Search

Facial expression recognition is a fine-grained task because different emotions have subtle facial movements. This paper proposes to learn inter-class optical flow difference using generative adversarial networks (GANs) for facial expression recognition. Initially, the proposed method employs a GAN to produce inter-class optical flow images from the difference between the static fully expressive samples and neutral expression samples. Such inter-class optical flow difference is used to highlight the displacement of facial parts between the neutral facial images and fully expressive facial images, which can avoid the disadvantage that the optical flow change between adjacent frames of the same video expression image is not obvious. Then, the proposed method designs four-channel convolutional neural networks (CNNs) to learn high-level optical flow features from the produced inter-class optical flow images, and high-level static appearance features from the fully expressive facial images, respectively. Finally, a decision-level fusion strategy is adopted to implement facial expression classification. The proposed method is validated on two public facial expression databases, BAUM_1a, SAMM and AFEW5.0, demonstrating its promising performance.

Video-based facial expression recognition (VFER) is the fundamental feature of various computer vision applications. Visual features are the key factors for facial expression recognition. However, the gap between the visual features and the emotions is large. In order to bridge the gap, the proposed method utilises convolutional neural networks (CNNs) and histogram of oriented gradient (HOG) to obtain the more comprehensive feature for VFER. Firstly, it extracts shallow features from the video frame through a number of convolutional kernels in CNNs, which has the characteristics of displacement, scale and deformation invariance. Then, the HOG is employed to extract HOG features from CNN's shallow features, which are strongly correlated with facial expressions. Finally, the support vector machine (SVM) is employed to conduct the task of facial expression recognition. The extensive experiments on RML, CK+ and AFEW5.0 database show that this framework takes on the promising performance and outperforming the state of the arts.

Increasing studies have begun to focus on biodiversity-productivity relationships for soil microorganisms through molecular ecology methods. However, most of these studies involve controlled experiments, and whether the relationship remains at large spatial scales is still largely unknown. To unravel this issue, archived desert soils from long-term experiments were analysed using high-throughput sequencing, and satellite-derived vegetation datasets were acquired to quantify productivity. Most of the abundant genera were significantly different between low- and high-productivity conditions, and soil bacterial communities were strongly impacted by productivity. Soil bacterial biodiversity, including observed OTUs (operational taxonomic units) and the Chao1, Shannon, and Faith's PD indexes, increased rapidly with productivity at low levels and then reached a relatively stable state, and similar phenomena were observed at multiple taxonomic ranks and for most of the dominant groups. Furthermore, we discovered that the mechanisms resulting in the observed relationship might be ecosystem resource availability in large-scale regions and species competition in local regions. Collectively, these results enhance our understanding of the linkage between belowground microorganisms and aboveground vegetation in arid and semi-arid areas and confirm the potential value of satellite-derived datasets in research on soil microbial diversity at large spatial scales. This article is protected by copyright. All rights reserved.

The spatial differences in microbial communities have been widely investigated in many environments through culture-independent molecular biological methods. Currently, for archived soils, there is a concern whether the spatial differences are still preserved in them after long-term storage. Considering plenty of archived soils being stored by institutions, we tried to explore whether they could provide certain information about microbial biogeography. In the present study, high-throughput sequencing of the 16S rRNA gene was conducted for archived soil samples, which were collected in different years from four agricultural long-term experimental stations of the Chinese Ecosystem Research Network and had been stored for approximate 1, 6 and 11 years. After processing the raw high-throughput sequencing data, a total of 10,623 OTUs belonging to 46 phyla were detected across all 72 soil samples. The results of OTU analysis showed that the overlap ratio (>57.35%) of bacterial OTUs between sampling years at the same station was higher than that (

The vast diversity of soil bacteria provides essential ecosystem services that support agricultural production. Variation in the diversity and composition of soil biota may have predictive values for soil nutrient cycling and resilience of ecosystem services, thus providing valuable insights to improve food production. The North China Plain (NCP) is one of the world’s key agricultural regions, supplying more than 50% of the cereal consumed in Asia. However, it is unknown whether soil microbial diversity is predictable across the NCP. Using the MiSeq Illumina platform, we examined bacterial community variation in relation to spatial and environmental factors from 243 soils in wheat-maize double cropping rotation fields across the NCP, which cover nearly 0.3 million km2. Based on observed bacterial communities and their relationships with environmental factors, we generated a map of bacterial communities across the NCP. The highest bacterial diversity was found in the middle part of the NCP, with most of the variation in diversity attributable to differences in the community similarity of Actinobacteria and Alphaproteobacteria. These findings provide important baseline information for analyzing the relationships between microbial community, soil functionality and crop yields.

