No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
On the use of remote sensing techniques for monitoring spatio-temporal soil organic carbon dynamics in agricultural systems
Abstract
Soil organic carbon (SOC) dynamics affect soil quality, agricultural productivity and atmospheric CO2 concentration. Despite the need for spatial assessments of SOC content over time, reliable estimates from traditional field survey methods are limited by data availability; where measurements are often made at discrete point locations, at a coarse sample spacing or over a limited spatial extent. Remote sensing (RS) is in a strong position to provide spatially distributed, reproducible, scale-appropriate and resource-efficient measurements of SOC content and fluxes at field, landscape and regional scales. This paper provides a critical review of optical RS techniques for such applications. The first part of the paper reviews the methods, instruments and techniques used for developing predictive models for monitoring spatial SOC content. Secondly, sources of spatio-temporal SOC variations are examined, including the lateral transfer of SOC by erosion, soil structural breakdown and land management practices, in the context of RS data and techniques. The key challenges of using RS to monitor SOC contents are discussed along with opportunities for improving SOC predictions within a spatial framework. Such opportunities include the use of ancillary data, scale-specific methods, improved development of spectral libraries and better integration of RS technologies into empirical and simulation SOC models. This paper aims to provide a transparent assessment and practical guide to RS techniques and products in order to further advance and better incorporate the use of RS methods within soil science.
... According to Croft et al. (2012), there have been considerably fewer studies using airborne and satellite platforms compared to proximal RS. Access to data, cost and training requirements affect the accessibility of airborne/satellite-derived reflectance products. ...
... Access to data, cost and training requirements affect the accessibility of airborne/satellite-derived reflectance products. Moreover, there is an increased complexity in deriving reflectance data from a pixel compared to controlled laboratory conditions, due to changes in illumination angles, terrain effects, atmospheric attenuation and low signal-to-noise ratios (Ben-Dor et al., 2002;Croft et al., 2012). Thus, using satellite RS to study soil properties involves multiple challenges. ...
... RS imagery has been used for several decades to determine soil properties, but the data have limitations in terms of access, cost, training requirements, and spatial and spectral resolution (Croft et al., 2012). Vis-NIR spectroscopy is therefore presented as a means to study soil ...
Reflectance spectroscopy can be used to non-destructively characterize materials for a wide range of applications. In this study, visible-near infrared (Vis-NIR) spectroscopy was evaluated for the prediction of diverse soil properties (clay content, SOC, TN, and pH) related to different soil samples from the Eastern Cape Province in South Africa. Soil samples were scanned by a portable spectrometer at 1 nm wavelength resolution from 350 to 2500 nm. Calibrations between soil properties obtained from digital soil maps and reflectance spectra were then developed using cross-validation under partial least squares regression (PLSR) and support vector machine regression (SVMR). Raw reflectance and Savitzky-Golay first derivative data were used separately for all the samples in the data set. Key wavelengths to predict clay content, SOC, TN, and pH were identified using the variable importance projection (VIP) and Boruta algorithms. Data were additionally divided into two random subsets of 70 and 30% of the full data, which were each used for calibration and validation. The results indicated that Vis-NIR spectroscopy can be successfully used to predict soil clay content, SOC, TN and pH.
... Practical limitations include the cost of taking soil inventories and the inability to determine the spatiotemporal changes in soil resulting from changes in land use and land management. Spatiotemporal methods using remote sensing technologies provide a means of measuring, monitoring and verifying SOC stocks (Post et al, 2001, Croft et al, 2012, Smith et al, 2012, Minasny et al, 2013. ...
... Such uncertainty in the direction of change is explained by the complex interaction between the soil and the vegetation, land management methods (e.g. manure management, crop-fallow succession), erosion processes (Croft et al, 2012), climate and soil properties (Guo and Gifford, 2002) and whether the study design is based on a chronosequence or time series (Fujisaki et al, 2015). . DSM sets out to create soil databases at a given resolution by using field and laboratory observation methods coupled with spatial and non-spatial environmental data (covariates) through quantitative relationships (Boentinger et al, 2010). ...
... The datasets have the reflectance of the land surface in the visible and near infra-red regions which is widely used to indicate the primary and ecological productivity. The reflectance in the middle and thermal infrared bands depends mostly on the soil properties (Mulder et al, 2011, Croft et al, 2012. ...
Le changement d’usage des terres, lié à l’agriculture et à la foresterie, engendre une perte importante de biodiversité et représente une part importante de nos émissions de gaz à effet de serres à l’origine du changement climatique. Le mécanisme de Réduction des Émissions liéesà la Déforestation et à la Dégradation des forêts, conservation, gestion durable et restauration des stocks de carbone (REDD+) initié il y a dix ans peine à se mettre en place du fait de nombreuses contraintes politiques et scientifiques. Malgré l’existence de lignes directrices élaborées par la communauté scientifique internationale, des outils et données sont nécessaires afin de fournir des informations précises, à moindre coût et utilisables à différentes échelles. L’objectif de cette thèse est de développer des méthodologies innovantes pour réduire les incertitudes sur les estimations des émissions et séquestrations de CO2 associées à la déforestation, dégradation et régénération des terres. Madagascar, pays engagédans la REDD+ depuis huit ans, et soumis à des pertes importantes de biodiversité et de couvert forestier, est pris comme exemple. Trois études complémentaires ont été réalisées : i) le suivi de la déforestation en région tropicale humide et sèche par satellites, ii) l’estimation des stocks de carbone dans les sols et les forêts et iii) la modélisation des changements d’usage de terres. Nous avons développé une nouvelle méthodologie de suivi de la déforestation à Madagascar permettant de tenir compte de la définition des forêts et d'améliorer la prise en compte des petites parcelles de défriche brûlis. Les chiffres de la déforestation ont ainsi été actualisés jusqu’en 2013. Une méthodologie innovante de cartographie des stocks de carbone dans le sol à des résolutions fines et à des échelles régionales a été mise au point en couplant plusieurs facteurs environnementaux etdesinventaires de terrain à l’aide d’un modèle d’arbres de « forêts aléatoire » (Random Forests). Ce modèle spatial du carbone a été appliqué sur des images satellites acquises vingt années plus tôt afin d’évaluer la dégradation des stocks de carbone du sol et leur régénération potentielle. Des facteurs de perte et gain de carbone dans le sol ont pu ainsi être estimés. Enfin, une approche de modélisation des changements d’usage des terres a permis de mieux comprendre les facteurs biophysiques et socio-économiques liés à la déforestation, dégradation des terres et régénération, et de proposer des scénarios spatialisés pour aider les décideurs. Les résultats obtenus dans cette thèse et les méthodologies développées permettentd’alimenter les discussions et documents concernant la stratégie REDD+ de Madagascar. Elle contribue plus largement à fournir des informations spatiales justes, précises spatialement et cohérentes à large échelle dans le but d’améliorer la gestion de nos écosystèmes terrestres.
... Soil reflectance varies according to chemical factors, such as soil mineralogy, SOM content and soil moisture, and also physical structure, such as surface roughness and particle size. Soil spectral signatures are defined by the reflectance of electromagnetic radiation by chemical substances as a function of wavelength (Croft et al. 2012). The color of the soil is usually closely related to its organic matter content, with darker soils being higher in organic matter, which indicates the relationship between soil organic matter content and its visible light reflectance (Aghababaie et al. 2018). ...
... To date, there is no versatile model, which fits all over the world, and the waveband selection for different study areas is also diverse (He et al. 2009). The soil generally has reflectance spectra in the 1100-2500 nm range, including three distinct absorption peaks around 1400, 1900 and 2200 nm with a few small absorption peaks between 2200 and 2500 nm (Croft et al. 2012). Organic matter affects the spectra by decreasing the overall reflectance (in the visible wavelengths (Blue, Green, Red; 400-700 nm), near-infrared (NIR; 700-1400 nm) and shortwave infrared (SWIR; 1400-2500 nm) regions of the electromagnetic spectrum (Croft et al. 2012), therefore bands around 1100, 1600, 1700 to 1800, 2000, and 2200 to 2400 nm have been identified as being particularly important for SOC calibration (Stenberg et al. 2010). ...
... The soil generally has reflectance spectra in the 1100-2500 nm range, including three distinct absorption peaks around 1400, 1900 and 2200 nm with a few small absorption peaks between 2200 and 2500 nm (Croft et al. 2012). Organic matter affects the spectra by decreasing the overall reflectance (in the visible wavelengths (Blue, Green, Red; 400-700 nm), near-infrared (NIR; 700-1400 nm) and shortwave infrared (SWIR; 1400-2500 nm) regions of the electromagnetic spectrum (Croft et al. 2012), therefore bands around 1100, 1600, 1700 to 1800, 2000, and 2200 to 2400 nm have been identified as being particularly important for SOC calibration (Stenberg et al. 2010). ...
Regarding the importance of soil organic carbon (SOC), there is an increasing demand for knowledge of the spatiotemporal variability of SOC. Because of difficulties with traditional analytical measurement (time-consuming, expensive, and also soil data of sampling points are discrete and incapable of providing continuous and complete information regarding the total study area), spatial variability of SOC through field soil sampling cannot be obtained. Using remote sensing data to provide an indirect measurement of SOC to inform digital soil mapping has the potential to provide a more reliable and cost-effective estimation. The objective of this paper is to review the remote sensing data for mapping and evaluating SOC. There are several statistical methods, including regression models, principal component analysis, the ‘soil line’ approach, and geostatistics, which have been applied to investigate the accuracy of such estimates. Results suggest that these techniques have proven valuable; however, the limitations of each technique must be tested under wider environmental circumstances to choose the best technique for predictions. It seems that predictive equations aren’t universal, and every scene needs new regression models. However, an important benefit of remotely sensed data is to present a sampling strategy that can lead to improved representation of spatial heterogeneity of SOC.
... Very few studies have quantified temporal changes in soil properties at a regional scale. Moreover, one of the greatest challenges facing the broad-scale adoption of RS methods in soil science is the site-specific nature of the relationships between RSmeasured variables and soil properties (Croft et al., 2012). In addition, based on the close relationships between soil organic carbons (SOCs) and the spectrum (i.e., visible and mid-infrared bands), most research has been applied to investigations of SOCs, whereas soil acidification models have hardly been applied to agricultural soils. ...
... Multivariate statistical approaches have been applied to find quantitative relationships between soil properties and soil reflectance (Roelofsen et al., 2015;Xu et al., 2017;Grinand et al., 2017). In this study, reflectance in the middle and thermal infrared bands of Landsat imagery had close relationships with soil pH, which was a result similar to those observed in other studies, confirming that the above bands depended mostly on soil properties (Mulder et al., 2011;Croft et al., 2012). In addition, some researchers suggested employing auxiliary predictors, including climate, landcover, land use, DEM, soil roughness, agricultural management, clay content and soil surface moisture availability, to improve the accuracy of the prediction model at a regional scale (Croft et al., 2012;Grinand et al., 2017). ...
... In this study, reflectance in the middle and thermal infrared bands of Landsat imagery had close relationships with soil pH, which was a result similar to those observed in other studies, confirming that the above bands depended mostly on soil properties (Mulder et al., 2011;Croft et al., 2012). In addition, some researchers suggested employing auxiliary predictors, including climate, landcover, land use, DEM, soil roughness, agricultural management, clay content and soil surface moisture availability, to improve the accuracy of the prediction model at a regional scale (Croft et al., 2012;Grinand et al., 2017). This was also applied in this paper and the result was just as Schillaci et al. (2017) showed, that the integration of remote sensing imagery with other auxiliary environmental predictors increased the predictive ability compared to the models built without environmental data. ...
... Mapping SOC has also been studied with remote sensing using airborne and satellite platforms to cover extended areas although with lower precision. Consequently, it should be integrated with field and laboratory measurements and complementary sensor data for better results [50]. Our results showed the feasibility of using on-the-go soil spectra for mapping SOC with appropriate reliability, having an accuracy closer to laboratory measurements than remote sensing data. ...
... Long-term monitoring can help to identify trends and seasonal Mapping SOC has also been studied with remote sensing using airborne and satellite platforms to cover extended areas although with lower precision. Consequently, it should be integrated with field and laboratory measurements and complementary sensor data for better results [50]. Our results showed the feasibility of using on-the-go soil spectra for mapping SOC with appropriate reliability, having an accuracy closer to laboratory measurements than remote sensing data. ...
Agricultural soils serve as crucial storage sites for soil organic carbon (SOC). Their appropriate management is pivotal for mitigating climate change. Continuous monitoring is imperative to evaluate spatial and temporal changes in SOC within agricultural fields. In-field datasets of Vis-NIR soil spectra were collected on a long-term experimental site using an on-the-go spectrophotometer. Data processing for continuous SOC prediction involves a two-step modeling approach. In Step 1, a partial least square (PLSR) regression model is trained to establish a relationship between the SOC content and spectral information, including spectral preprocessing. In Step 2, the predicted SOC content obtained from the PLSR models is interpolated using ordinary kriging. Among the tested spectral preprocessing techniques and semivariogram models, Savitzky–Golay and the Gap-Segment derivative preprocessing along with a Gaussian semivariogram model, yielded the best performance resulting in a root mean square error of 1.24 and 1.26 g kg−1. A striping effect due to the transect-based data collection was addressed by testing the effectiveness of extending the spatial separation distance, employing data aggregation, and defining the distribution based on treatment plots using block kriging. Overall, the results highlight the high potential of on-the-go spectral Vis-NIR data for field-scale spatial-temporal monitoring of SOC.
