Estimation Of Soil Organic Matter Research Articles

Improving the Spatiotemporal Transferability of Hyperspectral Remote Sensing for Estimating Soil Organic Matter by Minimizing the Coupling Effect of Soil Physical Properties on the Spectrum: A Case Study in Northeast China

Soil organic matter (SOM) is important for the global carbon cycle, and hyperspectral remote sensing has proven to be a promising method for fast SOM content estimation. However, because of the neglect of the spectral response of soil physical properties, the accuracy and spatiotemporal transferability of the SOM prediction model are poor. This study aims to improve the spatiotemporal transferability of the SOM prediction model by alleviating the coupling effect of soil physical properties on spectra. Based on satellite hyperspectral images and soil physical variables, including soil moisture (SM), soil surface roughness (root-mean-square height, RMSH), and soil bulk weight (SBW), a soil spectral correction model was established based on the information unmixing method. Two important grain-producing areas in Northeast China were selected as study areas to verify the performance and transferability of the spectral correction model and SOM content prediction model. The results showed that soil spectral corrections based on fourth-order polynomials and the XG-Boost algorithm had excellent accuracy and generalization ability, with residual predictive deviations (RPDs) exceeding 1.4 in almost all the bands. In addition, when the soil spectral correction strategy was adopted, the accuracy of the SOM prediction model and the generalization ability after the model migration were significantly improved. The SOM prediction accuracy based on the XG-Boost-corrected spectrum was the highest, with a coefficient of determination (R2) of 0.76, a root-mean-square error (RMSE) of 5.74 g/kg, and an RPD of 1.68. The prediction accuracy, R2 value, RMSE, and RPD of the model after the migration were 0.72, 6.71 g/kg, and 1.53, respectively. Compared with the direct migration prediction of the model, adopting the soil spectral correction model based on fourth-order polynomials and XG-Boost reduced the RMSE of the SOM prediction results by 57.90% and 60.27%, respectively. This performance comparison highlighted the advantages for considering soil physical properties in regional-scale SOM predictions.

ESTIMASI KANDUNGAN BAHAN ORGANIK TANAH DI LAHAN TANAMAN JERUK, KECAMATAN DAU, KABUPATEN MALANG MENGGUNAKAN INDEKS VEGETASI DAN SISTEM INFORMASI GEOGRAFIS

Citrus plants (Citrus sp.) are most plants that grow in Petungsewu Village, Dau District. The decreased productivity of citrus plants causes the availability of citrus plants also to decrease. This needs to be increased to meet citrus plants' needs. The influence of soil organic matter is the main factor causing a decrease in the productivity of citrus plants in Petungsewu Village, Dau District. Therefore, it is necessary to estimate soil organic matter using the NDVI vegetation density index to obtain adequate and efficient results. Reliability testing was also carried out to know the accuracy level of the estimates made. The research was conducted at the Citrus Plantation and Subtropical Fruit Research Institute in Petungsewu Village, Dau District, Malang Regency. Estimation of soil organic matter using the NDVI vegetation density index showed an accuracy rate of 81.1%. The results of the accuracy analysis were strengthened by the presence of a paired t-test with a value of t = 0.01 and a value of p = 0.991.

Estimation of soil organic matter by in situ Vis-NIR spectroscopy using an automatically optimized hybrid model of convolutional neural network and long short-term memory network

