and spatial analysis for enhanced assessment of soil contamination: A prospective review. Geoderma. 2015;241:180-209.  doi: 10.1016/j.geoderma.2014.11.024.
9. Cozzolino D. Near infrared spectroscopy as a tool to monitor contaminants in soil, sediments and water-State of the art, advantages and pitfalls. Trends in Environmental Analytical Chemistry. 2016;9:1-7. doi: 10.1016/j.teac.2015.10.001.
10. Jia X, O'Connor D, Shi Z, Hou D. VIRS based detection in combination with machine learning for mapping soil pollution. Environ Pollut. 2021 Jan 1;268(Pt A):115845. doi: 10.1016/j.envpol.2020.115845. Epub 2020 Oct 13. PMID: 33120345.
11. Liu J, Xie J, Han J, Wang H, Sun J, Li R, Li S. Visible and near-infrared spectroscopy with chemometrics are able to predict soil physical and chemical properties. Journal of Soils and Sediments. 2020;20:2749-2760. doi: 10.1007/s11368-020-02623-1.
12. Wang Y, Chen S, Hong Y, Hu B, Peng J, Shi Z. A comparison of multiple deep learning methods for predicting soil organic carbon in Southern Xinjiang, China. Computers and Electronics in Agriculture. 2023;212:108067. doi: 10.1016/j.compag.2023.108067.
13. Viscarra Rossel RA, Behrens T, Ben-Dor E, Brown DJ, Demattê JAM, Shepherd KD, Shi Z, Stenberg B, Stevens A, Adamchuk V, Aïchi H, Barthès BG, Bartholomeus HM, Bayer AD, Bernoux M, Böttcher  K, Brodský L, Du CW, Chappell A, Fouad Y, Genot V, Gomez C, Grunwald S, Gubler A, Guerrero C, Hedley CB, Knadel , Morrás HJM, Nocita M, Ramirez-Lopez L, Roudier P, Rufasto Campos EM, Sanborn P, Sellitto VM, Sudduth KA, Rawlins BG, Walter C, Winowiecki LA, Hong SY, Ji W. A global spectral library to characterize the world's soil. Earth-Science Reviews.2016;155:198-230. doi: 10.1016/j.earscirev.2016.01.012.
14. Rodrigues M, Argenta JC, Cezar E, dos Santos GLAA, Özal Ö, Reis AS, Nanni MR. The use of Vis-NIR-SWIR spectroscopy in the prediction of soil available ions after application of rock powder. Information Processing in Agriculture. 2022.
15. Lelago A, Bibiso M. Performance of mid infrared spectroscopy to predict nutrients for agricultural soils in selected areas of Ethiopia. Heliyon.2022;8(3). doi: 10.1016/j.heliyon.2022.e09050
16. Sleep B, Mason S, Janik L, Mosley L. Application of visible near-infrared absorbance spectroscopy for the determination of Soil pH and liming requirements for broad-acre agriculture. Precision Agriculture. 2022;23(1):194-218. doi:10.1007/s11119-021-09834-7.
17. Hume R, Marschner P, Mason S, Schilling RK, Hughes B, Mosley LM. Measurement of lime movement and dissolution in acidic soils using mid-infrared spectroscopy. Soil and Tillage Research. 2023;233:105807. doi:10.1016/j.still.2023.105807.
18. Garrett LG, Sanderman J, Palmer DJ, Dean F, Patel S, Bridson JH, Carlin T. Mid-infrared spectroscopy for planted forest soil and foliage nutrition predictions, New Zealand case study. Trees, Forests and People.2022;8:100280. doi: 10.1016/j.tfp.2022.100280.
19. Lotfollahi L, Delavar MA, Biswas A, Fatehi S, Scholten T. Spectral prediction of soil salinity and alkalinity indicators using visible, near-, and mid-infrared spectroscopy. Journal of Environmental Management. 2023;345:118854. doi: 10.1016/j.jenvman.2023.118854.
20. Ramifehiarivo N, Barthès BG, Cambou A, Chapuis-Lardy L, Chevallier T, Albrecht A, Razafimbelo T. Comparison of near and mid-infrared reflectance spectroscopy for the estimation of soil organic carbon fractions in Madagascar agricultural soils. Geoderma Regional. 2023;33:e00638. doi: 10.1016/j.geodrs.2023.e00638.
