Search

High-resolution remote sensing image-based vegetation monitoring is a hot topic in remote sensing technology and applications. However, when facing large-scale monitoring across different sensors in broad areas, the current methods suffer from fragmentation and weak generalization capabilities. To address this issue, this paper proposes a multisource high-resolution remote sensing image-based vegetation extraction method that considers the comprehensive perception of multiple features. First, this method utilizes a random forest model to perform feature selection for the vegetation index, selecting an index that enhances the otherness between vegetation and other land features. Based on this, a multifeature synthesis perception convolutional network (MSCIN) is constructed, which enhances the extraction of multiscale feature information, global information interaction, and feature cross-fusion. The MSCIN network simultaneously constructs dual-branch parallel networks for spectral features and vegetation index features, strengthening multiscale feature extraction while reducing the loss of detailed features by simplifying the dense connection module. Furthermore, to facilitate global information interaction between the original spectral information and vegetation index features, a dual-path multihead cross-attention fusion module is designed. This module enhances the differentiation of vegetation from other land features and improves the network's generalization performance, enabling vegetation extraction from multisource high-resolution remote sensing data. To validate the effectiveness of this method, we randomly selected six test areas within Anhui Province and compared the results with three different data sources and other typical methods (NDVI, RFC, OCBDL, and HRNet). The results demonstrate that the MSCIN method proposed in this paper, under the premise of using only GF2 satellite images as samples, exhibits robust accuracy in extraction results across different sensors. It overcomes the rapid degradation of accuracy observed in other methods with various sensors and addresses issues such as internal fragmentation, false positives, and false negatives caused by sample generalization and image diversity. [ABSTRACT FROM AUTHOR]

The complexity of surface characteristics in rural areas poses challenges for accurate extraction of built-up areas from remote sensing images. The Artificial Surface Index (ASI) emerged as a novel and accurate built-up land index. However, the absence of short-wave infrared (SWIR) bands in most high-resolution (HR) images restricts the application of index-based methods in rural built-up land extraction. This paper presents a rapid extraction method for high-resolution built-up land in rural areas based on ASI. Through the downscaling techniques of random forest (RF) regression, high-resolution SWIR bands were generated. They were then combined with visible and near-infrared (VNIR) bands to compute ASI on GaoFen-2 (GF-2) images (called A S I G F ). Furthermore, a red roof index (RRI) was designed to reduce the probability of misclassifying built-up land with bare soil. The results demonstrated that SWIR downscaling effectively compensates for multispectral information absence in HR imagery and expands the applicability of index-based methods to HR remote sensing data. Compared with five other indices (UI, BFLEI, NDBI, BCI, and PISI), the combination of ASI and RRI achieved the optimal performance in built-up land enhancement and bare land suppression, particularly showcasing superior performance in rural built-up land extraction. [ABSTRACT FROM AUTHOR]

This study investigates the application of hyperspectral image space–spectral fusion technology in lithologic classification, using data from China's GF-5 and Europe's Sentinel-2A. The research focuses on the southern region of Tuanjie Peak in the Western Kunlun Range, comparing five space–spectral fusion methods: GSA, SFIM, CNMF, HySure, and NonRegSRNet. To comprehensively evaluate the effectiveness and applicability of these fusion methods, the study conducts a comprehensive assessment from three aspects: evaluation of fusion effects, lithologic classification experiments, and field validation. In the evaluation of fusion effects, the study uses an index analysis and comparison of spectral curves before and after fusion, concluding that the GSA fusion method performs the best. For lithologic classification, the Random Forest (RF) classification method is used, training with both area and point samples. The classification results from area sample training show significantly higher overall accuracy compared to point samples, aligning well with 1:50,000 scale geological maps. In field validation, the study employs on-site verification combined with microscopic identification and comparison of images with actual spectral fusion, finding that the classification results for the five lithologies are essentially consistent with field validation results. The "GSA+RF" method combination established in this paper, based on data from GF-5 and Sentinel-2A satellites, can provide technical support for lithological classification in similar high-altitude regions. [ABSTRACT FROM AUTHOR]