It is attractive nowadays to estimate soil information based on in situ Visible and Near-infrared (Vis-NIR) spectroscopy. However, there exist a lot of errors, mainly due to modeling between soil properties and spectra. A hybrid of deep learning (DL) models seems an ideal approach for the modeling, e.g., convolutional neural network (CNN) and long short-term memory (LSTM). Nevertheless, it is not easy for a user, or not efficient for an automatic technique, such as Bayesian optimization (BO), to select hundreds of parameters of a hybrid of DL models. Recently, tree Parzen Estimator (TPE) based HyperBand within BO (BOHB) has been widely used in other fields for establishing a hybrid of DL models efficiently and automatically. Further, it is not quite clear if a DL model or a hybrid DL model performs well for different sample sizes and different numbers of spectral data. This study aimed to investigate these issues by evaluating performances of two hybrids of DL models (CNN-LSTM and LSTM-CNN) established using BOHB, based on soil spectra and organic matter contents of 670 soil samples collected from a forest in Nanning, southwest China. The performances were also compared with those of separate DL models and two commonly used methods, i.e., partial least square regression (PLSR) and random forest (RF), while the effect of sample sizes on the performances was also investigated. Besides, the effect of different numbers of spectra was also investigated, and the spectra were augmented using two methods, one by stacking scans of all measurement points of a sample and one by stacking preprocessed spectra using several methods. Results showed that, given less than 600 samples, accuracies of CNN-LSTM and LSTM-CNN were not higher than those of separate DL models, PLSR and RF. However, BOHB optimization largely improved the accuracies of the two hybrid models which were close to or higher than those of separate DL models, PLSR and RF. With the increase of sample sizes, accuracies of all models used in this study gradually increased and the DL models seemed to be more promising for large sample sizes. Furthermore, augmenting spectra with preprocessed data provided much more benefits than sample sizes, modeling methods, and optimization methods. Thus, it is promising to estimate soil using in situ Vis-NIR spectra based on a hybrid of DL models optimized using BOHB, particularly for large datasets or augmented spectra.

Impact of Spectral Resolution and Signal-to-Noise Ratio in Vis–NIR Spectrometry on Soil Organic Matter Estimation

Recently, considerable efforts have been devoted to the estimation of soil properties using optical payloads mounted on drones or satellites. Nevertheless, many studies focus on diverse pretreatments and modeling techniques, while there continues to be a conspicuous absence of research examining the impact of parameters related to optical remote sensing payloads on predictive performance. The main aim of this study is to evaluate how the spectral resolution and signal-to-noise ratio (SNR) of spectrometers affect the precision of predictions for soil organic matter (SOM) content. For this purpose, the initial soil spectral library was partitioned into to two simulated soil spectral libraries, each of which were individually adjusted with respect to the spectral resolutions and SNR levels. To verify the consistency and generality of our results, we employed four multiple regression models to develop multivariate calibration models. Subsequently, in order to determine the minimum spectral resolution and SNR level without significantly affecting the prediction accuracy, we conducted ANOVA tests on the RMSE and R2 obtained from the independent validation dataset. Our results revealed that (i) the factors significantly affecting SOM prediction performance, in descending order of magnitude, were the SNR levels > spectral resolutions > estimation models, (ii) no substantial difference existed in predictive performance when the spectral resolution fell within 100 nm, and (iii) when the SNR levels exceeded 15%, altering them did not notably affect the SOM predictive performance. This study is expected to provide valuable insights for the design of future optical remote sensing payloads aimed at monitoring large-scale SOM dynamics.

A new method to estimate soil organic matter using the combination model based on short memory fractional order derivative and machine learning model

In order to improve the quantitative inversion accuracy of soil organic matter in arid areas, the soils affected by different human disturbances in Xinjiang were used as research objects (Regions I, II and III). This study compared and analyzed the preprocessing hyperspectral data effects of GMFOD (global memory fractional order derivative) with 1759 memory length and SMFOD (short memory fractional order derivative) with different memory lengths (L = 1500, 1200, 1000, 800, 500, 300, 250, 200, 150, 100), machine learning models based on RBF (radial basis function) and ELM (extreme learning machine) were used to explore the best model for predicting soil organic matter content. Simulation results showed that (1) SMFOD had shorter computational runtime compared to GMFOD, and the 100 SMFOD had the shortest running time. When all orders of fractional order derivative were calculated, the time required for 100 SMFOD was 115.27 s, while the time required for GMFOD was 1025.32 s. These indicated that the running time of the former was 789.49% shorter than that of GMFOD. (2) Comparing the correlation coefficients of hyperspectral and soil organic matter in all fractional orders calculated based on GMFOD and SMFOD, it was found that the maximum correlation coefficient of 100 SMFOD and GMFOD differed by 0.0103, 0.0058 and 0.0287 in Regions I, II and III, which means that GMFOD only increased by 0.94%, 1.71% and 4.25%. These indicated that the maximum correlation coefficient of SMFOD and GMFOD was very similar at each fractional order. (3) The best model for inversion of soil organic matter in all three regions was SMFOD-CC1-RBF (CC1 denotes bands passed the 0.01 significance test), which was located at the fractional orders of 1.4, 1.8 and 1.1, respectively. Meanwhile, R2 (coefficient of determination), RMSE (root mean square error), RPD (ratio of the performance to deviation) were 0.8249, 0.7656, 1.9256 in Region I; were 0.7630, 1.4422, 1.9561 in Region II; and were 0.6799, 1.5343,1.8513 in Region III, which indicated that the higher-order SMFOD had better performance in inverting organic matter than the lower-order SMFOD, and the prediction ability of this model for soil organic matter in the three regions was good.