Many precious air-dried soil samples are currently stored by institutions across the world, and they have been proved useful in retrieving historical information of soil microbial ecology, such as the shifts of microbial populations and the effects of agricultural managements on soil microbial communities. However, most of those retrospective studies on microorganisms in archived soils were mainly conducted through PCR-DGGE or T-RFLP analysis, which was less sensitive to low-abundance species. In order to assess whether archived soils could reveal the spatial distribution of microbial communities at large spatial scale, especially the rare microbial taxa, a total of 117 serial time-historical samples from eight long-term field experimental stations of the Chinese Ecosystem Research Network were analyzed using high-throughput sequencing. The non-metric multidimensional scaling (NMDS) analyses showed that both the total bacterial communities and the rare bacterial subcommunities were distinctly clustered into different groups corresponding to each experimental station, and that analogous distribution patterns of bacterial communities existed between each two sampling years for both the total and the rare taxa, though the microbial community composition changed over time. The corresponding hierarchical clustering analyses demonstrated that soil bacterial communities for both the total and the rare taxa could be clustered into five groups for all samples and for the sub-samples of each sampling year, which corresponded to five eco-geographic regions in the Eco-Geographic Regional System of China. The results from this investigation highlighted the great value of serial time-archived soils in revealing the spatial distribution patterns of bacterial communities for both the total and the rare taxa at large spatial scale.

One key challenging issues of facial expression recognition (FER) in video sequences is to extract discriminative spatiotemporal video features from facial expression images in video sequences. In this paper, we propose a new method of FER in video sequences via a hybrid deep learning model. The proposed method first employs two individual deep convolutional neural networks (CNNs), including a spatial CNN processing static facial images and a temporal CN network processing optical flow images, to separately learn high-level spatial and temporal features on the divided video segments. These two CNNs are fine-tuned on target video facial expression datasets from a pre-trained CNN model. Then, the obtained segment-level spatial and temporal features are integrated into a deep fusion network built with a deep belief network (DBN) model. This deep fusion network is used to jointly learn discriminative spatiotemporal features. Finally, an average pooling is performed on the learned DBN segment-level features in a video sequence, to produce a fixed-length global video feature representation. Based on the global video feature representations, a linear support vector machine (SVM) is employed for facial expression classification tasks. The extensive experiments on three public video-based facial expression datasets, i.e., BAUM-1s, RML, and MMI, show the effectiveness of our proposed method, outperforming the state-of-the-arts.

In arid and semiarid desert areas, climate factors distinctly impact soil microbial community, which can also be greatly altered after agricultural practices at multiple spatial scales. However, it is still poorly unknown whether the effects of climate on soil microbial diversity change after intensive agriculture at a large spatial scale. To uncover this concern, we used time-interval archived soils, taken from paired desert and agricultural experiments at five field stations of the Chinese Ecosystem Research Network across northern China, and performed high-throughput sequencing. Herein, we discovered that the clustering pattern of soil microbial communities was influenced by precipitation at some extent in desert ecosystem, while not impacted by climate factors in agricultural ecosystem. In addition, the analyses on microbial communities presented that the effects of climate factors on the communities decreased after agricultural practices. Soil microbial richness was significantly correlated with environmental temperature in deserts (R = -0.39, P0.001) and croplands (R = 0.34, P = 0.004), while the coefficients were opposite; the richness-precipitation relationship was significant in deserts (R = 0.63, P0.001) while nonsignificant in croplands (R = -0.03, P = 0.815). Moreover, for the dominant microbial groups (the top 10 phyla), the relationships between their richness and climate factors differed in two land use types, and fewer significant correlations were observed in croplands. In summary, it can be indicated that the influences of climate on soil microbial communities are shifted after intensive agriculture, and the relations of the richness with climate factors are also weakened for both the total and dominant microbial groups. These results improve our comprehension about the effects of climate on soil microbial diversity after intensive agriculture in desert areas, which can help to project microbial diversity in varied land uses under the context of global climate changes.