... Robust quantification of change in SOC that is cost-effective and provides a statistical assessment of uncertainty is challenging; this is particularly the case in the presence of large spatial variability and slow SOC gains (Smith et al, 2020;Paustian et al., 2019;Stanley et al., 2023). For these reasons, data from universally available sources-such as freely available remote sensing imagery-that can provide an indication of temporal changes (or spatial differences) at landscape scales can help inform more effective decision making and reduce the costs associated with measurement of SOC (Ladoni et al., 2010;Croft at al., 2012;Paustian et al., 2019). ...
... Many studies have utilised remote sensing data to better model the spatial and spatio-temporal variation of SOC (Yang et al., 2009;Page et al., 2013;Wilson et al., 2017;Venter et al., 2021). However, relationships between remote-sensing variables and the temporal variation of SOC can be very site-specific (Croft et al., 2012), making it difficult to build broadly applicable predictive models of the spatiotemporal variation of SOC. Nonetheless, site-specific indicators of temporal changes can still be useful for informing the likelihood of increases or decreases in SOC. ...
Temporal variation of soil organic carbon (SOC) is driven by land use/management practice, ecosystem conditions and climatic variation. Robust quantification of changes in SOC that is cost-effective and provides a statistical assessment of uncertainty is challenging, particularly in the face of large spatial variability and slow soil SOC changes. Remote-sensing indicators of above-ground vegetation provide some indication of the amount of fresh organic material being supplied to the soil. Although, because of the time taken for this organic material to decay and become incorporated into the soil, there will be a lag between the changes in the indicator of vegetation growth and the resulting changes in SOC. In this work, we investigate how a remotely sensed indicator of vegetation cover can be used with a lag period to predict or indicate changes in SOC for grazed pasture sites at a long-term monitoring study, which has been monitoring soil under different land uses for over forty years. We assessed how well this worked for indicating the SOC changes for different depths in the soil profile. Results suggested that a lagged remotely sensed vegetation cover-the average cover of the two preceding years-provides some indication of SOC changes for the 0-10 cm soil depth, but changes for deeper soil depths were not well predicted. Further, we investigated the potential of using soil data from a point-in-time spatial dataset (e.g. data from a baseline sampling round) to calibrate a relationship between the remotely sensed cover and SOC, which can then be applied to predict or indicate the temporal variation of SOC. Results showed this approach gave large prediction errors, likely because the temporal variation (at a fixed point in space) and spatial variation (for a fixed point in time) of SOC that is predictable by cover differences are not interchangeable.
... SOC is one of the core indices of soil fertility and plays an important role in the global carbon cycle. Although SOC is only a small part of the total soil mass, it plays a positive role in soil fertility, environmental protection, plant growth, and agricultural production (Croft et al. 2012;Nawar et al. 2016;Tiessen et al. 1994;Wang et al. 2009). Farmland soils have a relatively high carbon sequestration potential and can effectively mitigate climate change (Hutchinson et al. 2007;Liu et al. 2019). ...
... Farmland SOC sequestration is the objective of international initiative "4 per 1000". The decrease in SOC content will not only lead to the degradation of farmland ecology, but also reduce the sustainable use of farmland (Croft et al. 2012;Tiessen et al. 1994;Zhang et al. 2021). SOC stocks in topsoil are more abundant but less stable under climate change. ...
Background and aimsSolar-induced chlorophyll fluorescence (SIF) is closely related to vegetation photosynthesis and can sensitively reflect the growth and health of vegetation. Using the advantages of SIF in photosynthetic physiological diagnosis, this study carried out a collaborative study of SIF, land attributes and image reflectance spectra to estimate soil organic carbon (SOC) content in typical agricultural areas of the Qinghai-Tibet plateau (QTP).Methods
The spectral reflectance (R), first derivative of reflectance (FDR), second derivative of reflectance (SDR) of spectral band of Landsat 8 Operational Land Imager (OLI) data were selected together with land attributes (i.e. elevation, slope, soil temperature, and soil moisture content) and SIF index and vegetation indices to establish the SOC content estimation models using the random forest (RF), back propagation neural network (BPNN) and partial least squares regression (PLSR), respectively.ResultsSIF index can significantly improve the SOC content estimation compared to the vegetation indices. The accuracy of the BPNN model established by combining SIF index with the FDR of Landsat 8 OLI data and land attributes was the highest (R2 = 0.977, RMSEC = 2.069 g·kg− 1, MAE = 0.945 g·kg− 1, RPD = 3.970, d-factor = 0.010).Conclusion
This study confirmed the good effect of BPNN model driven by SIF index, land attributes, and Landsat 8 OLI data on the estimation of SOC content, which can provide a new way for the accurate estimation of the soil internal components in the agricultural areas.
... It is assumed that reduced spectral resolution, as in the case of multispectral data, results in a reduction in the model's predictive capability. Despite this, a number of studies dealing with multispectral data [32][33][34] have shown that results can be satisfactorily applied to the needs of precision farming, especially with regard to acquisition costs. ...
... Moreover, accuracy and prediction ability are often affected by other factors, such as different soil conditions, variability of the analyzed characteristics, the condition of the studied surface (moisture and surface roughness affecting vegetation and crop residues), various atmospheric conditions and light incidence geometry during image acquisition [48,49]. This leads to reduced predictive ability compared to that obtained with soil laboratory spectroscopy [33,50] and makes it difficult to map SOC at a large scale, especially in temperate regions, due to crop cover and various types of land parcel management. ...
The image spectral data, particularly hyperspectral data, has been proven as an efficient data source for mapping of the spatial variability of soil organic carbon (SOC). Multispectral satellite data are readily available and cost-effective sources of spectral data compared to costly and technically demanding processing of hyperspectral data. Moreover, their continuous acquisition allows to develop a composite from time-series, increasing the spatial coverage of SOC maps. In this study, an evaluation of the prediction ability of models assessing SOC using real multispectral remote sensing data from different platforms was performed. The study was conducted on a study plot (1.45 km2) in the Chernozem region of South Moravia (Czechia). The adopted methods included field sampling and predictive modeling using satellite multispectral Sentinel-2, Landsat-8, and PlanetScope data, and multispectral UAS Parrot Sequoia data. Furthermore, the performance of a soil reflectance composite image from Sentinel-2 data was analyzed. Aerial hyperspectral CASI 1500 and SASI 600 data was used as a reference. Random forest, support vector machine, and the cubist regression technique were applied in the predictive modeling. The prediction accuracy of models using multispectral data, including Sentinel-2 composite, was lower (RPD range from 1.16 to 1.65; RPIQ range from 1.53 to 2.17) compared to the reference model using hyperspectral data (RPD = 2.26; RPIQ = 3.34). The obtained results show very similar prediction accuracy for all spaceborne sensors (Sentinel-2, Landsat-8, and PlanetScope). However, the spatial correlation between the reference mapping results obtained from the hyperspectral data and other maps using multispectral data was moderately strong. UAS sensors and freely available satellite multispectral data can represent an alternative cost-effective data source for remote SOC mapping on the local scale.
... Until now, their use was limited for soil observation due to (i) the required atmospheric, geometric and radiometric data corrections, (ii) simultaneous ground observations, (iii) the difficulty in finding large bare soil areas within a single image [44] and (iv) obstacles related to vegetation cover [45]. Consequently, there are few studies using satellite sensors for SOC estimation [46]. ...
... Certain articles were excluded due to inadequate results justification. Considering Scopus may fail to retrieve a significant percent of related works [56], previous reviews and surveys [7,9,19,27,38,46,54,[57][58][59][60][61][62][63][64][65] were further examined for related work and references from the selected articles were also evaluated, resulting in 28 articles in total. ...
Towards the need for sustainable development, remote sensing (RS) techniques in the Visible-Near Infrared–Shortwave Infrared (VNIR–SWIR, 400–2500 nm) region could assist in a more direct, cost-effective and rapid manner to estimate important indicators for soil monitoring purposes. Soil reflectance spectroscopy has been applied in various domains apart from laboratory conditions, e.g., sensors mounted on satellites, aircrafts and Unmanned Aerial Systems. The aim of this review is to illustrate the research made for soil organic carbon estimation, with the use of RS techniques, reporting the methodology and results of each study. It also aims to provide a comprehensive introduction in soil spectroscopy for those who are less conversant with the subject. In total, 28 journal articles were selected and further analysed. It was observed that prediction accuracy reduces from Unmanned Aerial Systems (UASs) to satellite platforms, though advances in machine learning techniques could further assist in the generation of better calibration models. There are some challenges concerning atmospheric, radiometric and geometric corrections, vegetation cover, soil moisture and roughness that still need to be addressed. The advantages and disadvantages of each approach are highlighted and future considerations are also discussed at the end.
... For example, RS technologies have been widely used in monitoring atmosphere (Mashyanov & Reshetov, 1995;Deng, Tian, Wang, & Chen, 2003;Emeis & Sch€ afer, 2006;Hu, Li, Zhang, & Wang, 2006;Belic, Radosavljevic, Milincic, & Sabi c, 2012;Prud'homme et al., 2013), water (Hom, 1968;Welch, 1971;Ishaq, 1985;Domenico, Crisafi, Magazz u, Puglisi, & Rosa, 1994;Hoogenboom, Dekker, & Althuis, 1998;Deng, Liu, Ke, Cheng, & Liu, 2004;Parsons, Rodriguez-Lado, Reuter, & Montanarella, 2009;Oparin, Potapov, Giniyatullina, & Andreeva, 2012), plant (Airola, 1990Houborg, Anderson, & Daughtry, 2009;Darvishzadch et al., 2008;Li, Lv, & Altermann, 2010;Emengini, Blackburn, & Theobald, 2013), snow (Tang et al., 2011;Oparin et al., 2014), and soil (Din, Dousari, & Literathy, 2008;Peng et al., 2016;Mirzaee, Ghorbani-Dashtaki, Mohammadi, Asadi, & Asadzadeh, 2016;Neto, Teixeira, Moreira, & Galvão, 2017;Xu, Smith, Grunwald, Abd-Elrahman, & Wani, 2017). In particular, the applications of RS in observing geochemical properties of soil profiles mainly focus on soil salinity (Hick & Russell, 1990;Dwivedi, Sreenivas, & Ramana, 1999;Douaoui, Nicolas, & Walter, 2006;Bouaziz, Matschullat, & Gloaguen, 2010;Neto et al., 2017), hydrocarbon pollution (Howari, 2004;Din et al., 2008), heavy-metal contamination (Swayze et al., 2000;Wu, Chen, Ji, Tian, & Wu, 2005;Choe et al., 2008;Wu, Zhang, Liao, & Ji, 2011;Asmaryan, Muradyan, Sahakyan, Saghatelyan, & Warner, 2014;Peng et al., 2016), mapping soil nutrients (Skidmore, Varekamp, Wilson, Knowles, & Delaney, 1997;Galvâo, Pizarro, & Epiphanio, 2001; Lopez-Granados, Jurado-Exposito, Pena-Barragan, & Garc ıa-Torres, 2005; Garcia- Gomez & Maestre, 2011;Lu, Wang, Niu, Li, & Zhang, 2013;Wang & Shen, 2015;Xu et al., 2017), organic matter and organic carbon (Frazier & Cheng, 1989;Palacios-Orueta, Pinzon, Ustin, & Roberts, 1999;Chen, Kissel, West, & Adkins, 2000;Ben-Dor, Patkin, Banin, & Karnieli, 2002;Fox, Sabbagh, Searcy, & Yang, 2004;Sullivan, Shaw, Rickman, Mask, & Luvall, 2005;Selige, B€ ohner, & Schmidhalter, 2006;Gomez, Viscarra Rossel, & McBratney, 2008;Ladoni, Alavipanah, Bahrami, & Noroozi, 2010;Croft, Kuhn, & Anderson, 2012;Mirzaee et al., 2016). ...
... Oerke, Gerhards, Menz, and Sikora (2010) presented the potential use of reflectance spectroscopy for retrieving useful soil parameters based on several case studies to illustrate the existing limitations for retrieving soil properties over large heterogeneous areas. Ge, Thomasson, and Sui (2011) reviewed recent publications on the subject of RS of soil properties in precision agriculture and found that a large array of agriculturally-important soil properties (including textures, organic and inorganic carbon content, macro-and micro-nutrients, moisture content, cation exchange capacity, electrical conductivity, pH, and iron) can be successfully quantified with RS. Croft et al. (2012) provided a critical review on the use of RS techniques for monitoring spatiotemporal soil organic carbon dynamics in agricultural systems to further advance and better incorporate the use of RS methods within soil science. Grunwald, Vasques, and Rivero (2015) presented an in-depth overview of proximal and RS technologies that are used in the realm of digital soil assessment and contrasted the benefits and constraints of proximal and RS, fusion of soil-environmental data, and integration pathways to mashup data and methods into complex soil assessments. ...
As a cross-disciplinary scientific domain, a great deal of research on landscape geochemistry has been conducted since its establishment. However, the used methods lack the ability of revealing the relationships of landscape geochemistry with other sciences, and the obtained knowledge is thus fragmented and isolated. In this study, the state-of-the-art regarding the applications of relatively new geospatial technologies including fractal theory, geographic information system and remote sensing to landscape geochemistry was reviewed and analyzed to provide deep insights of current research and a roadmap for furthering the development of landscape geochemistry as a cross-disciplinary discipline. The results showed that substantial research on the applications of fractal theory, GIS and RS technologies for analyzing the processes and data of landscape geochemistry has been conducted by using the advantages of the geospatial technologies. However, the great challenges still exist when the geospatial technologies were individually utilized due to the limitations of the technologies themselves and the complexity of landscape geochemistry. In the end, opportunities and challenges for advancing the further studies of landscape geochemistry were discussed in detail and new directions of studying landscape geochemistry using multidisciplinary or integrated approaches to enhance understanding the relationships among the relevant disciplines were suggested.