Using Soil Testing Data to Monitor Changes in Maine Agricultural Soil Organic Carbon from 1995 to 2021

ABSTRACT Estimates of soil organic matter (SOM) are commonly part of soil testing, and typically use the loss-on-ignition (LOI) method. To explore the utility of existing soil testing databases as a resource for estimating soil organic carbon (SOC) concentrations over time, we determined the relationship between SOM estimated by LOI and SOC measured using IR detection of CO2 for Maine soils. We then examined trends in SOC concentrations in Maine agricultural soils over the past 27 years using data from 97,713 standard soil samples received by the Maine Soil Testing Service. For 91 agricultural soil samples representative of Maine cropping systems, we found that the standard LOI temperature used by the Maine Soil Testing Service, 375°C, produced LOI values that correlated more strongly with SOC measured by IR detection of CO2 than the three higher temperatures we studied, although all temperatures produced data that was well correlated (R2 values from 0.95 to 0.98). Soil texture did not influence the correlation. Our analysis of soil testing data from 1995 to 2021 suggested increases in SOC concentrations in soils planted to grain and hay, and possibly corn, broadly corresponding with changing agricultural practices occurring over this time frame. Of our four focal crop categories, only potato soils decreased in SOC concentration from 1995 to 2021, underscoring the unique challenges associated with potato soil management relative to other cropping systems in Maine. Our study confirms the potential utility of soil testing databases in assessing changes in SOC concentrations in agricultural soils over decades.

Estimating Soil Organic Matter (SOM) Using Proximal Remote Sensing: Performance Evaluation of Prediction Models Adjusted at Local Scale in the Brazilian Cerrado

The quantification of soil organic matter (SOM) has increased over the years, especially in the Brazilian Cerrado region, one of the most important areas for grain production in the country. In this area, SOM content tends to be low, which directly impacts the physical, chemical and biological quality of soils. Thus, the use of spectroradiometry has been widely evaluated to investigate whether it can be used as a faster, more reliable and cheaper solution to meet the SOM estimation. In this context, the objective of the present paper was to evaluate the performance of a local spectral model for SOM prediction generated through the spiking strategy. The research was developed in the municipality of Passos, Minas Gerais State, located in the Brazilian Cerrado. Soil samples (0–0.2 m and 0.2–0.4 m depths) were collected in a zigzag pattern and split in calibration of the local models from a test area (90 soil samples) and recalibration and validation from a target area (46 soil samples). After this stage, the SOM contents were determined in a laboratory, and the spectral responses (350–2500 nm) of each soil sample were collected. From the target area, 10, 25 and 50% of soil spectra were selected for recalibration of the local models generated for the test area. Although median results were observed in the post-recalibration, due to the type of sample selected and the relative similarity among the spectral curves of both areas, improvement was observed for all statistical indices, especially when using 50% (23) of samples for recalibration of the local models, reaching r2 = 0.43, RMSEP = 2.34 gdm−3 and RPIQ = 4.58. These results are important for the SOM estimation in the Brazilian Cerrado considering its importance to the food security and socioeconomic activities. However, considering the lack of similar research in the study area, it is necessary to further investigate the development of spectral models on a local scale and their contribution to improve the identification of SOM spatial variability.