Since there has been concern about food security, accurate prediction of wheat yield prior to harvest is a key component. Random Forest (RF) has been used in many classification and regression applications, such as yield estimation, and the performance of RF has improved by tuning its hyperparameters. In this paper, different changes are made to traditional RF for yield estimation, and the performance of RF is evaluated. Accordingly, RFs constructed using various weak learners, as well as a combined RF consisting of different weak learners are assessed by growing weak Gaussian Process Regression (GPR), Decision Tree (DT), Neural Network (NN), and Stepwise Regression (SW) models in the forest. The input data to DTs are also partitioned into leaves in N (e.g., two) dimensional feature space by using clustering in each parent node. In addition, a subset of the training set is randomly sampled with replacement for training a learner in the forest, instead of randomly sampling the whole training set in traditional RF. Using clustering in DTs added flexibility while utilizing NN as a weak learner yielded the most favorable outcomes in our research. The number of input training samples (Itree) to each tree was also identified as a new hyperparameter to the forest, and the prediction results were more influenced by the Itree compared to the known hyperparameters, such as the number of trees in the forest (ntree) and the number of features for each tree (mtry). [ABSTRACT FROM AUTHOR]

Efficiently obtaining tobacco planting area is significant for rationally allocating tobacco resources and realizing the balance between supply and demand. However, tobacco fields are characterized by fragmentation in hilly areas, which brings various challenges to extracting the tobacco planting area. A 16-square-kilometre tobacco planting region in Xuancheng City, Anhui Province, China, was selected as the study area in this paper. First, the single-temporal full-feature sets (STFFS), a time-series full-feature set (TFFS) and a time-series optimal-feature set (TOFS) were constructed from multi-period Sentinel-2 images, respectively. Then, we applied Random Forest (RF), Support Vector Machine (SVM), Neural Network Classification (NNC) and Maximum Likelihood Classification (MLC) to classify the feature sets and compare the classification accuracy. The experimental results demonstrate that: (1) The best growth stage of tobacco remote sensing recognition is the mulching film phase of the spherical plant stage (ST1 period). (2) The classification accuracy indicates that TOFS outperforms the STFFS. (3) TOFS can still maintain the same classification accuracy as TFFS with fewer feature dimensions. (4) The RF classifier has great stability and robustness, which achieve an overall accuracy (OA) of 90.65%, 93.50%, and 93.09% based on the feature sets of ST1 period, TFFS, and TOFS, respectively. [ABSTRACT FROM AUTHOR]

Processing of hyperspectral remote sensing datasets poses challenges in terms of computational expense pertaining to data redundancy. As such, band selection becomes indispensable to address redundancy while preserving the optimal spectral information. This paper proposes a novel architecture using Genetic Algorithm (GA) optimizing technique with Random Forest (RF) classifier for efficient band selection with the Hyperspectral Precursor of the Application Mission (PRISMA) dataset. The optimal bands are BLUE (λ = 492.69 nm), NIR (λ = 959.52 nm), and SWIR 1 (λ = 1626.78 nm). This paper also involves an application of the selected bands to accurately identify and quantify built-up pixels by means of a new spectral index named Hyperspectral Imagery-based Built-up Index (HIBI). The proposed index was used to map built-up pixels in six cities around the world namely Jaipur, Varanasi, Delhi, Tokyo, Moscow and Jakarta to establish its robustness. This analysis shows that the proposed index has an accuracy of 94.02%, higher than all the other indices considered for this study. Moreover, the spectral separability analysis also establishes the efficiency of the proposed index to differentiate built-up pixels from spectrally similar land use or land cover classes. [ABSTRACT FROM AUTHOR]