The profound effects of land-use types on soil microbial communities have been uncovered extensively in areas where land-use conversions are unidirectional. However, little is known about the response of soil microbes to land-use types in areas undergoing frequent bidirectional land-use conversions. Here, 36 paired cropland and grassland soils across soil types were collected in the agro-pastoral ecotone of northern China (APEN), and soil microbial communities were examined by high-throughput sequencing. According to the land-use history interpreted from remote sensing images from 2008 to 2017, the average frequency of conversion between cropland and grassland was 3 to 4 at all sites, and the duration of current land uses ranged from 1 to 9 years. Paired t-tests revealed that soil microbial diversity, measured as observed OTUs and Faith’s PD, did not differ significantly between cropland and grassland. In contrast, soil microbial community composition remarkably varied with the land-use type based on the nonmetric multidimensional scaling (NMDS) and analysis of similarity (ANOSIM). The above phenomena were also observed for different soil types (Kastanozems and Cambisols). The similarity percentage (SIMPER) analysis showed that the microbial taxa contributing most to the variation in the community composition were Pseudarthrobacter, Azotobacter, Bacillus, and Blastococcus, which are closely associated with the vegetation type (maize vs. grass). Meanwhile, there were no significant differences in soil properties between cropland and grassland, suggesting that vegetation types alone might drive changes in the community composition. Moreover, soil microbial diversity decreased with the duration of the croplands but did not change with the duration of the grasslands. The community composition between cropland and grassland differed more over a long duration (5 to 9 years) than over a short duration (1–4 years). Therefore, the longer the land-use type lasted, the more pronounced the differences in soil microbes between the two land-use types became. In conclusion, compared to soil microbial diversity, the community composition was more susceptible to land-use types under frequent bidirectional conversions, but both responses were enhanced with the duration of current land uses.

Hyper-spectral technology has been proven to be an effective method for the fast and non-destructive monitoring of crop biomass. However, the biomass estimation accuracy of this method is limited due to the effects of background factors, such as soils and water. In this study, a spectral separation method, non-negative matrix factorization (NMF), was proposed to alleviate the effects of soil on spectra. With the application of the NMF method, pure vegetation spectra were extracted from the field-observed spectra of wheat canopy, which were collected in four growing seasons from the tillering to the booting stages of wheat. Then, prediction models of wheat biomass (WB) were established and validated using the extracted spectra with the partial least squares regression (PLSR) method. The results showed that the NMF method could effectively separate the vegetation spectra from the mixed canopy spectra. Based on the extracted vegetation spectra, the WB prediction accuracy could be greatly improved with an increase of 31.7% for the R2p and an increase of 46.6% for the ratio of performance to deviation (RPD) as compared to the original spectra, indicating that the NMF method could significantly improve the performance of the WB prediction model. This method has potential application in the estimation of biomass using remote sensing technology.

Remote sensing of percentage residue cover (PRC) is required to manage crop residue and to provide a key parameter for several biogeochemistry models. However, the growth of the next crop, crop residue decomposition, and precipitation or irrigation can affect the spectral features of crop residue and decrease the spectral difference between crop residue and soil. These factors limit the time window of image acquisition for estimating PRC, especially for double cropping systems (two crops harvested in one year). Remote sensing sensors with a short revisit interval allow imagery to be captured for a specific period (e.g., soon after harvest) when these effects are minimal. SPOT imagery has a short revisit interval of two to three days because of its oblique viewing capacity, and provides balance between high resolution and wide-area coverage. The study aimed to investigate the potential of SPOT 5 imagery for predicting PRC in a typical area of China with a wheat (Triticum aestivum)–corn (Zea mays L.) double cropping system. A SPOT 5 image was acquired with three visible and near infrared bands and one shortwave infrared (SWIR) band, from which six normalized difference indices (NDI) were built. Wheat PRC values (n = 94) were measured immediately after wheat harvest using ground vertical photographs and divided into a calibration set (n = 62) and a validation set (n = 32). To estimate PRC, linear spectral analysis and three regression methods were used. Simple linear regression (SLR) was used for its simplicity, and stepwise regression (SR) and principal component regression (PCR) were used because they could exploit the spectral information from all available image bands. Besides, the bootstrap was used to assess the prediction uncertainty of the regression methods. The results demonstrated that the NDI41 index (derived from SWIR and green bands) performed the best among the four bands and the six indices for estimating PRC when using the SLR method, with a coefficient of determination (R2) of 0.588 and a root mean square error (RMSE) of 0.114 for the validation set. This estimation accuracy was similar to that observed for the SR and PCR models and much better than that for the linear spectral analysis (R2 and RMSE were 0.326 and 0.208, respectively). However, NDI41 accompanied by SLR was the optimal choice for estimating PRC because of its simplicity. These results indicate that SPOT 5 imagery could be used to estimate PRC. Therefore, this study should contribute to crop residue management.