A novel semi-empirical soil multi-factor radiative transfer model for soil organic matter estimation based on hyperspectral imagery

Soil organic matter (SOM) concentration is an important factor affecting soil quality, and rapid and wide-scale monitoring of SOM concentration is a key step toward sustainable agriculture. Hyperspectral technology is widely used in soil composition monitoring, due to its rich spectral information. The complex imaging environment of hyperspectral imagery and the mixed pixel problem have led to the current applications of soil condition estimation mostly using data-driven methods. However, the estimation process based on data-driven methods cannot be explained by radiative transfer theory. Therefore, in this paper, a semi-empirical soil multi-factor radiative transfer (SESMRT) model combining a soil radiative transfer model and data-driven model is proposed for SOM estimation based on hyperspectral imagery. The radiative transfer model in the SESMRT model fully considers the soil components, such as SOM, soil moisture, soil iron oxides, and particle size distribution, and achieves high-precision simulation of soil spectra by resolving the mechanism of the influence of soil components on spectra. The SOM spectra calculated by the radiative transfer model are then used to estimate the SOM concentration based on a data-driven model. The SOM spectra eliminate the interference of other factors in the spectrum and significantly enhance the correlation between the spectrum and SOM. Thus, the SOM concentration can be estimated with a high degree of accuracy (R2 = 0.660 ± 0.034, RMSE = 3.923 ± 0.236, RPIQ = 2.003 ± 0.118 for GF5 and R2 = 0.685 ± 0.030, RMSE = 3.543 ± 0.311, RPIQ = 2.309 ± 0.207 for HyMap). Finally, the comparison of the estimated results for 2017 and 2019 shows that the SOM concentration of the cultivated soils in the study area increased, while the opposite is observed for the soils around the mining areas.

The modified model for hyperspectral estimation of soil organic matter using positive and inverse grey relational degree

PurposeIn order to improve the estimation accuracy of soil organic matter, this paper aims to establish a modified model for hyperspectral estimation of soil organic matter content based on the positive and inverse grey relational degrees.Design/methodology/approachBased on 82 soil sample data collected in Daiyue District, Tai'an City, Shandong Province, firstly, the spectral data of soil samples are transformed by the first order differential and logarithmic reciprocal first order differential and so on, the correlation coefficients between the transformed spectral data and soil organic matter content are calculated, and the estimation factors are selected according to the principle of maximum correlation. Secondly, the positive and inverse grey relational degree model is used to identify the samples to be identified, and the initial estimated values of the organic matter content are obtained. Finally, based on the difference information between the samples to be identified and their corresponding known patterns, a modified model for the initial estimation of soil organic matter content is established, and the estimation accuracy of the model is evaluated using the mean relative error and the determination coefficient.FindingsThe results show that the methods of logarithmic reciprocal first order differential and the first-order differential of the square root for transforming the original spectral data are more effective, which could significantly improve the correlation between soil organic matter content and spectral data. The modified model for hyperspectral estimation of soil organic matter has high estimation accuracy, the average relative error (MRE) of 11 test samples is 4.091%, and the determination coefficient (R2) is 0.936. The estimation precision is higher than that of linear regression model, BP neural network and support vector machine model. The application examples show that the modified model for hyperspectral estimation of soil organic matter content based on positive and inverse grey relational degree proposed in this article is feasible and effective.Social implicationsThe model in this paper has clear mathematical and physics meaning, simple calculation and easy programming. The model not only fully excavates and utilizes the internal information of known pattern samples with “insufficient and incomplete information”, but also effectively overcomes the randomness and grey uncertainty in the spectral estimation of soil organic matter. The research results not only enrich the grey system theory and methods, but also provide a new approach for hyperspectral estimation of soil properties such as soil organic matter content, water content and so on.Originality/valueThe paper succeeds in realizing both a modified model for hyperspectral estimation of soil organic matter based on the positive and inverse grey relational degrees and effectively dealing with the randomness and grey uncertainty in spectral estimation.