Selecting training samples is crucial in remote sensing image classification. In this paper, we selected three images—Sentinel-2, GF-1, and Landsat 8—and employed three methods for selecting training samples: grouping selection, entropy-based selection, and direct selection. We then used the selected training samples to train three supervised classification models—random forest (RF), support-vector machine (SVM), and k-nearest neighbor (KNN)—and evaluated the classification results of the three images. According to the experimental results, the three classification models performed similarly. Compared with the entropy-based method, the grouping selection method achieved higher classification accuracy using fewer samples. In addition, the grouping selection method outperformed the direct selection method with the same number of samples. Therefore, the grouping selection method performed the best. When using the grouping selection method, the image classification accuracy increased with the increase in the number of samples within a certain sample size range. [ABSTRACT FROM AUTHOR]

The emergency response is crucial in preventing and mitigating urban floods. Both remote sensing and social media data offer distinct advantages in large-scale flood monitoring and near-real-time flood monitoring. However, current research lacks a thorough exploration of the application of social media data and remote sensing imagery in the urban flood emergency response. To address this issue, this paper, while extracting disaster information based on social media data, deeply mines the spatiotemporal distribution characteristics and dynamic spatial accessibility of rescue points. Furthermore, SAR imagery and social media data for monitoring urban flooding are compared. This study took the Zhengzhou 7.20 urban flood as a case study and created a methodological framework to quickly extract flood disaster information (flood, landslide, and rescue points) using these two types of data; spatiotemporal analysis and random forest classification were also conducted to mine valuable information. Temporally, the study revealed that disaster information did not increase proportionally with the amount of rainfall during the rainfall process. Spatially, specific regions with higher susceptibility to flooding, landslides, and rescue points were identified, such as the central region characterized by low drainage standards and high-density urban areas, as well as the eastern region with low-lying terrain. Moreover, this study examined the spatial accessibility of rescue resources in real flood scenarios and found that their service coverage varied throughout the day during and after the disaster. In addition, social media excelled in high-density urban areas' flood point extraction, while SAR performed better in monitoring floods at the edges of low-density urban areas and large water bodies, allowing them to complement each other, to a certain extent. The findings of this study provide scientific reference value for the optimal selection of rescue paths and the allocation of resources in the emergency response to urban floods caused by extreme rainstorms. [ABSTRACT FROM AUTHOR]

Northeast China plays a pivotal role in producing commodity grains. The precipitation and temperature distribution during the growth season is impacted by geographical and climate factors, rendering the region vulnerable to drought. However, relying on a single index does not reflect the severity and extent of drought in different regions. This research utilised the random forest (RF) model for screening remote sensing indices. Relative soil moisture (RSM) was employed to compare seven commonly used indices: the temperature vegetation dryness index (TVDI), vegetation supply water index (VSWI), vegetation condition index (VCI), temperature condition index (TCI), vegetation health index (VHI), multi-band drought index (MBDI), and normalised difference drought index (NDDI). The effectiveness of these indices for monitoring drought during different developmental stages of spring maize was evaluated. Trend rates were employed to investigate the temporal changes in drought patterns of spring maize from 2003 to 2020, and the Sen + Mann–Kendall test was used to analyse spatial variations. The results showed the following: (1) The seven remote sensing indices could accurately track drought during critical growth stages with the TVDI demonstrating higher applicability than the other six indices. (2) The application periods of two TVDIs with different parameters differed for the drought monitoring of spring maize in different developmental periods. The consistency and accuracy of the normalised difference vegetation index (NDVI)-based TVDI (TVDIN) were 5.77% and 34.62% higher than those of the enhanced vegetation index (EVI)-based TVDI (TVDIE), respectively, in the early stage. In contrast, the TVDIE exhibited 13.46% higher consistency than the TVDIN in the middle stage, and the accuracy was the same. During the later stage, the TVDIE showed significantly higher consistency and accuracy than the TVDIN with consistency increases of 9.61% and 38.64%, respectively. (3) The drought trend in northeast China increased from 2003 to 2020, exhibiting severe spring drought and a weakening of the drought in summer. The southern, southwestern, and northwestern parts of northeast China showed an upward drought trend; the drought-affected areas accounted for 37.91% of the study area. This paper identified the most suitable remote sensing indices for monitoring drought in different developmental stages of spring maize. The results provide a comprehensive understanding of the spatial–temporal patterns of drought during the past 18 years. These findings can be used to develop a dynamic agricultural drought monitoring model to ensure food security. [ABSTRACT FROM AUTHOR]