21. Hong Y, Munnaf MA, Guerrero A, Chen S, Liu Y, Shi Z, & Mouazen AM. Fusion of visible-to-near-infrared and mid-infrared spectroscopy to estimate soil organic carbon. Soil and Tillage Research. 2022; 217:105284. doi: 10.1016/j.still.2021.105284.
22. Liu S, Chen J, Guo L, Wang J, Zhou Z, Luo J, Yang R. Prediction of soil organic carbon in soil profiles based on visible–near-infrared hyperspectral imaging spectroscopy. Soil and Tillage Research. 2023;232:105736. doi: 10.1016/j.still.2023.105736.
23. Pärnpuu S, Astover A, Tõnutare T, Penu P, Kauer K. Soil organic matter qualification with FTIR spectroscopy under different soil types in Estonia. Geoderma Regional. 2022;28:e00483. doi: 10.1016/j.geodrs.2022.e00483.
24. Wehrle R, Coulouma G, Pätzold S. Portable mid-infrared spectroscopy to predict parameters related to carbon storage in vineyard soils: Model calibrations under varying geopedological conditions. Biosystems Engineering. 2022;222:1-14. doi: 10.1016/j.biosystemseng.2022.07.012.
25. Sousa JCGD, Silva YJABD, Martins V, Rodrigues S, Teixeira MPR, Dalto PH, Barbosa RS. Diffuse Reflectance Spectroscopy for Mapping Soil Carbon Stock in the Gilbués Desertification Region at Brazilian Cerrado. Land.2023;12(9):1812. doi: 10.3390/land12091812.
26. Bai Z, Chen S, Hong Y, Hu B, Luo D, Peng J, Shi Z. Estimation of soil inorganic carbon with visible near-infrared spectroscopy coupling of variable selection and deep learning in arid region of China. Geoderma.2023;437:116589. doi: 10.1016/j.geoderma.2023.116589.
27. Cambou A, Barthès BG, Moulin P, Chauvin L, Masse D, Chevallier T, Chapuis-Lardy L, Faye E. Prediction of soil carbon and nitrogen contents using visible and near infrared diffuse reflectance spectroscopy in varying salt-affected soils in Sine Saloum (Senegal). Catena.2022;212:106075. doi: 10.1016/j.catena.2022.106075.
28. Gholizadeh A, Saberioon M, Pouladi N, Ben-Dor E. Quantification and depth distribution analysis of carbon to nitrogen ratio in forest soils using reflectance spectroscopy. International Soil and Water Conservation Research. 2023;11(1):112-124. doi: 10.1016/j.iswcr.2022.06.004.
29. Krzebietke S, Daszykowski M, Czarnik-Matusewicz H, Stanimirova I, Pieszczek L, Sienkiewicz S,  Wierzbowska J. Monitoring the concentrations of Cd, Cu, Pb, Ni, Cr, Zn, Mn and Fe in cultivated Haplic Luvisol soils using near-infrared reflectance spectroscopy and chemometrics. Talanta. 2023;251:123749. doi: 10.1016/j.talanta.2022.123749.
30. Naibo G, Ramon R, Pesini G, Moura-Bueno JM, Barros CA, Caner L, Tiecher T. Near-infrared spectroscopy to estimate the chemical element concentration in soils and sediments in a rural catchment. Catena.2022;213:106145. doi: 10.1016/j.catena.2022.106145.
31. Maia AJ, Nascimento RC, da Silva YJAB, do Nascimento CWA, de Sousa Mendes W, Neto JGV, Tiecher T, Araújo J, da Silva YJAB. Near-infrared spectroscopy for prediction of potentially toxic elements in soil and sediments from a semiarid and coastal humid tropical transitional river basin. Microchemical Journal. 2022;179:107544. doi: 10.2139/ssrn.4031424.
32. Kästner F, Sut-Lohmann M, Ramezany S, Raab T, Feilhauer H, Chabrillat S. Estimating heavy metal concentrations in Technosols with reflectance spectroscopy. Geoderma. 2022;406:115512. doi: 10.1016/j.geoderma.2021.115512.
33. Nyarko F, Tack FMG, Mouazen AM. Potential of visible and near infrared spectroscopy coupled with machine learning for predicting soil metal concentrations at the regional scale. Sci Total Environ. 2022 Oct 1;841:156582. doi: 10.1016/j.scitotenv.2022.156582. Epub 2022 Jun 14. PMID: 35714741.