Soil Organic Matter Inversion Based on Imaging Spectral Data in Straw-Covered Noncultivated Land

The rapid inversion of soil organic matter is of great significance for agricultural soil testing and fertilization in order to protect and utilize land resources effectively. This study selected the developmental base in Lishu County with typical characteristics of China’s black soil as its research area. This study established a 600 nm “bow curvature difference” spectral index and the partial least squares regression model, and the accuracy of their results was compared. The correlation between the 600 nm “bow curvature” spectral index and the soil organic matter of straw cover no-tillage is analyzed. The average soil organic matter content in the study area was 16.67 g·kg-1, and the organic matter increased significantly from NT-0 to NT-100 by 16.26 g·kg-1. The study provided a deep insight to improve the quantitative inversion methods to estimate soil organic matter.

Mapping Forest Cover and Estimating Soil Organic Matter by GIS-Data and an Empirical Model at the Subnational Level in Mexico

Forests play an essential role in climate change as they are the terrestrial ecosystems that store the highest C content in their soils and biomass. Despite this, the lack of information at the subnational level hinders their proper management and conservation. This study aimed to identify the extension and distribution of forests and to develop an empirical model for the spatial prediction of soil organic matter (SOM) in Ixtacamaxtitlan, Puebla, Mexico, based on environmental variables generated through Geographical Information Systems. A supervised classification in Landsat 8 images was used to define the forest cover, and environmental variables related to topography, climate and vegetation were generated. Finally, a Multiple Linear Regression model validated with the leave-one-out cross-validation method was used to examine the relationships between the covariates and the SOM and estimate its content in forest. The results show that the forest cover extension is 41%, with an overall accuracy of 97.7%. The model shows a good fit (R2cv = 0.69, RMSEcv = 1.53). The mean of SOM was 5.2%, and upper values were consistent with higher altitude, precipitation and cooler temperature. Estimating SOM content in forest areas is essential in developing planning strategies at the subnational level to mitigate the harmful effects of climate change.

Estimation of soil organic matter content in Yinchuan Plain based on fractional derivative combined with spectral indices.

Soil organic matter (SOM) is a crucial indicator of soil fertility. Field hyperspectral reflectance and laboratory SOM data of soil samples from the Yinchuan Plain were used to explore the performance of models based on fractional derivative combined with different spectral indices. Following reciprocal and logarithmic transformation, the reflectance data were processed using fractional derivative from 0 to 2 orders (interval 0.20). Then, the difference index (DI), ratio index (RI), brightness index (BI), normalized difference index (NDI), renormalized difference index (RDI), and generalized difference index (GDI) were constructed. The two-dimensional correlation between the six indices and SOM content were analyzed. The optimal spectral indices were selected to establish SOM estimation models with principal component regression (PCR), partial least square regression (PLSR), back propagation neural network (BPNN), support vector machine (SVM), and geographically weighted regression (GWR). Results showed that the maximum absolute correlation coefficient (MACC) values between DI, RI, NDI, BI, GDI, RDI, and SOM contents increased firstly and then decreased, with the highest values observed at 1.0, 0.6, 1.4, and 1.6 orders. The 0.2-2.0 order RDI under fractional derivative variation could be used for subsequent model construction, in which the optimal combinations of bands for MACC values were mainly concentrated at 400-600 nm and 1300-1700 nm. Among the different models based on the single spectral index RDI, the model based on SVM achieved the highest estimation accuracy, whose modeling determination coefficient, verification determination coefficient and relative percentage difference reached 0.86, 0.87 and 2.32. Our results would provide a scientific reference for quick and accurate SOM assessment and mapping in areas with relatively low SOM content.