... However, there are still limitations to the spaceborne measurement of SOC due to temperature, air pressure, spectral calibration, and the height of the crop. Moreover, the satellite is super expensive, so it's unsuitable for general-purpose use [23]. ...
... Machine learning techniques relate coarse-fine resolution pairs of observed images to anticipate unknown images of finer resolution. A range of methods have been employed in spatiotemporal data fusion, such as dictionary-pair learning , extreme learning (Tsagkatakis et al. 2019); regression tree (Hilker et al. 2009); random forest (Saini and Ghosh 2017); deep convolutional neural networks (Yu et al. 2020); and artificial neural networks (Croft et al. 2012). The SPSTFM was the first method to apply dictionary-pair learning approaches to spatiotemporal data fusion for natural picture super-resolution. ...
For many years, spectral remote sensing has been essential for research on the Earth’s surface. The data from a single satellite sensor is sometimes insufficient to fulfil the expanding needs of remote sensing applications. Spatial-temporal fusion techniques have become an effective way for merging spectral data from many sources and times, enabling improved data analysis and interpretation. The goal of this review paper is to offer a thorough examination of the historical growth of spatio-temporal fusion techniques for spectral remote sensing. The classification of all currently used fusion approaches, such as Unmixing, Weight-based, Bayesian-based, machine learning-based, and hybrid methods, is covered in detail. Additionally, it evaluates pixel-level, decision-level, and feature-level-based data fusion techniques and compares and contrasts their advantages and disadvantages. The report also discusses spatiotemporal fusion’s difficulties and recommends future advances. For those working in remote sensing research and practice, it offers an invaluable resource. In conclusion, this review paper provides a comprehensive overview of spatio-temporal fusion systems for spectral remote sensing, including an analysis of their comparative benefits and drawbacks and a description of their historical development. It aims to stimulate further research and development of spatio-temporal fusion methods for spectral remote sensing. In summary, this review paper presents a comprehensive overview of spatio-temporal fusion methods for spectral remote sensing, including their historical development, categorization of existing techniques and applications, and a comparative analysis of their strengths and limitations. It also discusses the current challenges and future research directions, providing a valuable resource for the remote sensing community.
... There is a certain correlation between the content of SOC and the reflectance of visible near-infrared spectroscopy, providing a basis for remotely sensing SOC [11]. Therefore, remote sensing can provide a rapid, repeatable, and cost-effective means of quantitatively assessing the spatial distribution of soil properties [12]. Hyperspectral remote sensing has been widely applied in predicting SOC content in spatially small areas, such as farmland and forests [13,14]. ...
The North China agro–pastoral zone is a large, ecologically fragile zone in the arid and semi-arid regions. Quantitative remote sensing inversion of soil organic carbon (SOC) in this region can facilitate understanding of the current status of degraded land restoration and provide data support for carbon cycling research in the region. Deep learning (DNN) for SOC inversion has been W.a hot topic over the past decade, but there have been few studies at the regional scale in the arid and semi-arid zones. In this study, a DNN model with five hidden layers and five skip connections was established using 644 spatially distributed SOC samples and Landsat 8 OLI imagery. The model was compared with the random forest algorithm in terms of generalization ability. The main conclusions were as follows: 1. The DNN algorithm can establish a high-precision SOC inversion model (R2 = 0.52, RMSE = 0.7), with 90% of errors concentrated in the range of −2.5 to 2.5 kg·C/m2; 2. the Boruta variable-screening algorithm can effectively improve the model accuracy of the random forest algorithm, but due to the DNN’s better ability to mine hidden information in the data, the improvement effect on the DNN model accuracy is limited; 3. the SOC samples in arid and semi-arid areas are highly positively skewed, with a significant impact on the modeling accuracy of DNN, and conversion is required to obtain a model with better generalization ability; and 4. in arid and semi-arid regions, SOC has a weak correlation with vegetation indices but a stronger correlation with temperature, elevation, and aridity. This study established a reliable deep learning model for SOC density in a large arid and semi-arid region, providing a reference and framework for the establishment of SOC inversion models in other regions.
... Mapping SOC has also been studied with remote sensing using airborne and satellite platforms to cover extended areas although with lower precision. Consequently, it should be integrated with field and laboratory measurements and complementary ancillary data for better results (Croft et al., 2012). Our results showed the feasibility of using on-the-go soil spectra for mapping SOC with appropriate reliability, having an accuracy closer to laboratory measurements than remote sensing data. ...
Agricultural soils serve as crucial storage sites for soil organic carbon (SOC). Their appropriate management is pivotal for mitigating climate change. To evaluate spatial and temporal changes in SOC within agricultural fields, continuous monitoring is imperative. In-field data sets of Vis-NIR soil spectra were collected on a long-term experimental site using an on-the-go spectrophotometer. Data processing for continuous SOC prediction involves a two-steps modelling approach. In Step 1, a Partial Least Square (PLSR) regression model is trained to establish a relationship between the SOC content and the spectral information also including spectral preprocesisng. In Step 2, the predicted SOC content obtained from the PLSR models is interpolated using ordinary kriging. Among the tested spectral preprocessing techniques and semivariogram models, SG and gapDer preprocessing along with a Gaussian semivariogram model, yielded the best performance resulting in a root mean square error of of 1.24 and 1.26 g kg-1. A striping effect due to the transect-based data collection was addressed by testing the effectiveness of extending the spatial separation distance, employing data aggregation, and defining the distribution based on treatment plots using block kriging. Overall, the results highlight the immense potential of on-the-go spectral Vis-NIR data for field-scale spatial-temporal monitoring of SOC.
... However, areas with yellow Ferralsols also showed increased goethite abundance at depth, while it did not occur for haematite in red Ferralsols (Fig. 10b). This phenomenon is possibly explained by the masking of iron oxides by SOC in the 0-0.2 m layer (Croft et al., 2012;Heller Pearlshtien and Ben-Dor, 2020). This effect is more pronounced for goethite than for haematite, as haematite pigmentation directly affects the spectral response of the visible region. ...
... However, soil properties are also estimated with a lower spectral resolution using satellite and airborne multispectral sensors. Imaging data from these sensors are recorded in only a few bands of the VNIR range and can be used to estimate the content of soil organic carbon (SOC) (Croft et al. 2012, de Paul Obade, Lal 2013 and clay (Nanni, Dematte 2006, Demattê, Fiorio 2009). However, better results can be obtained by combining satellite data with hyperspectral measurements (Peng et al. 2015). ...
Remote sensing techniques based on soil spectral characteristics are the key to future land management; however, they still require field measurement and an agrochemical laboratory for the calibration of the soil property model. Visible and near-infrared diffuse reflectance spectroscopy has proven to be a rapid and effective method. This study aimed to assess the suitability of multispectral data acquired with the agricultural digital camera in determining soil properties. This 3.2-Mpx camera captures images in three spectral bands – green, red and near-infrared. First, the reference data were collected, which consist of 151 samples that were later examined in the laboratory to specify the granulometric composition and to quantify some chemical elements. Second, additional soil properties such as cation exchange capacity, organic carbon and soil pH were measured. Finally, the agricultural digital camera photograph was taken for every soil sample. Reflectance values in three available spectra bands were used to calculate the spectra indices. The relationships between the collected data were calculated using the independent validation regression model such as Cubist and cross-validation model like partial least square in R Studio. Additionally, different types of data normalisation multiplicative scatter correction, standard normal variate, min–max normalisation, conversion into absorbance] were used. The results proved that the agricultural digital camera is suitable for soil property assessment of sand and silt, pH, K, Cu, Pb, Mn, F, cation exchange capacity and organic carbon content. Coefficient of determination varied from 0.563 (for K) to 0.986 (for soil organic carbon). Higher values were obtained with the Cubist regression model than with partial least squares.
... Although aerial RS employing Unmanned Aerial Vehicles (UAV) or manned flights can provide high resolution data (cm resolution), it cannot cover as big an area as satellite RS in the same amount of time [26]. Finally, proximal RS, which can be performed using Unmanned Ground Vehicles (UGV), vehicles with an operator, or on foot, can retrieve high-resolution data (mm resolution), but requires long and heavy effort [5,27,28]. ...
Various Remote Sensing (RS) technologies and platforms have been widely used in olive cultivation studies over the last 16 years. These technologies and platforms have been applied throughout the olive cultivation cycle, providing significant insights into olive growth and productivity. The goal of this review was to determine the importance of RS technologies and platforms in specific agronomic focus areas in order to determine the equipment and platform requirements in olive cultivation studies. For this reason, frequency and correspondence studies were carried out. Unmanned aerial vehicles, multispectral sensors, and vigour assessment found to be important in olive cultivation studies. Additionally, each agronomic focus area in an olive cultivation study presents different needs in equipment, proximity of sensing, and coverage area, indicating that this must be taken into account during field experiments with RS. Further technological improvements will permit the use of other RS technologies and platforms during future studies. Finally, future studies are expected to focus more on RS data processing as well as on the use of unmanned aerial and ground vehicles in swarms for data collection and performance of actions.
... Particularly, remote sensing data provide abundant predictors, which are area data. The intrinsic correlations of the soil point data with the remote sensing area data are established, and then the conversion of soil data from point to area can be realized based on the intrinsic correlations by means of mathematical models (Croft et al., 2012;Chi et al., 2018a;Paul et al., 2020;Dharumarajan et al., 2021). In recent years, the soil organic carbon mapping in three-dimensional space, i.e., horizontal and vertical spatial simulation at a time point, was explored in certain areas. ...
In the context of multiple disturbances, soil organic carbon stock (SOCS) in coastal wetlands experiences drastic spatiotemporal variations, which involve four dimensions. However, the finiteness and discontinuity of historical field soil data hinder the four-dimensional SOCS reconstruction in coastal wetlands. In this study, the zonal and progressive simulations were integrated to reconstruct the four-dimensional characteristics of coastal wetland SOCS by using field soil data in current time point and remote sensing data during the last decades. The zonal simulation was adopted to conduct the two-dimensional simulation at a low cost of field survey. The spatial and temporal progressive simulations were implemented based on the close relationships among soil factors in different depths and at different time points, respectively, for realizing the three- and four-dimensional simulations. The demonstration of the study in Chongming Island, an important coastal wetland in China, validated the low cost, high accuracy, and good applicability of the approach. Over the entire island during the last decades, SOCS in surface layer showed overall increasing characteristics, while SOCS in intermediate and bottom layers did not exhibit distinct change trend. Generally, SOCS in surface layer increased and that in intermediate and bottom layers decreased along the gradient from the shoreline to the inner island. Continuous sediment discharge and deposition enlarged the areas of coastal wetlands and thus increased SOCS in the alongshore areas. Human activities distinctly increased SOCS through long-term and large-scale agricultural activities in the inner island, while decreased it through urbanization. Then, healthy, coordinated, and sustainable measures for increasing coastal wetland SOCS were proposed from perspectives of scale control, spatial configuration, and quality promotion. Therefore, this approach could reconstruct SOCS four-dimensional characteristics and achieve the conversions of soil data from point to area, from surface to bottom, and from present to past.
... Even when studies sampled SOC multiple times to investigate changes in SOC (46,73,79,96,97,102), models were built separately for each year rather than based on use of A covariates likely due to the difficulty to resample the same soil profile. The review of Croft et al. (178) showed that RS data may be promising for modeling temporal SOC changes through the monitoring of soil structural changes, soil erosion, agricultural practices in time, but the accuracy of these covariate data obtained at the large spatial extent need to be tested further. ...
To explore how well large spatial scale digital soil mapping can contribute to efforts to monitor soil organic carbon (SOC) stocks and changes, we reviewed regional and national studies quantifying SOC within lands dominated by agriculture using SCORPAN approaches that rely on soil (S), climate (C), organisms (O), relief (R), parent material (P), age (A), and space (N) covariates representing soil forming factors. After identifying 79 regional (> 10,000 km2) and national studies that attempted to estimate SOC, we evaluated model performances with reference to soil sampling depth, number of predictors, grid-distance, and spatial extent. SCORPAN covariates were then investigated in terms of their frequency of use and data sources. Lastly, we used 67 studies encompassing a variety of spatial scales to determine which covariates most influenced SOC in agricultural lands using a subjective ranking system. Topography (used in 94% of the cases), climate (87%), and organisms (86%) covariates that were the most frequently used SCORPAN predictors, aligned with the factors (precipitation, temperature, elevation, slope, vegetation indices, and land use) currently identified to be most influential for model estimate at the large spatial extent. Models generally succeeded in estimating SOC with fits represented by R2 with a median value of 0.47 but, performance varied widely (R2 between 0.02 and 0.86) among studies. Predictive success declined significantly with increased soil sampling depth (p < 0.001) and spatial extent (p < 0.001) due to increased variability. While studies have extensively drawn on large-scale surveys and remote sensing databases to estimate environmental covariates, the absence of soils data needed to understand the influence of management or temporal change limits our ability to make useful inferences about changes in SOC stocks at this scale. This review suggests digital soil mapping efforts can be improved through greater use of data representing soil type and parent material and consideration of spatio-temporal dynamics of SOC occurring within different depths and land use or management systems.
... There are many studies that estimate the SOC content based on climate humidity, or on plant AGB or remotely-sensed vegetation index such NDVI (Croft et al., 2012;Sun et al., 2019;, and little attention is paid to the effects of soil properties such as DUL or sand % on the SOC content in different climate zones. Our results suggest that soil water retention parameters modulate the relations between climate and vegetation biomass or SOC content, and DUL is a better predictor than climate humidity for predicting SOC content, and it is necessary to incorporate soil physical and hydraulic parameters for the improvement of the prediction of SOC and plant production in semi-arid steppes. ...