An efficient and reliable method for detecting soil salinity in deep saline soils using remote sensing is required to better manage soil salinization. In this study, a laboratory evaporation experiment was conducted to determine how subsurface soil salinity can be determined using soil surface spectra. Ten soil columns were observed, and measurements were taken at specific time intervals to monitor the variations in soil salinity and moisture content at depths of 5 cm and 15 cm relative to soil surface spectra during the evaporation process. The structural equation modelling (SEM) method was used to analyse the relationships between the spectral reflectance data and the salt content in the subsurface soil (15 cm). The results showed significant correlations (standard path coefficient = −0.37, t = −6.00) between the spectral reflectance of the soil surface and the subsurface soil salinity and between different types of multispectral data, such as the Landsat Thematic Mapper (TM) data, and the subsurface soil salinity. Overall, the results implied that surface soil spectral information can be used to capture some subsurface soil properties and that remote sensing images could provide an alternative method for monitoring changes in deeper soils.

Visible and near-infrared spectroscopy (Vis-NIR) can accurately predict soil organic C (SOC) and clay from the spectra of air-dried and ground (DG) samples. However, grinding generally requires a lot of time and labor. In this study, air-dried and unground (DU) samples were used to exploit the time and accuracy advantages of Vis-NIR. The reflectance of 117 samples using DG and DU pretreatments was measured in the laboratory. Five spectral pretreatments were used: no-pretreatment (NP), Savitzky–Golay (SG) smoothing, first derivative (FD), standard normal variate transformation, and multiplicative scatter correction (MSC). When calibrations and validations used the same sample pretreatment (DG or DU), good predictions (R² > 0.75; RPD > 2.0) could be obtained for both SOC and clay. There were no sample × spectral pretreatments interaction effects on the prediction accuracy, whereas the main effects of the pretreatments were significant for clay but not for SOC. However, when calibrations derived from the DG samples were applied to the DU samples, the prediction accuracy decreased compared with that of the DG samples regardless of the SOC and clay. In this case, good predictions could be acquired for clay using the FD spectra and satisfactory predictions for SOC (0.60 < R² < 0.75; 1.4 < RPD < 2.0) were obtained from the NP, SG, and MSC spectra. These results indicate that it is viable to predict SOC and clay content using DU samples with acceptable accuracy while saving a lot of time and labor compared with that needed for DG samples.

Soil moisture reduces the accuracy of in situ measurements of soil properties using visible and near infrared reflectance spectroscopy and limits the application scope of existing soil spectral libraries built from air-dried samples. The preprocessing method of external parameter orthogonalisation (EPO) has been successfully used to improve the prediction of soil organic matter (SOM) in moist samples. However, the traditional strategy of EPO development (EPOI) requires a complex experimental design. In this study, we proposed a new EPO strategy (EPOII) that only uses a single sample but containing various soil moisture content ( SMC) levels. Reflectance spectra (350– 2500 nm) of 130 samples with different SMC levels were measured in the laboratory. We built calibration models using air-dried samples and partial least squares regression before and after EPOI and EPOII, which were validated using air-dried samples and moist samples, respectively. For the validation of moist samples, we first classified the SMC into four groups (A: all SMC levels; B: SMC < 0.1 g g−1; C: 0.1 < SMC < 0.2 g g−1; and D: SMC > 0.2 g g−1) and then used a Monte Carlo method to simulate SMC distributions in the field with a result of 500 cases for each SMC group. Before EPO, an apparent decrease in the accuracy of the SOM predictions for moist samples occurred in each SMC group when using the calibration model derived from air-dried samples, compared with that of air-dried samples. Both EPOI and EPOII improved the SOM prediction of moist samples for most scenarios of SMC variations (groups A–C), but did not apply to the scenario with both large SMC variation and high SMC (group D). When considering EPOII, it could be a feasible strategy of EPO implementation for removing moisture effects on SOM prediction using visible and near infrared spectroscopy, although the degree of improvement was less than that from EPOI. Also, EPOII will pave the way as a standard pretreatment especially for soil spectra measured in the field when projection matrices of EPOII have been derived and incorporated in spectroscopy-related software.

Timely and accurate monitoring of specific soil properties is essential. Visible and near-infrared (vis-NIR) reflectance spectroscopy has been suggested as a powerful tool to estimate soil properties, but the accuracy of soil property estimates can be strongly affected by soil moisture. External parameter orthogonalization (EPO) has been proposed to be an effective method to alleviate the effects of soil moisture on soil property estimation. The EPO algorithm decomposes the spectra into the useful (containing soil characteristics) part and the external part (affected by moisture). Most studies have focused on the useful part of spectra which contain soil information. However, it remains unclear whether the EPO process could also remove other useful soil information. Thus, the objectives of this study were to investigate the shape of the external part of the spectra, verify the relationship between soil moisture content and the external part, test whether other soil factors (e.g., soil texture) were removed by the EPO process, and propose a model for soil moisture using the spectra angle (EPO-SA) method. Soil samples with various soil moisture contents were analyzed. The results showed that the spectra of the external part, after the EPO process, were mainly correlated with the soil moisture content, with absolute correlation coefficients (|r|) greater than 0.8. These external spectra were not related to inherent soil properties, such as soil organic matter and clay content, with absolute correlation coefficients of less than 0.2. Furthermore, we proposed the EPO-SA method, which can be used to quantitatively predict the soil moisture content (R2 = 0.77, RMSE = 0.09 g kg−1). This approach clarified that useful soil spectral information was not removed by EPO, and it also proved the effectiveness of the EPO algorithm. Additionally, it is also demonstrated that the external spectra are useful for predicting the external factors since they are highly correlated with soil moisture.