A dynamic normalized difference index for estimating soil organic matter concentration using visible and near-infrared spectroscopy

Accurate, rapid, and non-destructive estimation of soil organic matter (SOM) is crucial for soil fertility diagnosis and precision farming. Due to the complicated and unstable spectral characteristics of SOM, few SOM spectral indexes have been proposed and widely used. In this paper, a new dynamic normalized difference index (DNDI) was proposed and constructed to estimate SOM using visible and near-infrared spectroscopy. A dynamic correction factor α was used to adjust the optimal wavelength range to obtain more robust features of SOM. Different spectral pre-processing methods were applied and compared. The support vector machine (SVM) model and Partial least square regression (PLSR) model were calibrated based on DNDI and applied to estimate SOM. To this end, a total of 111 soil samples were collected in the southern coastal plain of Laizhou Bay. The results showed that the DNDI index constructed by wavelength optimization could have a higher correlation with SOM than the two-dimensional normalized difference index (NDI). DNDI had the maximum correlation of 0.88 from the first derivative of reflectance, and the NDI correlations were most improved by standard normal variate transform (SNV), with the maximum correlation reaching 0.81. For SOM estimation models, DNDI exhibited better model performance, yielding a validation R2, RMSE, and RPD of 0.78, 0.17 g·kg−1, and 2.01, respectively. Our algorithm has strong application potential for estimating other soil properties and enhancing the application of ground- and satellite-based sensing.

A Framework for High-Resolution Mapping of Soil Organic Matter (SOM) by the Integration of Fourier Mid-Infrared Attenuation Total Reflectance Spectroscopy (FTIR-ATR), Sentinel-2 Images, and DEM Derivatives

Soil organic matter (SOM), as the greatest carbon storage in the terrestrial environment, is inextricably related to the global carbon cycle and global climate change. Accurate estimation and mapping of SOM content are crucial for guiding agricultural output and management, as well as controlling the climate issue. Traditional chemical analysis is unable to satisfy the dynamic estimation of SOM due to its low timeliness. Remote and proximal sensing have significant advantages in terms of ease of use, estimation accuracy, and geographical resolution. In this study, we developed a framework based on machine learning to estimate SOM with high accuracy and resolution using Fourier mid-infrared attenuation total reflectance spectroscopy (FTIR-ATR), Sentinel-2 images, and DEM derivatives. This framework’s performance was evaluated on a regional scale using 245 soil samples from northeast China. Results indicated that the calibration size could be shrunk to 50% while achieving a fair prediction performance for SOM content. The Lasso, partial least squares (PLS), support vector regression (SVR), and convolutional neural networks (CNN) performed well in predicting SOM from FTIR-ATR spectra, and the performance was enhanced further by using Sentinel-2 images and DEM derivates. The PLS, SVR, and CNN models created SOM maps with higher spatial resolution and variation than the Kriging approach. The PLS and SVR models provided enough variety and were more realistic in the local SOM map, making them usable at the field scale, and the suggested framework took a fresh look at high-resolution SOM mapping.

Estimation of soil organic matter in the Ogan-Kuqa River Oasis, Northwest China, based on visible and near-infrared spectroscopy and machine learning

Estimation of soil organic matter in the Ogan-Kuqa River Oasis, Northwest China, based on visible and near-infrared spectroscopy and machine learning

Grey fuzzy prediction model of soil organic matter content using hyper-spectral data

PurposeIn order to overcome the uncertainty and improve the accuracy of spectral estimation, this paper aims to establish a grey fuzzy prediction model of soil organic matter content by using grey theory and fuzzy theory.Design/methodology/approachBased on the data of 121 soil samples from Zhangqiu district and Jiyang district of Jinan City, Shandong Province, firstly, the soil spectral data are transformed by spectral transformation methods, and the spectral estimation factors are selected according to the principle of maximum correlation. Then, the generalized greyness of interval grey number is used to modify the estimation factors of modeling samples and test samples to improve the correlation. Finally, the hyper-spectral prediction model of soil organic matter is established by using the fuzzy recognition theory, and the model is optimized by adjusting the fuzzy classification number, and the estimation accuracy of the model is evaluated using the mean relative error and the determination coefficient.FindingsThe results show that the generalized greyness of interval grey number can effectively improve the correlation between soil organic matter content and estimation factors, and the accuracy of the proposed model and test samples are significantly improved, where the determination coefficient R2 = 0.9213 and the mean relative error (MRE) = 6.3630% of 20 test samples. The research shows that the grey fuzzy prediction model proposed in this paper is feasible and effective, and provides a new way for hyper-spectral estimation of soil organic matter content.Practical implicationsThe research shows that the grey fuzzy prediction model proposed in this paper can not only effectively deal with the three types of uncertainties in spectral estimation, but also realize the correction of estimation factors, which is helpful to improve the accuracy of modeling estimation. The research result enriches the theory and method of soil spectral estimation, and it also provides a new idea to deal with the three kinds of uncertainty in the prediction problem by using the three kinds of uncertainty theory.Originality/valueThe paper succeeds in realizing both the grey fuzzy prediction model for hyper-spectral estimating soil organic matter content and effectively dealing with the randomness, fuzziness and grey uncertainty in spectral estimation.