Soil hydraulic properties determine water transport, distribution and storage in soils, thus affect ecosystem functioning. Soil water retention parameters, such as saturation water content (SAT), water drained upper limit (DUL) and lower limit (DLL), characterize soil water-holding capacity, thus are important for plant growth and carbon and nutrient cycling in grassland ecosystem. However, we are still unclear to what a content these soil water retention parameters may regulate plant production and soil organic matter content in natural ecosystems in semi-arid grassland environment. We investigated plant species richness (SR) and aboveground biomass (AGB), soil organic carbon (SOC) and nitrogen (SON) contents, and soil water retention and other physical parameters in the steppe ecosystems along a climate transect, which spans the desert steppe, typical steppe and meadow steppe zones in Northern China. We analyzed the relative importance of climate factors, soil water retention parameters and other soil physical parameters in regulating plant AGB and SOC and SON contents using partial correlation analysis, and quantified the contribution of each separate and combined climate and soil physical parameters in predicting plant AGB and SOC and SON contents using geographic detectors method. We found that (i) plant SR and AGB, soil SAT and DUL, and SOC and SON are all significantly inter-correlated, and increased with the increase of climate humidity. (ii) Climate, but not soil physical parameters, was the most preponderant factor in predicting plant SR and AGB. (iii) Soil physical parameters were more important than climate parameters in regulating soil organic matter content; soil DUL was the best predictor to SOC content. Our results suggest that soil water retention parameters especially soil DUL, may play a greater role than climate factors in modulating SOC content, though climate is more predominant in regulating plant SR and production. Our results imply that we need to be cautious in estimating SOC content based on climate factors and vegetation biomass or remotely-sensed vegetation indices, and that soil physical properties need also to be incorporated along with vegetation parameters for improving SOC estimation at a regional scale.
... There are many studies that estimate the SOC content based on climate humidity, or on plant AGB or remotely-sensed vegetation index such NDVI (Croft et al., 2012;Sun et al., 2019;, and little attention is paid to the effects of soil properties such as DUL or sand % on the SOC content in different climate zones. Our results suggest that soil water retention parameters modulate the relations between climate and vegetation biomass or SOC content, and DUL is a better predictor than climate humidity for predicting SOC content, and it is necessary to incorporate soil physical and hydraulic parameters for the improvement of the prediction of SOC and plant production in semi-arid steppes. ...
Soil hydraulic properties determine water transport, distribution and storage in soils, thus affect ecosystem functioning. Soil water retention parameters, such as saturation water content (SAT), water drained upper limit (DUL) and lower limit (DLL), characterize soil water-holding capacity, thus are important for plant growth and carbon and nutrient cycling in grassland ecosystem. However, we are still unclear to what a content these soil water retention parameters may regulate plant production and soil organic matter content in natural ecosystems in semi-arid grassland environment. We investigated plant species richness (SR) and aboveground biomass (AGB), soil organic carbon (SOC) and nitrogen (SON) contents, and soil water retention and other physical parameters in the steppe ecosystems along a climate transect, which spans the desert steppe, typical steppe and meadow steppe zones in Northern China. We analyzed the relative importance of climate factors, soil water retention parameters and other soil physical parameters in regulating plant AGB and SOC and SON contents using partial correlation analysis, and quantified the contribution of each separate and combined climate and soil physical parameters in predicting plant AGB and SOC and SON contents using geographic detectors method. We found that (i) plant SR and AGB, soil SAT and DUL, and SOC and SON are all significantly inter-correlated, and increased with the increase of climate humidity. (ii) Climate, but not soil physical parameters, was the most preponderant factor in predicting plant SR and AGB. (iii) Soil physical parameters were more important than climate parameters in regulating soil organic matter content; soil DUL was the best predictor to SOC content. Our results suggest that soil water retention parameters especially soil DUL, may play a greater role than climate factors in modulating SOC content, though climate is more predominant in regulating plant SR and production. Our results imply that we need to be cautious in estimating SOC content based on climate factors and vegetation biomass or remotely-sensed vegetation indices, and that soil physical properties need also to be incorporated along with vegetation parameters for improving SOC estimation at a regional scale.
... Compared to other SOC remote sensing prediction models, the established models accounted for the impact of land-use types on SOC sensitive bands, and thereby improved the model precision. Several studies have confirmed that the use of ancillary data can improve the accuracy of soil prediction models [74,75]. ...
Coal mining, an important human activity, disturbs soil organic carbon (SOC) accumulation and decomposition, eventually affecting terrestrial carbon cycling and the sustainability of human society. However, changes of SOC content and their relation with influential factors in coal mining areas remained unclear. In the study, predictive models of SOC content were developed based on field sampling and Landsat images for different land-use types (grassland, forest, farmland, and bare land) of the largest coal mining area in China (i.e., Shendong). The established models were employed to estimate SOC content across the Shendong mining area during 1990–2020, followed by an investigation into the impacts of climate change and human disturbance on SOC content by a Geo-detector. Results showed that the models produced satisfactory results (R2 > 0.69, p < 0.05), demonstrating that SOC content over a large coal mining area can be effectively assessed using remote sensing techniques. Results revealed that average SOC content in the study area rose from 5.67 gC·kg−1 in 1990 to 9.23 gC·kg−1 in 2010 and then declined to 5.31 gC·Kg−1 in 2020. This could be attributed to the interaction between the disturbance of soil caused by coal mining and the improvement of eco-environment by land reclamation. Spatially, the SOC content of farmland was the highest, followed by grassland, and that of bare land was the lowest. SOC accumulation was inhibited by coal mining activities, with the effect of high-intensity mining being lower than that of moderate- and low-intensity mining activities. Land use was found to be the strongest individual influencing factor for SOC content changes, while the interaction between vegetation coverage and precipitation exerted the most significant influence on the variability of SOC content. Furthermore, the influence of mining intensity combined with precipitation was 10 times higher than that of mining intensity alone.
... Few studies use satellite sensors for SOC estimation [59]. However, the soil and the above-ground environment can be closely related allowing us to understand the biological, chemical and physical processes that govern the soil functions [60]. ...
Background
Soil organic carbon (SOC) affects essential biological, biochemical, and physical soil functions such as nutrient cycling, water retention, water distribution, and soil structure stability. The Andean páramo known as such a high carbon and water storage capacity ecosystem is a complex, heterogeneous and remote ecosystem complicating field studies to collect SOC data. Here, we propose a multi-predictor remote quantification of SOC using Random Forest Regression to map SOC stock in the herbaceous páramo of the Chimborazo province, Ecuador.
Results
Spectral indices derived from the Landsat-8 (L8) sensors, OLI and TIRS, topographic, geological, soil taxonomy and climate variables were used in combination with 500 in situ SOC sampling data for training and calibrating a suitable predictive SOC model. The final predictive model selected uses nine predictors with a RMSE of 1.72% and a R² of 0.82 for SOC expressed in weight %, a RMSE of 25.8 Mg/ha and a R² of 0.77 for the model in units of Mg/ha. Satellite-derived indices such as VARIG, SLP, NDVI, NDWI, SAVI, EVI2, WDRVI, NDSI, NDMI, NBR and NBR2 were not found to be strong SOC predictors. Relevant predictors instead were in order of importance: geological unit, soil taxonomy, precipitation, elevation, orientation, slope length and steepness (LS Factor), Bare Soil Index (BI), average annual temperature and TOA Brightness Temperature.
Conclusions
Variables such as the BI index derived from satellite images and the LS factor from the DEM increase the SOC mapping accuracy. The mapping results show that over 57% of the study area contains high concentrations of SOC, between 150 and 205 Mg/ha, positioning the herbaceous páramo as an ecosystem of global importance. The results obtained with this study can be used to extent the SOC mapping in the whole herbaceous ecosystem of Ecuador offering an efficient and accurate methodology without the need for intensive in situ sampling.
... The successful application of satellite RS images in predicting agricultural SOC has been proved in much research at the regional, national and global scale (Croft et al., 2012;Dvornikov et al., 2021;Hamzehpour et al., 2019;Mirzaee et al., 2016;Paul et al., 2020;Zhou et al., 2020a The relative importance of prediction features is presented in Fig. 8. Only 24 variables (10 features derived from S-2 and 14 features derived S-1) out of 48 variables were shown the high relative importance in the agricultural SOC. ...
Monitoring agricultural soil organic carbon (SOC) has played an essential role in sustainable agricultural management. Precise and robust prediction of SOC greatly contributes to carbon neutrality in the agricultural industry. To create more knowledge regarding the ability of remote sensing to monitor carbon soil, this research devises a state-of-the-art low cost machine learning model for quantifying agricultural soil carbon using active and ensemble-based decision tree learning combined with multi-sensor data fusion at a national and world scale. This work explores the use of Sentinel-1 (S1) C-band dual polarimetric synthetic aperture radar (SAR), Sentinel-2 (S2) multispectral data, and an innovative machine learning (ML) approach using an integration of active learning for land-use mapping and advanced Extreme Gradient Boosting (XGBoost) for robustness of the SOC estimates. The collected soil samples from a field survey in Western Australia were used for the model validation and indicators including the coefficient of determination (R2) and root - mean – square - error (RMSE) were applied to evaluate the model’s performance. A numerous features computed from optical and SAR data fusion were employed to build and test the proposed model performance. The effectiveness of the proposed machine learning model was assessed by comparing with the two well-known algorithms such as Random Forests (RF) and Support Vector Machine (SVM) to predict agricultural SOC. Results suggest that a combination of S1 and S2 sensors could effectively estimate SOC in farming areas by using ML techniques. Satisfactory accuracy of the proposed XGBoost with optimal features was achieved the highest performance (R2 = 0.870; RMSE= 1.818 tonC/ha) which outperformed RF and SVM. Thus, multi-sensor data fusion combined with the XGBoost lead to the best prediction results for agricultural SOC at 10 m spatial resolution. The innovative method described here could contribute significantly to various agricultural SOC retrieval studies globally. What makes this possible is its high prediction performance at 10 m spatial resolution and cost-effectiveness with reliable free-of-charge optical and SAR data acquired from Copernicus Open Access Hub.
... The authors [8] argue that SOC dynamics affect soil quality, agricultural productivity and CO2 concentration in the atmosphere. Despite the need for spatial estimates of SOC content over time, reliable estimates using traditional field research methods are limited by data availability because measurements are often made at discrete points, large sample spacing, or confined spaces. ...
The article solves the urgent problem of developing a system for remote monitoring of the soil conditions of grape agrocenoses in the southern regions of Russia. It is shown that such monitoring is required to be carried out constantly because of the high anthropogenic load and critical decrease in soil fertility. Experimental studies showing the possibility of satellite remote monitoring of the soil conditions have been carried out, while the accuracy of determining the humus content in the soil when shooting with a camera in a laboratory and using satellite images obtained with multispectral equipment is no worse than 10-20 percent compared to chemical laboratory methods. For the experiment, 22 experimental plots, the condition of the soil on which was determined in the laboratory according to Tyurin and by shooting and processing images of a high-resolution digital camera, were laid. Synchronous data from the Russian satellite Kanopus-V were studied for the same areas. Analytical dependences of the humus content in the soil on the brightness level of the red spectral channel were obtained. The structure of the information system for remote satellite monitoring of the soil conditions in the region is proposed.
... This characteristic plays an important role in soil mapping. The shorter the revisit time, the greater is the number of images that can be obtained and, therefore, the soil will be better monitored over time (Croft et al., 2012). Thus, the discussion about these three components of an image, has gain great importance to understand soil variability. ...
Soil maps at appropriate scales can aid decision-making in agriculture and the environment. In this sense, remote sensing products have shown their power to investigate soil properties, but the assessment of its spatial infor-mation can be hampered by the presence of other objects than soil. To overcome this issue, soil scientists have been studying spatial patterns from multi-temporal satellite images aiming at at improving soil property maps. In this work, we applied the cubist algorithm to predict topsoil properties (clay, sand, organic matter and iron contents, and soil color components) in south-eastern Brazil using multi-temporal (Landsat8-OLI, and Sentinel2- MSI) and single-date images (PlanetScope, Landsat8-OLI, and Sentinel2-MSI). We aimed to evaluate the influence of satellite’s spatial, spectral and temporal resolutions on soil mapping. Predictive models were constructed with 120 soil samples and using four (vis-NIR) and six (vis-NIR-SWIR) spectral bands as predictors in a 10-fold cross- validation procedure. The multi-temporal image obtained from the Sentinel2-MSI satellite (with 10 m pixel size and six spectral bands), showed the best model performances in cross-validation (R2 between 0.48 and 0.78). The PlanetScope image, which has only four bands in the vis-NIR region and spatial resolution of 3 m did not improve model performances, although the R2 values were higher for soil color components (R2 >0.5). Satellite images in different spatial, spectral and temporal resolution provides slightly different soil property maps which may promote different strategies regarding soil classification and management. However, satellite images should be used with caution, as they provide only surficial information about the soil variability and confirmation with field surveys is required.
... This reduces errors in quantifying soil carbon and other key properties that are often caused by spatial heterogeneity of soils. Infrared data can be integrated with geostatistic data (Cobo et al., 2010), remote sensing data and topographic information for digital soil mapping at the landscape level (Croft et al., 2012). Rossel et al. (2014), for instance, used infrared data to develop a soil carbon map of Australia. ...
... Soil physic and chemical properties, land control measures, tree species planted, vegetation types and environmental factors determine the change degree of SOC stocks and the timing of the switch between increase and decrease of stocks (Arai and Tokuchi, 2010;Don et al., 2011). In addition, it is also affected by the following factors, such as interactions between geochemistry and climate (i.e., precipitation and temperature) (Doetterl et al., 2015), previous land use type (England et al., 2016;van Straaten et al., 2015v), erosion processes (Croft et al., 2012), and whether the study design is based on a chrono sequence or time series (Fujisaki et al., 2015). However, different criteria of the research blocks, such as study design and scale, stock calculation method, and land management measures, which caused the considerable uncertainties in the SOC stock dynamics (Grinand et al., 2017). ...
