A novel framework for improving soil organic matter prediction accuracy in cropland by integrating soil, vegetation and human activity information.

Remote sensing is an important tool for monitoring soil information. However, accurate spatial modeling of soil organic matter (SOM) in areas with high vegetation coverage, typically represented by agroecosystems, remains a challenge for field-scale estimation using remote sensing. To date, studies have focused on using single-period or multi-temporal vegetation information to characterize SOM. Thus, the relationship between SOM content and time-series vegetation biomass has not yet been fully explored. In addition, most studies have ignored the effects of critical soil properties and human activities (e.g., soil salinization, soil particle size fractions, history of land-use changes) on SOM. By integrating information on vegetation, soil, and human activities, we propose a novel framework for assessing SOM in cotton fields of artificial oases in northwest China, where returned straw is one of the primary sources of SOM coming from vegetation. We developed an Annual Maximum Biomass Accumulation Index (AMBAI) using time-series Landsat images from 1990 to 2019. Subsequently, we quantified the information of the planting years (PY) of cropland using spectral index threshold and incorporated proximal sensing data (soil hyperspectral and apparent conductivity data) and soil particle size fractions to establish a predictive model of SOM using partial least squares regression (PLSR), random forest (RF), and convolutional neural network (CNN). The results revealed that AMBAI had the highest correlation coefficient (r) with SOM (0.76, P < 0.01). AMBAI, soil hyperspectral data, and PY were the most relevant predictors for estimating SOM. The CNN model integrating vegetation, soil, and human activity information performed best, with coefficient of determination (R2), relative analysis error (RPD), and root mean square error (RMSE) of 0.83, 2.38 and 1.38 g kg-1, respectively. This study confirmed that AMBAI and PY had great potential for characterizing SOM in arid and semi-arid regions, providing a reference for other relevant studies.

Diagnosis of cadmium contamination in urban and suburban soils using visible-to-near-infrared spectroscopy.

Previous studies have mostly focused on using visible-to-near-infrared spectral technique to quantitatively estimate soil cadmium (Cd) content, whereas little attention has been paid to identifying soil Cd contamination from a perspective of spectral classification. Here, we developed a framework to compare the potential of two spectral transformations (i.e., raw reflectance and continuum removal [CR]), three optimization strategies (i.e., full-spectrum, Boruta feature selection, and synthetic minority over-sampling technique [SMOTE]), and three classification algorithms (i.e., partial least squares discriminant analysis, random forest [RF], and support vector machine) for diagnosing soil Cd contamination. A total of 536 soil samples were collected from urban and suburban areas located in Wuhan City, China. Specifically, Boruta and SMOTE strategies were aimed at selecting the most informative predictors and obtaining balanced training datasets, respectively. Results indicated that soils contaminated by Cd induced decrease in spectral reflectance magnitude. Classification models developed after Boruta and SMOTE strategies out-performed to those from full-spectrum. A diagnose model combining CR preprocessing, SMOTE strategy, and RF algorithm achieved the highest validation accuracy for soil Cd (Kappa = 0.74). This study provides a theoretical reference for rapid identification of and monitoring of soil Cd contamination in urban and suburban areas.

Diagnosing cropland's allowable range and spatial allocation in China's typical mountainous plateau area: An evaluation framework based on ecological carrying capacity.

Natural capital serves as a major constraint that affects the sustainable development of mountainous plateau areas. Determining ecological carrying capacity (ECC), as the key to measuring the critical natural capital of cropland, is needed for sustainable development. This study aims to provide new insights into ECC by diagnosing whether human activities are within the allowable range of natural capital and whether the spatial allocation of natural capital is reasonable in such specific areas. Taking Yunnan Province, China as the study area, we proposed an improved Ecological Footprint (EF) model to evaluate cropland's ecological capacity (CEC), and then, a framework of balance evaluation and spatial optimal allocation was constructed to examine the cropland's allowable range and optimize its spatial allocation if found unreasonable. Results show the following. (1) The per capita CEC of Yunnan Province between 2009 and 2016 decreased from 0.103 ha/capita to 0.095 ha/capita, and the cropland ecological balance index (EBI) presented a "critical overload" state ranging from 0.433 to 0.463, at which the supply exceeded the demand. Hence, the cropland was not within the allowable range in terms of supply-demand balance. (2) The comprehensive Gini coefficient of CEC was 0.462-0.515, and the gravity center of CEC deviated from the geometric center and shifted toward the westward, thereby, CEC is neither balanced in terms of spatial allocation nor coordinated with the population, economy, and resource environment. (3) The spatial allocation pattern of the study area was grouped into five zones on the basis of the optimization model. These zones are key optimization zone, adjustment optimization zone, maintenance zone, reasonable reduction zone, and key reduction zone. Accordingly, the targeted and differentiated strategies were accordingly put forward. Our study can contribute to identifying the practical approach to sustainable ecosystem management in mountainous plateau areas from the perspective of ECC and are beneficial for decision-making as regards new policies on cropland protection and Chinese ecological civilization construction.

Estimating lead and zinc concentrations in peri-urban agricultural soils through reflectance spectroscopy: Effects of fractional-order derivative and random forest.