34. Shi L, O'Rourke S, de Santana FB, Daly K. Prediction of soil bulk density in agricultural soils using mid-infrared spectroscopy. Geoderma. 2023;434:116487. doi: 10.1016/j.geoderma.2023.116487.
35. Afriyie E, Verdoodt A, Mouazen AM. Potential of visible-near infrared spectroscopy for the determination of three soil aggregate stability indices. Soil and Tillage Research. 2022;215:105218. doi: 10.1016/j.still.2021.105218.
36. Clergue TC, Saby NP, Wadoux AMC, Barthès BG, Lacoste M. Estimating soil aggregate stability with infrared spectroscopy and pedotransfer functions. Soil Security. 2023;11:100088. doi: 10.1016/j.soisec.2023.100088.
37. Zhu Z, Minasny B, Field DJ, An S. Using mid-infrared diffuse reflectance spectroscopy to investigate the dynamics of soil aggregate formation in a clay soil. Catena. 2023;231:107366. doi: 10.1016/j.catena.2023.107366.
38. Munnaf MA, Mouazen AM. Removal of external influences from on-line vis-NIR spectra for predicting soil organic carbon using machine learning. Catena. 2022;211:106015. doi: 10.1016/j.catena.2022.106015.
39. Hong Y, Chen S, Hu B, Wang N, Xue J, Zhuo Z, Yang Y, Chen Y, Peng J, Liu Y, Mouazen AM, Shi Z. Spectral fusion modeling for soil organic carbon by a parallel input-convolutional neural network. Geoderma. 2023;437:116584. doi: 10.1016/j.geoderma.2023.116584.
40. Carvalho JK, Moura-Bueno JM, Ramon R, Almeida TF, Naibo G, Martins AP, Lenio SS, Clesio G, Tiecher T. Combining different pre-processing and multivariate methods for prediction of soil organic matter by near infrared spectroscopy (NIRS) in Southern Brazil. Geoderma Regional. 2022;29:e00530. doi: 10.1016/j.geodrs.2022.e00530.
41. Albinet F, Peng Y, Eguchi T, Smolders E, Dercon G. Prediction of exchangeable potassium in soil through mid-infrared spectroscopy and deep learning: From prediction to explainability. Artificial Intelligence in Agriculture. 2022;6:230-241. doi: 10.1016/j.aiia.2022.10.001.
42. Wang W, Yang W, Zhou P, Cui Y, Wang D, Li M. Development and performance test of a vehicle-mounted total nitrogen content prediction system based on the fusion of near-infrared spectroscopy and image information. Computers and Electronics in Agriculture. 2022;192:106613. doi: 10.1016/j.compag.2021.106613.
43. Ebrahimi MKV, Lee H, Won J, Kim S, Park SS. Estimation of soil texture by fusion of near-infrared spectroscopy and image data based on convolutional neural network. Computers and Electronics in Agriculture. 2023;212:108117. doi: 10.1016/j.compag.2023.108117.
44. Shepherd KD, Ferguson R, Hoover D, van Egmond F, Sanderman J, Ge Y. A global soil spectral calibration library and estimation service. Soil Security. 2022;7:100061. doi: 10.1016/j.soisec.2022.100061.
45. Mendes WS, Dematte JA, Rosin NA, da Silva Terra F, Poppiel RR, Urbina-Salazar DF, Luiz Boechat C, Benedet da Silva E, Curi N, Godinho Silva SH, Santos UJ, Valladares GS. The Brazilian soil mid-infrared spectral library: the power of the fundamental range. Geoderma. 2022;415:115776. doi: 10.1016/j.geoderma.2022.115776.
46. Ma Y, Roudier P, Kumar K, Palmada T, Grealish G, Carrick S, Lilburne L, Triantafilis J. A soil spectral library of New Zealand. Geoderma Regional. 2023;e00726. doi:  10.1016/j.geodrs.2023.e00726.
47. Hong Y, Chen Y, Chen S, Shen R, Hu B, Peng J, Wang N, Guo L, Zhuo Z, Yang Y, Liu Y, Mouazen AM, Shi Z.. Data mining of urban soil spectral library for estimating organic carbon. Geoderma. 2022;426:116102. doi: 10.1016/j.geoderma.2022.116102.
48. Luce MS, Ziadi N, & Rossel RAV. GLOBAL-LOCAL: A new approach for local predictions of soil organic carbon content using large soil spectral libraries. Geoderma. 2022;425:116048. doi: 10.1016/j.geoderma.2022.116048.