نوع مقاله : مروری
عنوان مقاله English
نویسنده English
Visible and Near-Infrared (Vis–NIR) spectroscopy has emerged as a rapid, non-destructive, and cost-effective technique for estimating soil properties and supporting digital soil mapping. This review critically evaluates the performance and limitations of Vis–NIR spectroscopy for predicting soil properties in arid and semi-arid regions by synthesizing findings from studies conducted in Iran and elsewhere. The review covers the fundamental principles of soil spectroscopy, sample preparation, spectral preprocessing techniques, multivariate modeling approaches, and recent advances in multi-source data fusion. The reviewed studies demonstrate that Vis–NIR spectroscopy provides satisfactory predictions for several soil properties, including soil organic matter, texture, moisture, electrical conductivity, and calcium carbonate, particularly when coupled with advanced machine learning algorithms. However, predictive performance varies widely in arid and semi-arid soils due to the complex effects of high carbonate, gypsum, and salinity contents, mineralogical heterogeneity, and differences in spectral preprocessing, calibration strategies, and modeling techniques. The review further indicates that variations in model performance are primarily associated with the quality and representativeness of reference datasets, the availability of regional spectral libraries, and site-specific calibration rather than the choice of prediction algorithm alone. Consequently, developing regional spectral libraries, standardizing spectral measurement protocols, and integrating Vis–NIR data with remote sensing and environmental covariates are identified as key priorities for improving model transferability and practical implementation. Overall, Vis–NIR spectroscopy has considerable potential to support precision agriculture, sustainable soil management, and digital soil mapping in arid and semi-arid environments, provided that region-specific calibration and robust validation procedures are adopted.
کلیدواژهها English
Conventional procedures for soil analysis are complicated, time-consuming, and expensive, and extraction procedures may alter the equilibrium between different phases in soil. Therefore, developing techniques that preserve the basic integrity of the soil system is essential. Moreover, small-scale soil resource inventories at national, continental, and global levels need sensing and scanning technologies to precisely and rapidly describe soil properties. Over the last 20 years, the application of visible (vis), near-infrared (NIR), and mid-infrared (mid-IR) diffuse reflectance spectroscopy (DRS) in soil science garnered considerable attention. Diffuse reflectance spectroscopy is one of the Proximal Soil Sensing (PSS) methods. PSS technologies include field-based methods that can be used to measure soil properties from a distance of less than 2 m above the soil surface. The advantages of PSS are the collection of larger amounts of spatial data using non-destructive, inexpensive, and simple methods with small sample sizes, minimal sample preparation, and less labor-intensive techniques. Furthermore, PSS measurements are made in situ, providing the information at field conditions and on time.
The objective of the review paper was to explain the principles and foundations of spectroscopy, introduce the steps of sample preparation, data pre-processing, and multivariate modeling, examine studies conducted in Iran and worldwide, and assess the capabilities and limitations of the DRS technique.
The standard method for soil physical and chemical analyses is air-drying and passing through a 2 mm sieve, which is also suitable for spectroscopic studies. Grinding and sieving are performed to remove particles larger than 2 mm and coarse plant debris. The presence of instrumental, environmental, and illumination-related errors in the spectral curve highlights the necessity of reviewing and examining the spectral curves separately and with great care after measurement. Some common pre-processing methods include: multiplicative scatter correction (MSC), first and second derivatives, orthogonal signal correction (OSC), simple additive baseline correction, standard normal variate (SNV) correction, and simple additive baseline correction (OSC). Common smoothing methods for reducing noise in spectral signals include spectral averaging, median and mean shift filters, and the Savitzky-Golay filter.
Diffuse reflectance spectra of soil in the visible-near infrared region are not completely specific due to common absorption and overlap of some soil components. This non-specificity and the occurrence of absorption by other phenomena require the separation of spectra by mathematical methods. Therefore, the analysis of diffuse reflectance spectra of soil requires multivariate calibrations. The most common calibration methods for soil applications are based on linear regressions called multiple linear regression (MLR), principal component regression (PCR), and partial least squares regression (PLSR). In addition, the use of data analysis techniques such as artificial neural networks (ANN) and boosted regression trees is increasing. The accuracy of the predicted model is examined by the residual predictive deviation (RPD), which is the ratio of the standard deviation of the measured values to the least squares error of the model.
Visible-Near Infrared (Vis-NIR) emerged in recent years as a fast, low-cost, and non-destructive tool for determining the physical, chemical, and biological properties of soils. The present research investigated a large number of spectroscopic studies in Iran, and results indicated that this method is more accurate for properties such as soil organic matter, soil texture, carbonates, and gypsum, and can be suitable for expensive and time-consuming laboratory methods. Combinations of evaluation with machine learning methods and advanced algorithms such as support vector regression SVR, Random Forest, and neural networks have improved models in recent years. Visible-Near Infrared (Vis-NIR) can be a fast and cost-effective way to estimate the properties of known soils, but in arid and semi-arid regions, the presence of soil salinity, high levels of gypsum, and carbonates can reduce accuracy. The presence of calcareous and gypsum compounds overlaps with organic matter and other features, limiting the accuracy of the models. To increase the reliability and accuracy of the models, the development of indigenously produced libraries, local calibration, the use of big data and super-fiber sensors and satellites, and the application of advanced artificial intelligence models are being used. In general, Vis–NIR measurements are powerful tools for precision agriculture, soil health assessment, and resource management in arid and semi-arid regions, but the application of this method in these regions should be accompanied by caution and extensive validation.
The study did not receive any funding.
Fatemeh khayamim: Conceptualization, methodology and formal analysis, investigation, resources, data curation, writing—original draft preparation, writing—review and editing, visualization, supervision, project administration.
No artificial intelligence tools were used in the article writing process.
Data from this study are applicable upon request.
The authors would like to thank anonymous reviewers for their valuable suggestions in manuscript revision.
The author avoided data fabrication, falsification, plagiarism, and any form of misconduct.
The authors declare no conflict of interest.