Remote sensing has been widely applied to map soil salinity in the last few decades. However, a notable decrease in the accuracy of soil salt content (SSC) predictions occurred when the soil surfaces were partially vegetated. To minimize the influence of partial vegetation cover on spectral reflectance, we applied a spectral separation method, non-negative matrix factorization (NMF), to extract soil spectral information from a controlled field experiment with three varying factors [vegetation coverage, soil moisture content (SMC), and SSC]. The method was applied without prior knowledge of, or restrictions on the mixed and source spectra. Soil samples and spectral reflectance collected on three periods were used to determine the effectiveness of NMF-extracted soil spectra with partial least squares regression (PLSR). The results indicated that SSC can be predicted by bare soil spectra. NMF effectively separated soil spectra from the observed spectra, and the SSC was successfully predicted from the extracted soil spectra within a wide range of vegetation cover (0%-64.7%) within defined moisture levels (-1 by weight). The approach proposed in this study will improve the prediction accuracy of SSC for partially vegetated surfaces and will expand the application of remote sensing.

Land-use/land-cover changes (LUCCs) have links to both human and nature inter- actions. China's Land-Use/cover Datasets (CLUDs) were updated regularly at 5-year inter- vals from the late 1980s to 2010, with standard procedures based on Landsat TM\ETM+ im- ages. A land-use dynamic regionalization method was proposed to analyze major land-use conversions. The spatiotemporal characteristics, differences, and causes of land-use changes at a national scale were then examined. The main findings are summarized as fol- lows. Land-use changes (LUCs) across China indicated a significant variation in spatial and temporal characteristics in the last 20 years (1990-2010). The area of cropland change de- creased in the south and increased in the north, but the total area remained almost un- changed. The reclaimed cropland was shifted from the northeast to the northwest. The built-up lands expanded rapidly, were mainly distributed in the east, and gradually spread out to central and western China. Woodland decreased first, and then increased, but desert area was the opposite. Grassland continued decreasing. Different spatial patterns of LUC in China were found between the late 20th century and the early 21st century. The original 13 LUC zones were replaced by 15 units with changes of boundaries in some zones. The main spatial characteristics of these changes included (1) an accelerated expansion of built-up land in the

Multispectral remote sensing technique has been extensively applied in recent years for the detection of soil salinity on bare soil; however, multispectral remote sensing is restricted in areas covered with vegetation, largely due to the mixed pixel problem. In the present study, non-negative matrix factorization (NMF) was implemented to separate soil spectral signal from mixed pixels of Landsat 5 Thematic Mapper (TM) to further estimate the soil salinity in a partially vegetated area. Four methods, namely, partial least squares regression (PLSR), least-squares support vector machine (LS-SVM), back propagation neural network (BPNN), and random forest (RF), were applied and compared. The results showed that the NMF-separated soil spectra could greatly improve the prediction accuracy compared with the mixed spectra, and among the four modeling methods, RF performed better than the rest of the methods, with the averaged results of determination of the prediction R2p = 0.67, a root mean square error of the prediction RMSEp = 0.73 ms cm−1, and the ratio of the standard deviation to RMSEp RPD = 1.61 after 100 times of random sampling and modeling. This approach could propose a new method for accurate and timely monitoring of soil salinity in a partially vegetation-covered area.

The proposed method has 30 streams, i.e., 15 spatial streams and 15 temporal streams. Each spatial stream corresponds to each temporal stream. Therefore, this work correlates with the symmetry concept. It is a difficult task to classify video-based facial expression owing to the gap between the visual descriptors and the emotions. In order to bridge the gap, a new video descriptor for facial expression recognition is presented to aggregate spatial and temporal convolutional features across the entire extent of a video. The designed framework integrates a state-of-the-art 30 stream and has a trainable spatial&ndash, temporal feature aggregation layer. This framework is end-to-end trainable for video-based facial expression recognition. Thus, this framework can effectively avoid overfitting to the limited emotional video datasets, and the trainable strategy can learn to better represent an entire video. The different schemas for pooling spatial&ndash, temporal features are investigated, and the spatial and temporal streams are best aggregated by utilizing the proposed method. The extensive experiments on two public databases, BAUM-1s and eNTERFACE05, show that this framework has promising performance and outperforms the state-of-the-art strategies.