In the context of global climate change, the preservation of soil productivity and the estimation of carbon budgets and cycles, the quantification of changes in carbon has important significance. In this study, we investigated the dynamics of soil aggregate associated organic carbon (OC) following temperate natural forest development in China. The objectives of this study were to examine the variation of soil aggregate associated OC decomposition rates, quantify the changes in the proportion of new and old soil aggregate OC, and explore the effects of controlling factors on SOC stocks, rate of total SOC increase and decomposition rate constants. The results showed that soil aggregate associated OC sequestration increased in 0−10 cm soil depth, while decreased in 10−30 cm soil depth. However, rate of aggregate associated OC increase, decomposition rate constants, and proportion of new OC increased at the early stage and then decreased along with the natural vegetation restoration. In addition, land use change had an important effect on soil aggregate associated OC dynamics, and soil particles, BD, MWD, C: N, plant diversity also played an important role. Moreover, SOC stocks had a negative relationship with clay and silt, while had a positive relationship with MWD and sandy soils. decomposition rate constants had a negative relationship with plant diversity, silt, and sand, while had a positive relationship with C: N and MWD. The proportions of new SOC had significant positive relationships with C: N, and it had a negative relationship with clay and silt. Therefore, it is necessary to clarify the formation mechanism of soil particles and aggregates, improve plant biodiversity, regulate the soil C: N ratio, and improve soil particle structure to increase soil carbon sequestration.
... Evaluating spatiotemporal variations of human activity intensity and ecological influence can produce accurate and practical results for reference; however, such an evaluation requires large amounts of regional data, which are costly due to the extensive field work required (De Gruijter et al., 2006;Chi et al., 2018c). Remote sensing provides an economical and convenient means of spatiotemporal evaluation (Croft et al., 2012;Suziedelyte Visockiene et al., 2019). Remote sensing data contains abundant ecological information, and various ecological indices and land cover data can be obtained (Liu et al., 2010a;Chi et al., 2018c). ...
Human activities have widely spread over nearly every corner of the world and been remarkably influencing the natural ecosystem since the 20th century. Identifying and quantifying the negative and positive influences of human activities are important for providing a solid basis for reasonable exploitation and effective conservation. This study focused on the negative and positive influences of human activities on five “macro to micro” aspects of an estuarine ecosystem, including island geomorphology, landscape pattern, plant community, physical quality, and chemical environment. An evaluation model was established using spatiotemporal ecological information from remote sensing, and three new indices, namely, human damage index (HDI), human regulation index (HRI), and human net influence index (HNII), were established to quantify the negative, positive, and net influences of human activities, respectively. Chongming Island in the Yangtze River Estuary of China was used as the study area, and four scenes of remote sensing images in 1988, 1995, 2007, and 2017 served as the data source. Results indicated that HDI initially increased and then decreased, HRI showed generally increasing characteristics, and HNII initially decreased and then increased in the entire study area from 1988 to 2017. Although the net influence was negative, ecological conservation and management since the 21 st century have clearly increased the HNII. Wetland vegetation, mudflat, and woodland had positive HNII; farmland, water area, and pond had HNII close to zero; and building, traffic, and industrial lands possessed negative HNII. The model was proven to greatly contribute to judging the ecological efficiencies of different types of land uses and optimizing the spatial configuration of human activities in estuarine areas.
... Most of them focus on the discussion on soil physical properties [18,19] and the spatial variation of soil salinity [20] from the perspective of natural science. As the research further develops, the geostatistical study of the spatial variability of soil nutrients is increasing [9,21,22], for example, the study of SOM [23,24], total N, NO -3-N, total P, available P, available K [25][26][27], nugget effect and degree of correlation [28][29][30], variable coefficient, etc. [31,32]. However, the extant study may be improved in the following aspects. ...
The study on soil organic matter (SOM) is of great importance to regional cultivated land use and protection. Based on data collected via continuous and high-density soil samples (0–20 cm) and socio-economic data collected from household survey and local bureau of statistics, this study employs geostatistics and economic statistical methods to investigate the spatial-temporal variation of SOM contents during 1980–2010 in the urban fringe of Sujiatun district in Shenyang City, China. We find that: (1) as to temporal variation, SOM contents in the study sites decreased from 30.88 g/kg in 1980 to 22.63 g/kg in 2000. It further declined to 20.07 g/kg in 2010; (2) in terms of spatial variation, the closer to city center, the more decline of SOM contents. Contrarily, SOM contents could even rise in outer suburb area; and (3) SOM content variation may be closely related to human factors such as farmers’ land use target and behaviour including inputs of chemical and organic fertilizers, types of crops and etc. These findings are conductive to grasp the overall trend of SOM variation and the influence of farmers’ land use behaviour on it. Furthermore, they could provide support for policymakers to agricultural planning and land use monitoring, which consequently aids the improvement of soil quality and food production in the urban fringe areas.
... Spectral absorption features are caused by vibrational stretching and bending of structural molecule groups and electronic excitation (Ben-Dor et al., 1999;Dalal and Henry, 1986). Molecule vibrations from hydroxyl, carboxyl, and amine functional groups produce absorption features related to soil organic matter in the mid-infrared (MIR) region of the spectra (Croft et al., 2012). In comparison, Vis-NIR spectra show only broad and unclear adsorption features related to overtone vibrations from the MIR, but instruments are less cost-intensive and available for field monitoring as well Viscarra Rossel et al., 2006a). ...
Soil organic carbon (SOC) plays a major role concerning chemical, physical, and biological soil properties and functions. To get a better understanding of how soil management affects the SOC content, the precise monitoring of SOC on long-term field experiments (LTFEs) is needed. Visible and near-infrared (Vis–NIR) reflectance spectrometry provides an inexpensive and fast opportunity to complement conventional SOC analysis and has often been used to predict SOC. For this study, 100 soil samples were collected at an LTFE in central Germany by two different sampling designs. SOC values ranged between 1.5% and 2.9 %. Regression models were built using partial least square regression (PLSR). In order to build robust models, a nested repeated 5-fold group cross-validation (CV) approach was used, which comprised model tuning and evaluation. Various aspects that influence the obtained error measure were analysed and discussed. Four pre-processing methods were compared in order to extract information regarding SOC from the spectra. Finally, the best model performance which did not consider error propagation corresponded to a mean RMSEMV of 0.12% SOC (R2 D 0:86). This model performance was impaired by 1RMSEMV D 0:04% SOC while considering input data uncertainties
(1R2 D 0:09), and by 1RMSEMV D 0:12% SOC (1R2 D 0:17) considering an inappropriate pre-processing. The effect of the sampling design amounted to a 1RMSEMV of 0.02% SOC (1R2 D 0:05). Overall, we emphasize the necessity of transparent and precise documentation of the measurement protocol, the model building, and validation procedure in order to assess model performance in a comprehensive way and allow for a comparison between publications. The consideration of uncertainty propagation is essential when applying Vis–NIR spectrometry for soil monitoring.
... It can measure soil parameters from a large number of samples in real time. NIRS is applicable to the development of precision agriculture and has attracted the attention of agriculture researchers [13][14][15]. SOM spectrometry is based on the spectral characteristics of soils and shows the reflectivity of organic matter at specific wave bands. ...
Soil organic matter (SOM) is a major indicator of soil fertility and nutrients. In this study, a soil organic matter measuring method based on an artificial olfactory system (AOS) was designed. An array composed of 10 identical gas sensors controlled at different temperatures was used to collect soil gases. From the response curve of each sensor, four features were extracted (maximum value, mean differential coefficient value, response area value, and the transient value at the 20th second). Then, soil organic matter regression prediction models were built based on back-propagation neural network (BPNN), support vector regression (SVR), and partial least squares regression (PLSR). The prediction performance of each model was evaluated using the coefficient of determination (R2), root-mean-square error (RMSE), and the ratio of performance to deviation (RPD). It was found that the R2 values between prediction (from BPNN, SVR, and PLSR) and observation were 0.880, 0.895, and 0.808. RMSEs were 14.916, 14.094, and 18.890, and RPDs were 2.837, 3.003, and 2.240, respectively. SVR had higher prediction ability than BPNN and PLSR and can be used to accurately predict organic matter contents. Thus, our findings offer brand new methods for predicting SOM.
... Because the physical foundation for the spectral response of STN in remote sensing data is still lacking, we collected several soil attributes that were closely related to STN and whose spectral responses were already clear, such as the soil organic carbon (Croft et al., 2012;Lu et al., 2018) and soil moisture (Sadeghi et al., 2017;Wang et al., 2007). Several spectral indices, such as the normalized difference vegetation index (NDVI), the difference vegetation index (DVI), the enhanced vegetation index (EVI), the normalized difference water index (NDWI), the modified soil adjusted vegetation index (MSAVI), and the transformed vegetation index (TVI), are generally employed in the estimation of soil organic carbon and soil moisture. ...
... As a result, it is now possible to accurately monitor the spatiotemporal distribution and variation of SOC at multiple temporal and spatial scales within the region. These advantages mean that remote sensing process modelling is now the main research approach used for simulating carbon cycles in the terrestrial ecosystem (Croft et al., 2012;Smith et al., 2012;Minansy et al., 2013;Liu, 2014;Chen, 2016). It is worth noting, however, that due to the limits of current satellite data collection revisit cycles and the impact of cloud coverage, it remains challenging to estimate the spatial distribution of SOC at high-resolution. ...
Accurate quantitative estimates of Soil Organic Carbon Density (SOCD) can effectively represent regional carbon cycle processes and regulation mechanisms, and can serve as reference data when making land management decisions. Limited research, however, has been carried out in arid or desert zones covered with sparse vegetation , despite the fact that these cover wide areas of the earth and play a significant role in global carbon cycles. In this study, the Otindag Sandy Land and its surroundings (OSLAIS) in the Inner Mongolia Autonomous Region of China was selected as the study area. The study introduces a useful technique for making high spatial coverage SOCD estimates for drylands by utilizing GF-1 WFV optical satellite images and a time series of MODIS satellite remote sensing datasets, and using these to optimize parameters for simulation models in conjunction with other technical procedures that are described. The results showed that the resulting model's accuracy was 77.87%, R 2 = 0.8627, and so the SOCD estimates modelled by soil basal respiration (SBR) could be used for SOCD estimation and for analyzing the spatial distribution patterns across the OSLAIS. The average SOCD was 1.22 kgC/m 2 for the whole of the OSLAIS, and it had a heterogenous distribution pattern. The SOCD was closely related to the way the land was used in each area, and the average SOCD for the main land use types were: forest land = 2.88 kgC/m 2 , farmland = 1.63 kgC/m 2 , shrub land = 1.41 kgC/m 2 , and grassland = 1.08 kgC/m 2. In conclusion, we believe that the proposed method, based on high-resolution GF-1 WFV data and optimized estimation models constructed by integrating climate and vegetation characteristic data, can effectively describe the spatial distribution patterns of SOC and SOCD in the OSLAIS area, in depth and in detail, especially for the areas where the SOCD values are high. We expect this research to provide useful technical support and scientific reference data for land management and for land degradation/desertification assessments, for the study area monitored, as well as across the whole dryland area of China.
... These techniques do not need chemical reagents or elaborate preparation, and they are faster and cheaper than conventional analyses so that the analysis of many soil samples is feasible (Viscarra Rossel et al., 2011). There are some reviews on the use of reflectance spectroscopy for soil total organic and inorganic C analyses (Reeves III, 2009;Stenberg et al., 2010;Bellon-Maurel and McBratney, 2011;Croft et al., 2012;Soriano-Disla et al., 2014;Viscarra Rossel et al., 2016). ...
Soil organic carbon (C) is an important indicator of agricultural and environmental quality. It improves soil fertility and helps to mitigate greenhouse gas emissions. Soil spectroscopy with either vis–NIR (350–2500 nm) or mid-IR (4000–400 cm⁻¹) spectra have been used successfully to predict organic C concentrations in soil. However, research to improve predictions of soil organic C by simply combining vis–NIR and mid-IR spectra to model them together has been unsuccessful. Here we use the Outer Product Analysis (OPA) to fuse vis–NIR and mid-IR spectra by bringing them into a common spectral domain. Using the fused data, we derived models to predict soil organic C and compared its predictions to those derived with vis–NIR and mid-IR models separately. We analyzed 1259 tropical soil samples from surface and subsurface layers across agricultural areas in Central Brazil. Soil organic C contents were determined by a modified Walkley-Black method, and vis–NIR and mid-IR reflectance spectra were obtained with a FieldSpec Pro and a Nicolet 6700 Fourier Transformed Infrared (FT-IR), respectively. Reflectances were log-transformed into absorbances. The mean content of soil organic C was 9.14 g kg⁻¹ (SD = 5.64 g kg⁻¹). The OPA algorithm was used to emphasize co-evolutions of each spectral domain into the same one by multiplying the absorbances from both sets of spectra to produce a matrix with all possible products between them. Support Vector Machine with linear kernel function was used for the spectroscopic modeling. Predictions of soil organic C using vis–NIR, mid-IR, and fused spectra were statistically compared by the Tukey's test using the coefficient of determination (R²), root mean squared error (RMSE), and ratio of performance to interquartile distance (RPIQ). Absorbances in vis–NIR and mid-IR were emphasized in the common spectral domain presenting stronger correlations with soil organic C than individual ranges. Soil organic C predictions with the OPA fused spectra were significantly better (R² = 0.81, RMSE = 2.42 g kg⁻¹, and RPIQ = 2.87) than those with vis–NIR (R² = 0.69, RMSE = 3.38 g kg⁻¹, and RPIQ = 2.08) or mid-IR spectra (R² = 0.77, RMSE = 2.90 g kg⁻¹, and RPIQ = 2.43). Fusing vis–NIR and mid-IR spectra by OPA improves predictions of soil organic C.