Mapping Soil Organic Matter Content during the Bare Soil Period by Using Satellite Data and an Improved Deep Learning Network

Soil function degradation has impaired global work in the implementation of sustainable development goals (SDGs), and soil organic matter (SOM) is a basic and the most important indicator. The deep learning neural network (i.e., DL network) has become a popular tool for mapping SOM content at a regional scale. However, outlier sample data caused by environmental factors (e.g., moisture and vegetation) and uncertain noise (e.g., random noise and instability effects) have interfered with the determination of the function mapping relationship between target soil properties and spectral features, leading to DL networks with low generalization capability. Therefore, we introduced a spatial association module into a deep neural network to remove outlier sample data in order to construct an optimal sample set for calibrating an improved deep learning (IDL) network for SOM mapping. A total of 707 soil samples and a Sentinel-2B multispectral image were acquired during the bare soil period in Weibei, China. The variable importance in the projection approach was used to select the SOM-responsive spectral features for model inputs. Measured SOM contents were taken as the dependent variable, and the IDL network was constructed and applied to map SOM at the regional scale. The results showed that Band 11 was the most important band for SOM prediction. The band difference transformation method was able to integrate multiple-band information and enhance the absorption signal of SOM. The optimal SOM-responsive spectral features included B11, B1–B11, B2−B11, B3−B11, B4−B11, and B1−B12. The IDL network exhibited better performance (R2 = 0.92; RPIQ = 4.57) regarding SOM estimation compared with the DL model performance (R2 = 0.84; RPIQ = 2.84), being improved by 9.52% (for R2) and 60.92% (for RPIQ). After introducing the spatial association module, the DL network generalization capability was enhanced. SOM distribution showed a high-value (>20 g kg−1) area in the south, and a low-value (<6 g kg−1) area in the north of the study area (the area affected by seawater intrusion). These results provide a strategy based on an IDL network and satellite data for effectively and accurately mapping SOM at the regional scale.

Combining Multitemporal Sentinel-2A Spectral Imaging and Random Forest to Improve the Accuracy of Soil Organic Matter Estimates in the Plough Layer for Cultivated Land

Soil organic matter (SOM) is vital for assessing the quality of arable land. A fast and reliable estimation of SOM is important to predict the soil carbon stock in cropland. In this study, we aimed to explore the potential of combining multitemporal Sentinel-2A imagery and random forest (RF) to improve the accuracy of SOM estimates in the plough layer for cultivated land at a regional scale. The field data of SOM content were utilized along with multitemporal Sentinel-2A images acquired over three years during the bare soil period to develop spectral indices. The best bands and spectral indices were selected as prediction variables by using the RF algorithm. Partial least squares (PLS), geographically weighted regression (GWR), and RF were employed to calibrate spectral indices for the SOM content, and the optimal calibration model was used for the mapping of the SOM content in arable land at a regional scale. The results showed the following. (1) The multitemporal image estimation model outperformed the single-temporal image estimation model. The estimation model that utilized the optimal bands and spectral indices as prediction variables usually had better accuracy than the models based on full spectral data. (2) For the SOM content estimates, the performance was better with RF than with PLS and GWR in almost all cases. (3) The most accurate SOM estimation in the case area was achieved by using multitemporal images from 2018 and the RF calibration model based on the optimal bands and spectral indices as prediction variables, with R2val (coefficient of determination of the validation data set) = 0.67, RMSEval (root mean square error of the validation dataset) = 2.05, and RPIQval (ratio of performance to interquartile range of the validation dataset) = 3.36. (4) The estimated SOM content in the plough layer for cultivated land throughout the study area ranged from 16.17 to 36.98 g kg−1 and exhibited an increasing trend from north to south. In the current study, we developed a framework that combines multitemporal remote sensing imagery and RF for the SOM estimation, which can improve the accuracy of quantitative SOM estimations, provide a dynamic, rapid, and low-cost technique for understanding soil fertility, and offer an early warning of changes in soil quality.