Heavy metal contamination of peri-urban agricultural soil is detrimental to soil environmental quality and human health. A rapid assessment of soil pollution status is fundamental for soil remediation. Heavy metals can be monitored by visible and near-infrared spectroscopy coupled with chemometric models. First and second derivatives are two commonly used spectral preprocessing methods for resolving overlapping peaks. However, these methods may lose the detailed spectral information of heavy metals. Here, we proposed a fractional-order derivative (FOD) algorithm for preprocessing reflectance spectra. A total of 170 soil samples were collected from a typical peri-urban agricultural area in Wuhan City, Hubei Province. The reflectance spectra and lead (Pb) and zinc (Zn) concentrations of the samples were obtained in the laboratory. Two calibration methods, namely, partial least square regression and random forest (RF), were used to establish the relation between the spectral data and the two heavy metals. In addition, we aimed to explore the use of spectral estimation mechanism to predict the Pb and Zn concentrations. Three model evaluation parameters, namely, coefficient of determination (R2), root mean squared error, and ratio of performance to inter-quartile range (RPIQ), were used. Overall, the spectral reflectance decreased with the increase in Pb and Zn contents. The FOD algorithm gradually removed spectral baseline drifts and overlapping peaks. However, the spectral strength slowly decreased with the increase in fractional order. High fractional-order spectra underwent more spectral noises than low fractional-order spectra. The optimal prediction accuracies were achieved by the 0.25- and 0.5-order reflectance RF models for Pb (validation R2 = 0.82, RPIQ = 2.49) and Zn (validation R2 = 0.83, RPIQ = 2.93), respectively. A spectral detection of Pb and Zn mainly relied on their covariation with soil organic matter, followed by Fe. In summary, our results provided theoretical bases for the rapid investigation of Pb and Zn pollution areas in peri-urban agricultural soils.

Rapid identification of soil organic matter level via visible and near-infrared spectroscopy: Effects of two-dimensional correlation coefficient and extreme learning machine.

Accurate estimation of soil organic matter (SOM) is essential in understanding the spatial distribution of SOM to identify areas that need fertilization and the required grade of those fertilizers. Visible and near-infrared spectroscopy is a promising alternative to time consuming and costly conventional soil assessment methods. However, this approach is highly dependent on selecting suitable preprocessing strategies and data mining techniques for regression analysis. In this study, 2D correlation coefficients, including ratio, difference, and normalized difference indices, were introduced to select sensitive spectral parameters. The performance of extreme learning machine (ELM) was evaluated via comparison with that of support vector machine (SVM) for SOM estimation. A total of 257 soil samples were collected from Hubei Province, Central China, with SOM contents and reflectance spectra measured in the laboratory. Five spectral pretreatments, except for the raw spectra, were applied. SVM and ELM models were calibrated on spectral parameters selected by one-dimensional and 2D correlation coefficients and subsequently applied to predict SOM. Results showed that 2D correlation coefficient can effectively highlight the detailed SOM information compared with that of one-dimensional correlation coefficient. The ELM models yielded superior predictability relative to SVM models in all eight established models. The most excellent estimation accuracy was obtained by 2D ratio index and ELM (TRI-ELM) method, with an independent validation R2 and a ratio of performance to interquartile range of 0.83 and 3.49, respectively. The SOM fertility levels of predicted SOM showed that TRI-ELM method presented the largest similarity to laboratory-measured SOM levels, and misclassified samples were all concentrated within one error level. In summary, our study indicates that the TRI-ELM model is a rapid, inexpensive, and relatively accurate method for identifying SOM fertility level.

[Inversion of Soil Organic Matter Content Using Hyperspectral Data Based on Continuous Wavelet Transformation].

Soil organic matter content (SOMC) is an important parameter that reflect soil fertility available for crop production, and monitoring of the SOMC dynamically has shown great importance to promote the development of precision agriculture. In recent years, many researchers have tried to use proximal soil sensing, especially using the proximal hyperspectral techniques to acquire different kinds of spectral data under the field and laboratory conditions, and various new algorithms are also introduced to build inversion models to predict SOMC from spectra for different regions and different kinds of soils. In this paper, the hyperspectral reflectance of different soil samples was measured using the ASD FieldSpec3 spectrum analyzer. At the same time, the SOMC of each soil sample was analyzed using potassium dichromate external heating method in the laboratory. The correlation analyses between raw soil spectral reflectance (R) and SOMC were done, and it could select sensitive wavebands reflectance when the determination coefficients (R2) exceeded 0.15. A continuous wavelet transform (CWT) was also performed on R and the continuum removal curves (CR) to generate a wavelet power scalogram in different scales, the correlation analyses were done between wavelet power coefficients and SOMC, and it could select the sensitive wavelet coefficients when the R2 exceeded 0.3. Then, after extracting wavebands reflectance from R and wavelet power coefficients from R-CWT, CR-CWT, the estimation models for SOMC had been successfully built by partial least squares regression (PLSR), BP neural network (BPNN), support vector machine regression (SVMR), respectively. The results showed that, compared to the R2 between SOMC and R, the R2 between SOMC and R- CWT, CR-CWT wavelet coefficients were increased by about 0.15 and 0.2. The CR-CWT-SVMR model was the best, its R2, RMSE and RPD value of validation set were 0.83, 4.02, 2.48, which could estimate SOMC comprehensively and stably. For the CR-CWT-PLSR model, although there was a slight gap in the prediction accuracy with that CR-CWT-BPNN and CR-CWT-SVMR models, it also had its own unique advantages: the model was simple and thus the computation speed was reduced significantly. In the future, the results can provide good potential for field proximal sensing researching.