... al., 2017). Meanwhile, the possibility of extending the results obtained from ground derived spectral data to hyperspectral data, has developed the applicability of this technique for mapping the heavy metal polluted areas on HyMAP images during the last decades (Choe et al., 2008;Zhang et al., 2010;Mulder et al., 2011;Liu et al., 2011;Chen et al., 2012;Croft et al., 2012;Omran, 2016). Considering these advantages, VNIR -SWIR spectroscopy is commonly used as a preliminary step in deciding upon sampling and analysis strategies in most geochemical mapping projects (Shi et al., 2007). ...
This study considered the possibility of using visible and near infrared (VNIR) spectral
absorption feature parameters (SAFPs) in predicting the concentration and mapping the distribution of heavy metals in sediments of the Takab area. In total, 60 sediment samples were collected along main streams draining from the mining districts and tailing sites, in order to measure the concentration of As, Co, V, Cu, Cr, Ni, Hg, Ti, Pb and Zn and the reflectance spectra (350–2500 nm). The quantitative relationship between SAFPs (Depth500nm, R610/500nm, R1344/778nm, Area500nm, Depth2200nm, Area2200nm,
Asym2200nm) and geochemical data were assessed using stepwise multiple linear regression (SMLR) and enter multiple linear regression (EMLR) methods. The results showed a strong negative correlation between Ni and Cr with Area2200nm, a significant positive correlation between As and Asym2200nm, Ni and Co with Depth2200nm, as well as Co, V and total values with Depth500nm. The EMLR method eventuated in a significant prediction result for Ni, Cr, Co and As concentrations based on spectral parameters, whereas the prediction for Zn, V and total value was relatively weak. The spatial distribution pattern of geochemical data showed that mining activities, along with the natural weathering of base metal occurrences and rock units, has caused high concentrations of heavy metals in sediments of the Sarough River tributaries.
... Soil organic carbon (SOC) accumulation is considered to be closely associated with soil quality, agricultural productivity and atmospheric carbon dioxide concentration (Croft et al., 2012). Therefore, SOC sequestration has a profound significance to ecosystem stability and agricultural sustainability (Lal, 2004;Ratnayake et al., 2017). ...
Currently, humanity is indeed in turbulent times. The current world affairs are no different from those during World War I (28th July 1914–11th November 1918) or during the time of World War II (1st September 1939–2nd September 1945). Scientists, especially, chemists, have a vital role to ensure that lifesaving materials are available at the disposal of common people. Renowned German chemist Raphael von Ostrejko has a few patents based on activated carbon from biomass during 1900–1920 and his inventions based on those patents, namely activated carbon based masks, saved many people from perishing due to the harmful effects of poisonous gases used during World War I. Carbon materials from biomass, in particular, are equally significant during these testing times of war, energy and food shortages and climate change. Biomass constitutes a rich source of carbon materials with diverse properties and applications impacting almost every sphere of human activity, ranging from agriculture, health, energy, environment, materials, safety, security, defence and many more. All the three vital components of biomass, namely cellulose, hemicellulose and lignin, offer unique structure and morphological features to the resulting carbon materials. Owing to the sustainability, diversity in properties and applications of these carbon materials produced from biomass, it is conceived that a Special Issue needed to be launched in the journal “C-journal of carbon research” published by Multidisciplinary Digital Publishing Institute (MDPI), wherein the latest scientific advancements throughout the globe could be assembled and made available freely, at a click, to the scientific, academic and industrial fraternity for their growth and development. Thus, this Special Issue was
launched with the title “Biomass—A renewable resource for carbon materials”.
TOPRAK KALİTESİ VE DEĞERLENDİRİLMESİ
Comprehensively revealing the spatial pattern of soil condition is vital for exploring the evolution mechanism of island ecosystem and providing reference for ecological maintenance. However, it is difficult to meet the demands of comprehensiveness and accuracy in mapping the soil condition in an archipelago that contains remote islands. Such an archipelago, namely, Shengsi Archipelago off the Yangtze River Estuary, China, was selected as the study area. Field survey was conducted on part of islands that have relatively high accessibility in the archipelago, and remote sensing data that cover the entire study area was adopted. An island soil index system, including soil content, storage, and quality indices, was proposed to represent the island soil integrated condition. Soil content indices are original soil measured data, and soil storage and quality indices are composite indices that are determined based on soil content indices. Then, a predictor system based on remote sensing and the 10-fold cross-validation were used to conduct the spatial simulations of soil indices. The results validated that the combination of soil measured data sourced by field survey could respond more sensitively to remote sensing data and integrate with it better than the original soil measured data, thereby increasing the simulation accuracies of soil storage and quality indices to higher levels than the corresponding soil content indices and achieving the spatial exhibition of soil integrated condition. The spatial pattern of soil indices in Shengsi Archipelago indicated that islands or areas with good vegetation condition but low soil salinity, land surface aridity, and human interference generally had good soil integrated condition. Then, suggestions to improve the island soil integrated condition were proposed from perspectives of different essential components of the island ecosystem. The study has provided a practical method for comprehensively mapping the soil condition in areas with low accessibility.
Soil organic carbon storage in agricultural soil constitutes a crucial potential for sustainable agricultural productivity and climate change mitigation. This paper aimed at assessing soil or-ganic carbon stock and its distribution in three particle size fractions across five cropping systems located in Kiti sub-watershed in Benin. Soil samples were collected using a grid sampling method on four soil depth layers: 0–10, 10–20, 20–30 and 30–40 cm in five cropping systems maize–cotton relay cropping (MCRC), yam–maize intercropping (YMI), teak plantation (TP), 5-year fallow (5YF) and above 10-year fallow (Ab10YF) from July to August 2017. Soil organic carbon stock (C stock) was estimated for the different soil layers and particle-size fractionation of soil organic matter was performed considering three fractions. The fractions coarse particulate organic matter (cPOM: 250–2000 µm), fine particulate organic matter (fPOM: 53–250 µm) and non-particulate organic matter (NOM: <53 µm) were separated from two soil depth layers: 0–10 and 10–20 cm. The results showed that fallow lands Ab10YF and 5YF exhibited the highest C stock, 22.20 and 17.74 Mg C·ha−1, while cultivated land under tillage MCRC depicted the lowest, C stock 11.48 Mg C·ha−1. The three organic carbon fractions showed a significant variation across the cropping systems with the NOM fraction holding the largest contribution to total soil organic carbon for all the cropping systems, ranging between 3.40 and 7.99 g/kg. The cPOM and fPOM were the most in-fluenced by cropping systems with the highest concentration observed in Ab10YF and 5YF. The findings provide insights for upscaling farm management practices towards sustainable agricultural systems with substantial potential for carbon sequestration and climate change mitigation.
Soil organic carbon (SOC) is one of the key soil components for cultivated soils. SOC is regularly monitored and mapped to improve the quality, health, and productivity of the soil. However, traditional SOC-level monitoring is expensive for land managers and farmers. Estimating SOC using satellite imagery provides an easy, efficient, and cost-effective way to monitor surface SOC levels. The objective of this study was to estimate the surface SOC distribution in selected soils of Major Land Resource Areas (MLRA), 102A (Rolling Till Plain, Brookings County, SD), and 103 (Central Iowa and Minnesota Till Prairies, Lac qui Parle County, MN), using satellite imagery with different resolutions (Landsat 8 and PlanetScope). The dominant soils in the study area are Haplustolls, Calciustolls, and Endoaquolls, which are formed in silty sediments, local silty alluvium, and till. Landsat 8 and PlanetScope spectral bands were used to develop SOC prediction models. Parametric and data-driven methods were employed to predict the SOC. Multiple linear regression and Linear Spatial Mixed Model (LSMM) were used on the Landsat 8 and PlanetScope data. In addition to the parametric models, Regression Trees and Random Forest were also employed on both satellite data. The results showed that reduced LSMM provided the lowest RMSE, which are 0.401 and 0.367 for Landsat 8 and PlanetScope, respectively. Furthermore, the random forest has the highest RPD and RPIQ for Landsat 8 (RPD of 2.67 and RPIQ of 2.49) and PlanetScope (RPD of 2.85 and RPIQ of 3.7). In all cases, models obtained from PlanetScope are better than those obtained from Landsat 8.
The Qinghai-Tibet Plateau (QTP) is a typical ecologically fragile area. Once the surface vegetation degenerates, it may not be restored. This requires the development of soil organic matter (SOM) monitoring method without destroying the surface, so as to ensure the sustainable development of plateau agriculture. This work investigated the environmental factors that are significantly related to SOM content in the river valley of the southern QTP. These environmental factors include soil hydrothermal factors (soil moisture content and soil temperature), topographic factors (elevation and slope) and vegetation factor (NDVI). The original band reflectivity (OR) of Landsat 8 OLI images and the band reflectivity after the first-order derivative (FDR) and the second-order derivative (SDR) processing were combined with the above environmental factors to estimate SOM content. The results showed that the accuracy of the model was improved obviously by adding environmental factors. The estimation effect of back propagation neural network (BPNN) model was better than that of geographically weighted regression (GWR) model, partial least squares regression (PLSR) model and multivariable linear regression (MLR) model. GWR model can also meet the estimation requirements, while PLSR and MLR models cannot achieve effectively the estimation of SOM content. FDR-BPNN model considering environmental factors was the best model for estimating SOM content, with R² being 0.947, RMSEC being 4.701 g·kg⁻¹ and MAEV being 5.485 g·kg⁻¹. Moreover, the model had the lowest uncertainty and the highest stability. This study will provide a good insight for the monitoring of SOM content in the future, and provide basic data support for the implementation of precision agriculture in the QTP.
The regular monitoring of soil physical, chemical, and biological properties is very essential, due to its role in soil ecosystem functions. A cost-effective alternative for soil monitoring corresponds to spectral sensing techniques. Soil spectral sensing techniques can support decision-making in agricultural systems at both time and spatial scales, maximizing food production while preserving an adequate soil condition. Due to the large number of ground, airborne, and orbital spectral sensors operating today, this technology has been increasingly assimilated by soil scientists. However, it is important to have an adequate comprehension about the technique principles and limitations. This chapter provides a wide perspective about the soil spectral sensing in the visible (vis: 350–700 nm), near-infrared (NIR: 700–1000 nm), and shortwave infrared (SWIR: 1000–2500 nm), considering reflectance data at different acquisition levels. Here, it is discussed how soil constituents interact with EMR and the resulting soil spectral behaviors. We describe the predictive potential of vis-NIR-SWIR data for quantitative assessment of soil and which soil attributes have been reliably estimated and the most commonly used vis-NIR-SWIR equipment, as well as their advantages and limitations. Finally, we discuss the current application in soil science and future perspectives.
The need for adopting negative emission technologies (NETs) is greater now than ever before. The “4 per 1000” (4p1000) initiative, launched during the COP 21 in Paris in 2015, based on translating the science of soil carbon sequestration into action at the global scale, is one example of a broader set of negative emission technologies (NETs). While governments and non-governmental organizations are voluntarily implementing the aspirational 4p1000, its adoption and pledges by industry or the private sector is the driving force to accelerate the process of re-carbonization of soil and the terrestrial biosphere. The pledge of achieving carbon (C) neutrality by 2030 or 2050 by some prominent industries is encouraging, and it can serve as a role model for other industries to follow. Adoption of the 4p1000 concept by agro-industries (e.g., fertilizers, pesticides, farm machinery, irrigation, food processing) can promote and accelerate adoption of recommended management practices which create a positive/soil ecosystem carbon budget and enhance C sequestration. This review article describes examples of the emission-neutral concept already adopted by some industries, outlines strategies to promote 4p1000 and NETs, and discusses approaches to advance site-specific NETs. These techniques may include conservation agriculture, cover cropping, precision agriculture, etc., facilitated by artificial intelligence, remote sensing, and robots. Innovative methods of soil sampling as well as remote sensing and mapping of SOC stock are pertinent to promoting NETs and 4p1000. Payments to land managers and farmers for provisioning of ecosystem services through C sequestration in soil and the biosphere can advance the mission of 4p1000. Amongst strategies of encouraging industries to adopt NETs, implementation of a “Healthy Soil Act” at the state, national, and global level could be a step in the right direction. Market-based economic incentives to industry may be another option to promote adoption of NETs. Similar to U.S. EPA Clean Air Act of 1963 and the U.S. EPA Clean Water Act of 1972, there is a strong need to enact a Healthy Soil Act globally. An effective implementation of the trinity of Acts (air, water, and soil) at the national and global scale will accelerate the process of re-carbonization of the soil and the terrestrial biosphere while also strengthening the provisioning of essential ecosystem services.
The time difference between the field sampling and acquired spaceborne imagery has always been ignored in satellite-based soil analyses. In this study, the impacts of the time difference on soil nutrient predictions and the underlying mechanism for these predictions were investigated using a case study from the North China Plain. Soil total potassium (TK) was sampled in 2016 and was subsequently analyzed, with the soil TK content then predicted using the original reflectances from eight Landsat TM/OLI images that spanned the 1986–2016 period as inputs to multiple linear regression (MLR) and artificial neural network (ANN) models. The results reveal a temporal paradox, where earlier satellite imagery yields higher accuracy in the predictions than does the more recent imagery. Historical soil nutrient data sets were used to explain this temporal paradox, which indicated that both an internal and external factor influenced the soil TK predictions. The internal factor, the deposition pattern across the study region, was found to strongly control the spatial distribution of potassium and exhibited minimal changes over the past 30 years, which influenced the good soil TK predictions using the early satellite imagery. The external factor, the soil organic matter (SOM), which has a stronger impact on the spectral reflectances than the soil TK, indicated that the increase and regional uniformization of SOM contents caused by Chinese agricultural development from the 1980s to the early 2000s masked the true spectral response of soil TK and explained the decline in the prediction accuracy of soil TK with time. This study reveals a potential limitation in remote-sensing-based soil TK predictions, and indicates that early satellite imagery should be considered as a potentially important factor in future research in order to accurately assess soil nutrient.