The estimation of soil organic carbon (SOC) stock in terrestrial ecosystems of China is of particular importance because it exerts a major influence on worldwide terrestrial carbon (C) storage and global climate change. Map-based estimates of SOC stocks conducted in previous studies have typically been applied on planimetric areas, which led to the underestimation of SOC stock. In the present study, SOC stock in China was estimated using a revised method on three-dimensional (3-D) surfaces, which considered the undulation of the landforms. Data were collected from the 1:4 M China Soil Map and a search work from the Second Soil Survey in China. Results indicated that the SOC stocks were 28.8 Pg C and 88.5 Pg C in soils at depths of 0–20 cm and 0–100 cm, corresponding to significant increases of 5.66% and 5.44%, respectively. Regression analysis revealed that the SOC stock accumulated with the increase of areas on 3-D surfaces. These results provide more reasonable estimates and new references about SOC stocks in terrestrial ecosystems of China. The method of estimation on 3-D surfaces has scientific meaning to promote the development of new approaches to estimate accurate SOC stocks.

The use of landscape covariates to estimate soil properties is not suitable for the areas of low relief due to the high variability of soil properties in similar topographic and vegetation conditions. A new method was implemented to map regional soil texture (in terms of sand, silt and clay contents) by hypothesizing that the change in the land surface diurnal temperature difference (DTD) is related to soil texture in case of a relatively homogeneous rainfall input. To examine this hypothesis, the DTDs from moderate resolution imagine spectroradiometer (MODIS) during a selected time period, i.e., after a heavy rainfall between autumn harvest and autumn sowing, were classified using fuzzy-c-means (FCM) clustering. Six classes were generated, and for each class, the sand (> 0.05 mm), silt (0.002–0.05 mm) and clay (< 0.002 mm) contents at the location of maximum membership value were considered as the typical values of that class. A weighted average model was then used to digitally map soil texture. The results showed that the predicted map quite accurately reflected the regional soil variation. A validation dataset produced estimates of error for the predicted maps of sand, silt and clay contents at root mean of squared error values of 8.4%, 7.8% and 2.3%, respectively, which is satisfactory in a practical context. This study thus provided a methodology that can help improve the accuracy and efficiency of soil texture mapping in plain areas using easily available data sources.

Spatial and temporal monitoring of soil properties in smelting regions requires collection of a large number of samples followed by laboratory cumbersome and time-consuming measurements. Visible and near-infrared diffuse reflectance spectroscopy (VNIR-DRS) provides a rapid and inexpensive tool to predict various soil properties simultaneously. This study evaluated the suitability of VNIR-DRS for predicting soil properties, including organic matter (OM), pH, and heavy metals (Cu, Pb, Zn, Cd, and Fe), using a total of 254 samples collected in soil profiles near a large copper smelter in China. Partial least square regression (PLSR) with cross-validation was used to relate soil property data to the reflectance spectral data by applying different preprocessing strategies. The performance of VNIR-DRS calibration models was evaluated using the coefficient of determination in cross-validation (R2cv) and the ratio of standard deviation to the root mean standard error of cross-validation (SD/RMSEcv). The models provided fairly accurate predictions for OM and Fe (R2cv > 0.80, SD/RMSEcv > 2.00), less accurate but acceptable for screening purposes for pH, Cu, Pb, and Cd (0.50 < R2cv < 0.80, 1.40 < SD/RMSEcv < 2.00), and poor accuracy for Zn (R2CV < 0.50, SD/RMSEcv < 1.40). Because soil properties in contaminated areas generally show large variation, a comparative large number of calibrating samples, which are variable enough and uniformly distributed, are necessary to create more accurate and robust VNIR-DRS calibration models. This study indicated that VNIR-DRS technique combined with continuously enriched soil spectral library could be a nondestructive alternative for soil environment monitoring.

Research has shown that the use of multitemporal images could obtain better classification over single date images and that seasonal variation of the normalized difference vegetation index (NDVI) could help improve classification accuracy. On consideration of different crop phenology and the seasonal changing characteristics of vegetation, water and bare soil indices from the Landsat Thematic Mapper (TM) images, the present paper uses multitemporal NDXI (NDVI, normalized difference water index (NDWI) and normalized difference soil index (NDSI)) to discriminate different landuse types in a rice planting area. From a comparison of the overall accuracy, Kappa coefficient and producer and user accuracies of different approaches, it is found that the NDXI approach is superior to the traditional classification that uses the original un-transformed images in the discrimination of different landuse types and agricultural land types in the rice planting area. The approach is expected to be more applicable to multitemporal Moderate Resolution Imaging Spectroradiometer (MODIS) images and to be used in discriminating different cropping systems in paddy areas.