Improving soil organic matter estimation accuracy by combining optimal spectral preprocessing and feature selection methods based on pXRF and vis-NIR data fusion

Rapid and accurate estimation of soil organic matter (SOM) content is of great significance for agricultural production and carbon stock estimation. Visible-near-infrared spectroscopy (Vis-NIR) and portable X-ray fluorescence spectroscopy (pXRF), with the fast, low-cost, and environmental friendly characteristics, can realize the rapid large-scale detection of SOM. However, due to the heterogeneity and complexity of soil, estimation of SOM using a single sensor often leads to high uncertainty in result. Multi-sensor data fusion is one of the approaches to solve this problem, but suitable spectral preprocessing methods and data fusion methods still need further research. In this study, 500 soil samples were collected from two watersheds with large differences in soil parent material, soil type, and climate in the Xinjiang Uygur Autonomous Region of China. The effects of different spectral preprocessing methods for vis-NIR and pXRF spectra on SOM estimation were compared, and the effects of five feature selection methods on single sensor estimation and three different multi-sensor fusion estimations were evaluated. The results showed that vis-NIR spectrum was better than pXRF spectrum for SOM estimation, and the 1.6th-order derivative and baseline correction were the optimal preprocessing method for vis-NIR and pXRF spectra (Beam 3), respectively. Feature selection could improve the accuracy of the single sensor and multi-sensor fusion estimations. The competitive adaptive reweighted sampling (CARS) method was the optimal feature selection method for single sensor estimation and Granger–Ramanathan averaging (GRA) fusion estimation, and the variable importance in projection (VIP) was the optimal feature selection method for direct concatenation and outer-product analysis fusion estimation. In general, the optimal SOM estimation (RMSEp = 3.59 g kg−1, LCCCp = 0.84, RPIQp = 2.67) could be achieved based on CARS feature selection method and GRA fusion. This study proves that the combination of spectral preprocessing and feature selection can effectively improve the accuracy of the single sensor and multi-sensor fusion estimations. Especially, the multi-sensor fusion estimation has higher accuracy than the single sensor estimation. This study will provide a reference for the application of proximal soil sensing.

Local wavelet packet decomposition of soil hyperspectral for SOM estimation

It is a critical work to accurate extraction of spectral characteristics of soil organic matter (SOM), which will help to improve the model performance for SOM estimation. In this paper, we propose a new SOM prediction model using local wavelet characteristics of soil spectrum. Specifically, six level local wavelet packets decomposition are considered with different moving window sizes and sliding lengths, and thus formulate two new spectral indicators, namely, local wavelet packet energy spectrum (LWE) and local logarithmic wavelet packet energy spectrum (LLWE). The LWE and LLWE are then viewed as model inputs, which are respectively used to forecast SOM content. We test the prediction performance of the LWE and LLWE based on the two prediction models, that is, multiple linear regression (MLR) and ridge regression with cross-validation (RCV). The result shows that the two new spectral indicators help to enhance the spectral response information of SOM. Among the four prediction models, the LLWE-MLR is the most outstanding. Compared to the original hyperspectral and energy features vectors, the LWE and LLWE bring the better model performance. Since both the high and low-frequency components of local information of soil hyperspectral are fully extracted, the two new spectral indicators significantly improve the prediction accuracy and reliability of SOM.