Remote sensing provides possibilities to support digital soil mapping at large scales, but this technology over complex landscapes is still challenging due to the high spatial heterogeneity of soil properties. The objective of this work was to check the predictive capability of SOC and soil pH over different ecoregions. First, we collected a series of satellite images from 2017 and ancillary variables (e.g., climate and terrain) as spectral and environmental indicators, respectively. A total of 201 topsoil samples (0–20 cm) from the Northeastern Black Soil Region of China were also used. Then, the relationships between the two soil properties and indicators were analyzed. The indicators that showed significant correlation with SOC or soil pH were involved in the subsequent models. Finally, the SOC and soil pH prediction models were constructed by using a stepwise regression model. The results showed that the visible and shortwave infrared bands, along with the thermal infrared band, had strong correlations with the two soil properties. Environmental factors such as precipitation, temperature, and elevation all had an impact on the distribution of SOC and soil pH. The SOC and soil pH prediction models had R2 values of 0.66 and 0.73, root mean square errors of 0.21 and 0.58, and residual prediction deviations of 1.46 and 1.53, respectively, indicating reliable precision and average prediction abilities. This approach to estimating SOC and soil pH can contribute to improving our understanding and the modeling of soil processes at a large scale and in areas with relatively complex landscapes.
Soil surface roughness, denoted by the root mean square height (RMSH), and soil moisture (SM) are critical factors that affect the accuracy of quantitative remote sensing research due to their combined influence on spectral reflectance (SR). In regards to this issue, three SM levels and four RMSH levels were artificially designed in this study; a total of 12 plots was used, each plot had a size of 3 m × 3 m. Eight spectral observations were conducted from 14 to 30 October 2017 to investigate the correlation between RMSH, SM, and SR. On this basis, 6 commonly used bands of optical satellite sensors were selected in this study, which are red (675 nm), green (555 nm), blue (485 nm), near infrared (845 nm), shortwave infrared 1 (1600 nm), and shortwave infrared 2 (2200 nm). A negative correlation was found between SR and RMSH, and between SR and SM. The bands with higher coefficient of determination R² values were selected for stepwise multiple nonlinear regression analysis. Four characterized bands (i.e., blue, green, near infrared, and shortwave infrared 2) were chosen as the independent variables to estimate SM with R² and root mean square error (RMSE) values equal to 0.62 and 2.6%, respectively. Similarly, the four bands (green, red, near infrared, and shortwave infrared 1) were used to estimate RMSH with R² and RMSE values equal to 0.48 and 0.69 cm, respectively. These results indicate that the method used is not only suitable for estimating SM but can also be extended to the prediction of RMSH. Finally, the evaluation approach presented in this paper highly restores the real situation of the natural farmland surface on the one hand, and obtains high precision values of SM and RMSH on the other. The method can be further applied to the prediction of farmland SM and RMSH based on satellite and unmanned aerial vehicle (UAV) optical imagery.
Airborne hyperspectral images provide high spatial and spectral resolution along with flexible temporal
resolution that are ideally suited for precision agricultural applications. In this study, we have explored the potential of aerial visible/infrared (VIR) hyperspectral imagery for characterizing soil fertility factors in midwestern agricultural fields. Two fields (SW and NW) in Illinois and two fields (GV and FO) in Missouri were considered in this study. Field data included hyperspectral VIR images and soil fertility parameters including pH, organic matter (OM), Ca, Mg, P, K, and soil electrical conductivity. The VIR images were geo-registered and calibrated into apparent reflectance values. The FO field had the highest average reflectance, followed by SW, GV, and NW. The Illinois fields (SW and NW) were high in soil minerals, OM,
and soil electrical conductivity. The measured soil fertility characteristics were modeled on first derivatives of the reflectance data using partial least square regression (PLSR). The PLSR model on derivative spectra was able to explain 66% of the overall variability in soil fertility variables considered in this study, with a predicted residual sum of square (PRESS) of 0.66. The model explained a higher degree of variability in some of the response variables, such as Ca (82%), Mg (72%), Veris shallow (86%), Veris deep (67%), and OM (66%), compared to factors such as pH (48%) and EM (50%). Analysis of the parameter estimates for each response variable showed that some of the wavebands, such as 625, 652, 658, 661, 754 and
784 nm, explained a high degree of variability in the model, whereas a large number of wavelengths had negligible contribution. In conclusion, this study showed that soil fertility factors important for precision agriculture applications can be successfully modeled on hyperspectral VIR remote sensing data with partial least square regression models.
Models of soil reflectance under changing moisture conditions are needed to better quantify soil and vegetation properties from remote sensing. In this study, measurements of reflected shortwave radiation (400-2500 nm) were acquired in a laboratory setting for four different soils at various moisture contents. The observed changes in soil reflectance revealed a nonlinear dependence on moisture that was well described by an exponential model, and was similar for different soil types when moisture was expressed as degree of saturation. Reflectance saturated at much lower moisture contents in the visible and near-infrared (VNIR) spectral region than in the shortwave-infrared (SWIR) spectral region, suggesting that longer wavelengths are better suited for measuring volumetric moisture contents above ∼20%. To explore the potential of a general reflectance model based solely on dry reflectance and moisture, we employed a Monte Carlo analysis that accounted for observed variability in the measured spectra. Modeling results indicated that the SWIR region offers significant potential for relating moisture and reflectance on an operational basis, with uncertainties less than half as large as in the VNIR. The results of this study help to quantify the strong influence of moisture on spectral reflectance and absorption features, and should aid in the development of operational algorithms as well as more physically based models in the future.
Methods for rapid estimation of soil properties are needed for quantitative assessments of land management problems. We developed a scheme for development and use of soil spectral libraries for rapid nondestructive estimation of soil properties based on analysis of diffuse reflectance spectroscopy. A diverse library of over 1000 archived topsoils from eastern and southern Africa was used to test the approach. Air-dried soils were scanned using a portable spectrometer (0.35-2.5 μm) with an artificial light source. Soil properties were calibrated to soil reflectance using multivariate adaptive regression splines (MARS), and screening tests were developed for various soil fertility constraints using classification trees. A random sample of one-third of the soils was withheld for validation purposes. Validation r2 values for regressions were: exchangeable Ca, 0.88; effective cation-exchange capacity (ECEC), 0.88; exchangeable Mg, 0.81; organic C concentration, 0.80; clay content, 0.80; sand content, 0.76; and soil pH, 0.70. Validation likelihood ratios for diagnostic screening tests were: ECEC <4.0 cmolc kg-1, 10.8; pH <5.5, 5.6; potential N mineralization >4.1 mg kg-1 d-1, 2.9; extractable P <7 mg kg-1, 2.9; exchangeable K >0.2 cmolc kg-1, 2.6. We show the response of prediction accuracy to sample size and demonstrate how the predictive value of spectral libraries can be iteratively increased through detection of spectral outliers among new samples. The spectral library approach opens up new possibilities for modeling, assessment and management of risk in soil evaluations in agricultural, environmental, and engineering applications. Further research should test the use of soil reflectance in pedotransfer functions for prediction of soil functional attributes.
The soil is important in sequestering atmospheric CO2 and in emitting trace gases (e.g. CO2, CH4 and N2O) that are radiatively active and enhance the ‘greenhouse’ effect. Land use changes and predicted global warming, through their effects on net primary productivity, the plant community and soil conditions, may have important effects on the size of the organic matter pool in the soil and directly affect the atmospheric concentration of these trace gases.
A discrepancy of approximately 350 × 1015 g (or Pg) of C in two recent estimates of soil carbon reserves worldwide is evaluated using the geo-referenced database developed for the World Inventory of Soil Emission Potentials (WISE) project. This database holds 4353 soil profiles distributed globally which are considered to represent the soil units shown on a 1/2° latitude by 1/2° longitude version of the corrected and digitized 1:5 M FAO–UNESCO Soil Map of the World.
Total soil carbon pools for the entire land area of the world, excluding carbon held in the litter layer and charcoal, amounts to 2157–2293 Pg of C in the upper 100 cm. Soil organic carbon is estimated to be 684–724 Pg of C in the upper 30 cm, 1462–1548 Pg of C in the upper 100 cm, and 2376–2456 Pg of C in the upper 200 cm. Although deforestation, changes in land use and predicted climate change can alter the amount of organic carbon held in the superficial soil layers rapidly, this is less so for the soil carbonate carbon. An estimated 695–748 Pg of carbonate-C is held in the upper 100 cm of the world's soils. Mean C: N ratios of soil organic matter range from 9.9 for arid Yermosols to 25.8 for Histosols. Global amounts of soil nitrogen are estimated to be 133–140 Pg of N for the upper 100 cm. Possible changes in soil organic carbon and nitrogen dynamics caused by increased concentrations of atmospheric CO2 and the predicted associated rise in temperature are discussed.
This review article gives an overview of how satellite observations are used to feed or tune crop models and improve their capability to predict crop yields in a region. Relations between crop characteristics which correspond to models state variables and satellite observations are briefly analysed, together with the various types of crop models commonly used. Various strategies for introducing short wavelength radiometric information into specific crop models are described, from direct update of model state variables to optimization of model parameter values, and some of them are exemplified. Methods to unmix crop-specific information from mixed pixels in coarse resolution-high frequency imagery are analysed. The conditions of use of the various methods and types of information are discussed.
Predictive models for soil organic carbon using reflectance were generated for soil samples collected from five Saskatchewan fields on a field-specific basis. Models were developed using calibration data sets to identify optimal regressor wavebands for multiple linear regressions and subsequently tested using validation data sets. Models were evaluated by comparing coefficient of determination and sum of squared error values, with a bias toward models with fewer regressors. The optimum model, including the number and wavelength of reflectance bands, for each field was identified. The four-regressor model, using reflectance in the 1500, 1720, 1740, and 1790-nm wavebands, was selected for the Hepburn field. Four-regressor models were also selected for Outlook and St. Louis fields, using reflectance at 480, 760, 1650, and 1700 nm and 530, 560, 2320, and 2360 nm, respectively. Reflectance data in the 1370, 2390, and 2450-nm wavebands were used in the three-regressor model for the Swift Current field, while the two-regressor model, comprised of reflectance data in the 1570 and 1600-nm wavebands, was chosen for the Watrous field. Coefficients of determination for these models were 0.85, 0.85, 0.89, 0.73, and 0.86 for the Hepburn, Outlook, St. Louis, Swift Current and Watrous fields, respectively.
There is a global concern about the increase in atmospheric concentrations of greenhouse gases. One method being dis-cussed to encourage greenhouse gas mitigation efforts is based on a trading system whereby carbon emitters can buy effective mitigation efforts from farmers implementing con-servation tillage practices. These practices sequester car-bon from the atmosphere, and such a trading system would require a low-cost and accurate method of verification. Re-mote sensing technology can offer such a verification tech-nique. This paper is focused on the use of standard image processing procedures applied to a multispectral Ikonos image, to determine whether it is possible to validate that farmers have complied with agreements to implement con-servation tillage practices. A principal component analysis (PCA) was performed in order to isolate image variance in cropped fields. Analyses of variance (ANOVA) statistical procedures were used to evaluate the capability of each Ikonos band and each principal component to discriminate between conventional and conservation tillage practices. A logistic regression model was implemented on the princi-pal component most effective in discriminating between conventional and conservation tillage, in order to produce a map of the probability of conventional tillage. The Ikonos imagery, in combination with ground-reference information, proved to be a useful tool for verification of conservation tillage practices.
For the determination of soil organic carbon (OC) concentrations, the availability of a fast, low-cost analysis method is required. The aim of the present study was to evaluate the possibilities of near infrared reflectance spectroscopy (NIRS) to build a spectral database and to develop calibrations for the prediction of organic carbon concentrations in grassland soils. NIRS spectra of 1626 soil samples from different grasslands (both agricultural and natural) were collected between 1100 and 2500 nm. NIRS calibrations were developed with modified partial least square regression and tested with independent validation samples. The best equations were obtained with the first derivative of the spectra without scatter corrections. For the global calibration, containing the samples of all origins, the standard errors of calibration (SEC) and of prediction (SEP) were respectively 3·70 g OC/kg dry soil (R2=0·89) and 3·95 g OC/kg dry soil (R2=0·88). The ratio of the standard deviation of the reference validation data to the SEP (RPD), indicating the performance of the calibration, was 2·9. Dividing the samples into groups according to their practice (agricultural or natural grassland), improved SEP by 5·8 and 7·7%, respectively. Dividing the samples into texture groups (clay, silt, sand) improved SEP for agricultural grassland by, on average, 7·4% and for natural grassland by 16·2%.
Farm managers are becoming increasingly aware of the spa- tial variability in crop production with the growing availabil- ity of yield monitors. Often this variability can be related to differences in soil properties (e.g., texture, organic matter, salinity levels, and nutrient status) within the field. To de- velop management approaches to address this variability, high spatial resolution soil property maps are often needed. Some soil properties have been related directly to a soil spectral response, or inferred based on remotely sensed mea- surements of crop canopies, including soil texture, nitrogen level, organic matter content, and salinity status. While many studies have obtained promising results, several inter- fering factors can limit approaches solely based on spectral response, including tillage conditions and crop residue. A number of different ground-based sensors have been used to rapidly assess soil properties "on the go" (e.g., sensor mounted on a tractor and data mapped with coincident po- sition information) and the data from these sensors compli- ment image-based data. On-the-go sensors have been devel- oped to rapidly map soil organic matter content, electrical conductivity, nitrate content, and compaction. Model and statistical methods show promise to integrate these ground- and image-based data sources to maximize the information from each source for soil property mapping.