This paper reports the dynamic changes of soil and water loss in the red soil region of Southern China since the 1950s. The red soil region covers eight provinces: Jiangxi, Zhejiang, Fujian, Anhui, Hubei, Hunan, Guangdong and Hainan. From the 1950s to 1986, the annual rate of soil erosion increased by 3.4%. From 1986 to 1996 and from 1996 to 2000, the annual rates of soil erosion decreased by 2.0% and 0.32%, respectively. Field surveys showed that from 2000 to 2005, the area of soil and water loss decreased annually by 1.2%. This decrease was a result of large-scale erosion control activities across China. Although the eroded soil has been restored, the restoration process is very slow and full restoration will take a long time. Our report suggests that controlling soil and water loss is a challenging task, and additional measures must be taken to effectively control the soil erosion in the red soil region.

Soil plays an important role in the global carbon cycle, and carbon sequestration in soil is important for mitigating global climate change. Historically, soil erosion led to great reductions of soil organic carbon (SOC) storage in China. Fortunately, with the economic development and remarkably effective soil erosion control measures in subtropical China over the past 20 years, soil erosion has been greatly decreased. As a result, soil organic carbon sequestration has gradually increased due to the rapid recovery of vegetation in the area. However, little information exists concerning the potential of soil carbon sequestration in the area. This paper introduces a case study in Xingguo County, Jiangxi Province, China, which used to be a typical area with significant soil loss in subtropical China. This work represents a systematic investigation of the interrelations of carbon sequestration potential with soil erosion types, altitudes, soil types and soil parent materials. In this study, 284 soil samples were collected from 151 sampling sites (51 are soil profile sites) to determine soil physicochemical properties including organic carbon content. Soil organic carbon distribution maps of the surface layer (0–20 cm) and whole profile (0–100 cm) were compiled by linking soil types to the polygons of digital soil maps using GIS. Assuming that SOC was lost following the destruction of native vegetation, these lands hold great promise for potentially sequestering carbon again. The potential of soil carbon sequestration in the study area was estimated by subtracting the organic carbon status in eroded soils from that in non-eroded soils under undisturbed forest. Results show that the potential of SOC in the surface layer is 4.47 Tg C while that in the whole profile is 12.3 Tg C for the entire county. The greatest potential for carbon sequestration (3.72 Tg C) is found in severely eroded soil, while non-eroded soil has the smallest potential. Also, soil carbon sequestration potential decreases with increasing altitude. Soils at altitudes of 800 m have the smallest potential (0.25 Tg C). Among various soil types, red earths (Humic Acrisols) have the greatest potential of carbon sequestration (5.32 Tg C), and yellow earths (Ferralic Cambisols) have the smallest (0.15 Tg C). As for soils derived from various parent materials, soils derived from phyllite possess the greatest carbon sequestration potential, and soils from Quaternary red clays have the smallest.

Accurate simulation of soil organic carbon (SOC) dynamics is vitally important in researching the carbon cycle in terrestrial ecosystems. Especially, the application of SOC model at the regional scale has major implications for regional and global carbon cycling. This paper addresses the regional simulation for SOC in the surface layer (0~15 cm) of paddy soils in Wuxi and Changzhou, Jiangsu Province, China from 1982~2000 by linking a model of agro-ecosystem SOC dynamic simulation with GIS data on the basis of 100 × 100 m grids. The dataset includes soil profile descriptions and chemical data of 187 paddy soil samples in the Second National Soil Survey of China in 1982 and 352 paddy soil samples taken in 2000. Soil maps at a scale of 1:50,000 were used for the simulation. Results showed that the area of differences between simulated and observed SOC in 2000 (in the range of −3.6 to 3.6 Mg C ha − 1) accounted for 50.6% of the total area. Also, when a 10% and 20% simulated increase, and a 10%, 20% and 30% simulated decrease in harvested crop straw and manure amendments were simulated to test the sensitivity of the model, simulated SOC would range from 10.26 to 50.40 Mg C ha − 1 with a standard deviation of 1.6 and 2.5, respectively. It is indicated that a key factor controlling SOC is the amount of residue and manure inputs. As for the assessment model suitability based on soil textures, soil types, and parent materials, the model seemed to perform better for clayey soils, groundwater-related paddy soils and soils derived from alluvial parent material than for loamy soils, surface water-related paddy soils and soil derived from loess material.