The ability of agricultural lands to sequester carbon from the atmosphere and help mitigate global warming has the potential to add value to farmland through the development of carbon-credit trading. Crucial to the creation of a market-based carbon credit trading system is the monitoring and verification of agricultural practices that promote carbon storage. Using remotely sensed images for this purpose could prove more efficient and cost-effective than traditional land-based methods. Landsat Enhanced Thematic Mapper Plus (ETM+) imagery and logistic regression had >95% accuracy in verifying no-till fallow fields. Further research is needed to investigate the potential for this low-cost technology to assist in the monitoring and verification of practices that sequester carbon. Development of an accurate, low-cost, efficient means of monitoring and verifying carbon sequestering practices will further the development of cropland carbon credits, thus helping to mitigate global warming, and will add value to U.S. farmland.
In situ or on-the-go visible and near infrared (VisNIR) diffuse reflectance spectroscopy has been proposed as a rapid and inexpensive tool for intensively mapping soil texture and organic carbon (SOC). While lab-based VisNIR has been established as a viable technique for estimating various soil properties, few experiments have compared the predictive accuracy of on-the-go and lab-based VisNIR. In this study, eight north central Montana wheat fields were intensively interrogated using on-the-go and lab-based VisNIR. The on-the-go VisNIR system employed a spectrophotometer (350–2224nm, 8-nm spectral resolution) built into an agricultural shank mounted on a toolbar and pulled behind a tractor. Regional (whole-field out cross-validation) and hybrid (regional model including randomly chosen “local” calibration samples) spectral models were calibrated using partial least squares regression. Lab-based spectral data consistently provided more accurate predictions than on-the-go data. However, neither in situ nor lab-based spectroscopy yielded even semi-quantitative SOC predictions. For hybrid models with nine local samples included in the calibrations, standard error of prediction (SEP) values were 2.6 and 3.4gkg−1 for lab and on-the-go VisNIR respectively, with σSOC=3.2gkg−1. With an SOC coefficient of variation (CV)=26.7%, even with a relatively low SEP values, there was little SOC variability to explain. For clay content, hybrid-7 calibrations yielded lab SEP=53.1gkg−1 and residual product differential (RPD)=1.8 with on-the-go SEP=69.4gkg−1 and RPD=1.4. With more variability (σclay=91.4gkg−1 and CV=49.6%), both lab and on-the-go VisNIR show better explanatory power. There a number of potential explanations for degraded on-the-go predictive accuracy: soil heterogeneity, field moisture, consistent sample presentation, and a difference between the spatial support of on-the-go measurements and soil samples collected for laboratory analyses. In terms of predictive accuracy, our results are largely consistent with those previously published by Christy (2008), but on-the-go VisNIR was not able to capture the subtle SOC variability in Montana soils. Though the current configuration of the Veris on-the-go VisNIR system allows for rapid field scanning, on-the-go soil processing (i.e. drying, crushing, and sieving) could improve predictions.
Near infrared diffuse reflectance spectrophotometry, within the wavelength range 1100 to 2500 nm, was investigated for use in the simultaneous prediction of the moisture, organic C, and total N contents of air‐dried soils. An infraAlyzer 500 C (Technicon Instruments Corp.) scanning spectrophotometer was used to obtain near infrared reflectance of soils at 2‐nm intervals. Calibration equations for each of the soil constituents studied were based upon selection of the best combination of three wavelengths in a multiple regression analysis. The wavelengths selected for moisture, organic C, and total N, respectively, were 1926, 1954, and 2150 nm, 1744, 1870 and 2052 nm, and 1702, 1870 and 2052 nm. The standard errors of prediction for finely ground samples (<0.25 mm) from the top layers (0‐0.1, 0.1‐0.2, 0.2‐0.3, 0.3‐0.6 m) were 0.58, 0.16, and 0.014% for moisture, organic C, and total N, respectively. The standard errors of prediction, however, were much larger for coarsely ground soils (<2 mm), soils containing low amounts of organic C (<0.3%) and total N (<0.03%), and for those with a wide range in colors. Within a narrow range in soil color and at moderate amounts of organic matter (0.3–2.5%C), the near infrared reflectance technique provides a rapid, nondestructive, and simultaneous measurement of moisture, organic C and total N in soils
The data acquired from the hyperspectral airborne sensor DAIS-7915 over Izrael Valley in northern Israel was processed to yield quantitative soil properties maps of organic matter, soil field moisture, soil saturated moisture, and soil salinity. The method adopted for this purpose was the Visible and Near Infrared Analysis (VNIRA) approach, which yields an empirical model for predicting the soil property in question from both wet chemistry and spectral information of a representative set of samples (calibration set). Based on spectral laboratory data that show a significant capability to predict the above soil properties and populations using the VNIRA strategy, the next step was to examine this feasibility under a hyperspectral remote sensing (HSR) domain. After atmospherically rectifying the DAIS-7915 data and omitting noisy bands, the VNIRA routine was performed to yield a prediction equation model for each property, using the reflectance image data. Applying this equation on a pixel-bypixel basis revealed images that described spatially and quantitatively the surface distribution of each property. The VNIRA results were validated successfully from a priori knowledge of the area characteristics and from data collected from several sampling points. Following these examinations, a procedure was developed in order to create a soil property map of the entire area, including soils under vegetated areas. This procedure employed a random selection of more than 80 points along non-vegetated areas from the quantitative soil property images and interpolation of the points to yield an isocontour map for each property. It is concluded that the VNIRA method is a promising strategy for quantitative soil surface mapping, furthermore, the method could even be improved if a better quality of HSR data were used.
There has been growing interest in the use of diffuse reflectance as a quick, inexpensive tool for soil characterization. Some studies, using techniques like Partial Least Squares (PLS) regression of 1st derivative spectra have reported predictive accuracies for soil Organic C (OC) and Inorganic C (IC) that approach the analytical limits of standard laboratory measures. We applied 1st derivative Visible and Near-Infrared (VNIR) reflectance PLS regression modeling to soil samples obtained from six sites with similar soils across three counties in north central Montana, with five completely random 30% test sets selected for model validation. We obtained–relative to estimated SEL (Standard Error of Laboratory reference measurements) of 1.07 and 0.97 g kg−1 for OC and IC, respectively–SECV (calibration Standard Error of Cross-Validation) values of 1.04–1.20 and 1.54–1.63 g kg−1, and SEP (validation Standard Error of Prediction) values of 1.09–1.27 and 1.43–1.63 g kg−1. These results, together with validation RPD (Residual Prediction Deviation) values ≥2, could suggest a stable, effective PLS calibration that could be applied to similar soils in the same physiographic region. However, when we attempted to predict soil C for each of the six sites in turn using the remaining five sites for calibration, the models failed completely at two of the six sites and gave inconsistent results at a third site despite pre-screening for spectral similarity. “One-off” local calibrations for this study required ∼20–35% of the full samples, which could be prohibitively expensive for many applications. The results of this study demonstrate that “pseudo-independent” validation (random selection of non-independent test samples) can overestimate predictive accuracy relative to independent validation. The spatial structure of calibration and validation samples matters a great deal. Greater care needs to be taken to ensure that validation samples are independent to a degree that matches intended model use.
A digital elevation model (DEM) can be extracted automatically from stereo satellite images. During the past decade, the most common satellite data used to extract DEM was the across-track SPOT. Recently, the addition of along-track ASTER data, which can be downloaded freely, provides another attractive alternative to extract DEM data. This work compares the automated DEM extraction results using an ASTER stereo pair and a SPOT stereo pair over an area of hilly mountains in Drum Mountain, Utah, when compared to a USGS 7.5-min DEM standard product. The result shows that SPOT produces better DEM results in terms of accuracy and details, if the radiometric variations between the images, taken on subsequent satellite revolutions, are small. Otherwise, the ASTER stereo pair is a better choice because of simultaneous along- track acquisition during a single pass. Compared to the USGS 7.5-min DEM, the ASTER and the SPOT extracted DEMs have a standard deviation of 11.6 and 4.6 m, respectively.
A controlled investigation of the spectral signature of soil's structural crust in the 1.2- to 2.5-μm spectral region was conducted to investigate the feasibility of accounting for the crust status solely from the spectral information. This research was conducted because there is no valid method for in situ assessment of soil-crust characteristics in agricultural fields while and after the crust is formed. Through the use of a laboratory rainfall simulator and a sensitive spectrometer, we showed that significant spectral differences occurred between the crust and the bulk soil for three selected soils from Israel. The study was divided into two parts: Part 1, in which qualitative observations of the three soils were conducted under one level of rainstorm energy, and Part 2, where a selected soil was further investigated under varying levels of rain energy. The spectral differences obtained for the crusted soil are related to the texture and mineralogy of the soil's surface. It was concluded that the relationship between structural crust and soil reflectance spectroscopy can be used as a tool for estimating soil properties governed by the physical crust formation, such as infiltration rate, soil runoff, and erosion. It was further suggested that this methodology be tested within a remote sensing domain, using field, air, or spaceborne hyperspectral sensors.
In this paper we consider how we may determine the relative proportions of ground cover components in a mixed pixel. We assume the usual linear model for signal mixing and examine a number of methods, closely related, for estimating the proportions. We also show how the precision of our estimates can be defined. We introduce a new estimator which is based on regularisation principles and which produces a smoother set of images than other methods, and gives more accurate estimates. The methods are compared on a simulated data set.
Ingleby, H.R. and Crowe, T.G. 2001. Neural network models for predicting organic matter content in Saskatchewan soils. Canadian Biosystems Engineering/Le génie des biosystèmes au Canada 43:7.1- 7.5. Neural networks were developed to predict soil organic carbon content using reflectance data as inputs. Networks were subsequently evaluated through validation testing, and their predictive performance was compared to that of previously developed multiple-linear- regression (MLR) models. Neural networks tended to outperform MLR models even though the inputs were selected based on optimum linear model performance. Wavelengths of input reflectance data were the same for the best network and the best MLR model in three of the five fields tested. Investigation into determining optimum inputs for nonlinear network development is recommended. Des réseaux neuronaux furent développés afin de prédire la teneur en carbone organique du sol à partir des données de réflectance. Ces réseaux furent ensuite validés et leur capacité à prédire la teneur du sol en carbone fut comparée à celle de modèles de régression linéaire multiple (RLM) qui avaient été développés auparavant. La performance des réseaux neuronaux était généralement meilleure que celle des modèles, même si les données d'entrée avaient été choisies à partir de la performance optimale du modèle linéaire. Dans trois des cinq champs testés, les longueurs d'onde des données de réflectance étaient les mêmes pour le meilleur réseau et le meilleur modèle RLM. Il est recommandé d'entreprendre des recherches afin de déterminer les données d'entrée optimales nécessaires au développement d'un réseau non-linéaire.
Over the last 3yr, a cooperative effort between scientists and users to update the USLE is nearing completion and will produce a revised version of the USLE known as the RUSLE. Some of the improvements in the RUSLE will include: A greatly expanded erosivity map for the western United States. Minor changes in R factors in the eastern United States. Expanded information on soil erodibility. A slope length factor that varies with soil susceptibility to rill erosion. A nearly linear slope steepness relationship that reduces computed soil loss values for very steep slopes. A subfactor method for computing values for the cover-management factor. Improved factor values for the effects of contouring, terracing, stripcropping, and management practices for rangeland. -from Authors
Soil reflectance data were collected with a portable near infrared (NIR) spectrophotometer and were correlated with soil organic matter content in laboratory and field tests. Laboratory calibrations yielded a correlation of 0.89 and a standard error of prediction of 0.40% organic matter with 30 representative Illinois soils at 1.5 MPa and 0.033 MPa moisture tension levels. Limited in-furrow field operation produced much higher errors, due to the movement of soil past the sensor during data acquisition. Estimation of cation exchange capacity and soil moisture content was also accomplished in the laboratory. -Authors
Spectral mixture modeling was performed for identification and mapping of land degradation features related to soil erosion processes in the Sacaba Valley, Bolivia. The model allowed use of up to five surface components to characterize the selected area, since six bands (1, 2, 3, 4, 5, and 7) of the Landsat TM sensor were used as inputs. Among the various methods commonly used to determine end-members from the satellite image, three were selected: a) identification of one “pure” pixel representing a particular surface component from false color composites; b) average of “pure” pixels to characterize a particular end-member; and c) a method based on principal components. The best characterization of end-members was achieved by using average pure pixel reflectance. The median of the abundance images showed that, in 95% of the cases, the individual pixel compositions were explained by the selected surface components. The research has demonstrated that regional patterns of soil surface erosion features can be reliably mapped using linear spectral mixture analysis. Extrapolation of this approach to other regions where soil degradation features are correlated with spectrally distinguishable surface characteristics is feasible, provided that an optimization of the unmixing model as a function of local or regional surface component types is completed.
In a laboratory rainfall simulator study soil surface roughness was measured using contact (roller chain, pin meter) and noncontact devices (laser scanner, photogrammetry). Soil surfaces with two initial roughness conditions (aggregates < 20 mm and < 63 mm) were investigated before and after 90 mm of simulated rainfall. Measured plot area was 50 by 55 cm. A comparison of soil roughness measurement techniques was undertaken with regard to data acquisition and computation efforts, resolution, precision and capability to represent soil surface features.
Accelerated erosion involves preferential removal of soil organic carbon (SOC) because it is concentrated in vicinity of the soil surface and has lower density than the mineral fraction. The SOC transported by water runoff is redistributed over the landscape and deposite