A new method to calculate the integrated index of soil erosion (EI) for the Red Soil Region of Southeast China is proposed. The median of erosion rates for the extreme severe erosion grade was calculated by using the method of polynomial fitting curve and others based on the medians and the corresponding grades of soil erosion in the Classification Standard of Soil Erosion (SL190-96); then the weight values were calculated for each erosion grade. Based on EI values in 1986, 1996, and 2002, the evolution of soil erosion was studied in a total of 8 provinces and in a total of 476 counties of southeastern China. The results show that the overall EI value of the whole Red Soil Region of Southeast China decreased from 3.81 to 3.02 from 1986 to 2002, although EI values increased in 163 counties, decreased in 285 counties, and remained stable in 28 counties, which reflects a general alleviation of soil erosion. The EI values decreased in Jiangxi, Fujian, and Zhejiang provinces, but increased in Guangdong and Hunan provinces. The increases are mainly attributed to rapid social and economic development with insufficient attention to soil erosion.

This study used GIS technology, remote sensing images of 1949, 1966, 1981, 1984, 1988, 1992, 1996 and 2000, to analyze the spatial process of urbanization and its impact on the soil resources in Yixing city, southeast China. The results showed that the urban area grew more than three-fold during 51 years and the loss of paddy soil accounted for 82.9% of all soil loss. According to the expansion dynamics and the results of spatial metrics, the urban expansion process in Yixing could be divided into four steps, the initial step (1949–1966), the almost standstill step (1966–1981), the preparation step (1981–1984) and the rapid growing step (after 1984). The study also indicated that the urban expansion was encouraged by the transportation system, however, restricted by the rivers, lakes and hills. The local government's decision also affected the spatial process of urban expansion. The spatial process of Yixing urban expansion might reflect the spatial process of urbanization of southeast China, the region with a rapidly growing economy.

Robust models for predicting soil salinity that use visible and near-infrared (vis-NIR) reflectance spectroscopy are needed to better quantify soil salinity in agricultural fields. Currently available models are not sufficiently robust for variable soil moisture contents. Thus, we used external parameter orthogonalization (EPO), which effectively projects spectra onto the subspace orthogonal to unwanted variation, to remove the variations caused by an external factor, e.g., the influences of soil moisture on spectral reflectance. In this study, 570 spectra between 380 and 2400 nm were obtained from soils with various soil moisture contents and salt concentrations in the laboratory; 3 soil types × 10 salt concentrations × 19 soil moisture levels were used. To examine the effectiveness of EPO, we compared the partial least squares regression (PLSR) results established from spectra with and without EPO correction. The EPO method effectively removed the effects of moisture, and the accuracy and robustness of the soil salt contents (SSCs) prediction model, which was built using the EPO-corrected spectra under various soil moisture conditions, were significantly improved relative to the spectra without EPO correction. This study contributes to the removal of soil moisture effects from soil salinity estimations when using vis-NIR reflectance spectroscopy and can assist others in quantifying soil salinity in the future.

This study was aimed to investigate the abundance and community shift of ammonia-oxidizing archaea (AOA) and bacteria (AOB) in air-dried forest soils in response to water addition, to explore the applicability of air-dried soil for microbial ecology study, and to elucidate whether AOA within the marine group 1. 1a dominate ammonia oxidizers communities in the acidic forest soils in China.Soil samples were collected from 10 forest sites of the China Ecosystem Research Network (CERN) and kept under air-drying conditions in 2010. In 2013 the air-dried soil samples were adjusted to 60% of soil maximum water holding capacity for a 28-day incubation at 28 degrees C in darkness. DGGE fingerprinting, clone library construction, pyrosequencing and quantitative PCR of amoA genes were performed to assess community change of ammonia oxidizers in air-dried and re-wetted soils.After incubation for 28 days, the abundance of bacteria and archaea increased significantly, up to 3,230 and 568 times, respectively. AOA increased significantly in 8 samples, and AOB increased significantly in 5 of 10 samples. However, pyrosequencing of amoA genes reveals insignificant changes in composition of AOA and AOB communities. Phylogenetic analysis of amoA genes indicates that archaeal ammonia oxidizers were predominated by AOA within the soil group 1. 1b lineage, while the Nitrosospira-like AOB dominate bacteria ammonia oxidizer communities. There was a significantly positive correlation between AOA/AOB ratio and total nitrogen (r2 = 0.54, P0.05), implying that soil ammonia oxidation might be dominated by AOA in association with ammonium released from soil mineralization.Phylogenetic analysis suggest that AOA members within the soil group 1. 1b lineage were not restricted to non-acidic soils as previously thought. The abundance rather than composition of AOA and AOB changed in response to water addition. This indicates that air-dried soil could be of help for microbial biogeography study.