As one of the central cities in the Wuling Mountains, Huaihua is regarded as the main body and core economic development area in the process of Wuling's future new urbanization, exerting a pivotal influence on the five provincial border areas (Hunan, Hubei, Chongqing, Guizhou and Guangxi). But studies on urban expansion in this city are few, so this paper makes efforts to simulate its urban land expansion by constructing the Cellular Automata model (the Weight-CA model) taking into account differences in land types conversion. Firstly, the paper obtains the time-series Landsat spectral bands (Red, NIR) and NDVI images using the time–space fusion model; then the optimal combination of classification features is screened based on J Bh distance and the object-oriented random forest algorithm is applied for land cover classification; the paper finally constructs the Weight-CA model for the simulation and validation of urban land expansion in Huaihua. Study results in the paper have shown that the Landsat spectral bands obtained with the ESTARFM have a relatively high correlation with real Landsat images. The land cover classification based on the optimal classification feature combination and the object-oriented random forest algorithm is high in accuracy, the total of which exceeding 84 % over 4 years. Meanwhile, based on the Weight-CA model, the simulation accuracy of urban land expansion in Huaihua in 2010 and 2015 reached 89.98 % and 89.60 % compared with the real values. It is operationally ready and practical and of strong practical significance to extract land cover types by Landsat and MODIS data fusion, and simulate urban land expansion using the Weight-CA method. [ABSTRACT FROM AUTHOR]

The evaluation of satellite soil moisture is a big challenge owing to the large spatial mismatch between pixel-based satellite soil moisture products and point-based in situ measurements. Upscaling in situ measurements to obtain the "true value" of soil moisture content at the satellite grid/footprint scale can make up for the scale difference and improve the validation. Many existing upscaling methods have strict requirements regarding the spatial distribution and quantity of soil moisture sensors. However, in reality, soil-moisture-monitoring networks are commonly sparse with low sensor density, which increases the difficulty of obtaining accurate upscaled soil moisture data and limits the validation of satellite products. For this reason, this paper proposes a scheme to upscale in situ measurements using five machine learning methods along with Landsat 8 datasets and DEM data to validate the accuracy of a SMAP-enhanced passive soil moisture product for a sparse network on the Qinghai–Tibet Plateau. The proposed scheme realizes the upscaling of in situ soil moisture data to the pixel scale (30 m × 30 m) and then to the coarse grid scale (9 km × 9 km) by using multi-source remote sensing data as the bridge of scale conversion. The long-time SMAP SM products since April 2015 on the Qinghai–Tibet Plateau were validated based on upscaled soil moisture data. The results show that (1) random forest regression performs the best, and the upscaled soil moisture data reflect the region-average soil moisture conditions that can be used for evaluating SMAP data; (2) the SMAP product meets its scientific measurement requirements; and (3) the SMAP product generally underestimates the soil moisture in the study area. [ABSTRACT FROM AUTHOR]

Land cover (LC) classification remains a challenging task due to the diversity of terrain and topography, limited prior knowledge, and complex data sets. These variations can lead to differences in illumination and shading, which can affect the appearance of objects in the image. The local binary pattern (LBP) model is an effective technique for capturing local texture information in an image, which can help to overcome the effects of topography diversity by analysing patterns in different image regions, even if the illumination or shading conditions are different. However, LBP alone is inadequate for characterizing high-resolution remotely sensed images with complex semantic content as it only utilizes sign information in the local region. In this paper, a new texture characterization descriptor, known as completed homogeneous LBP (CHLBP), is proposed as an improved version of homogeneous LBP (HLBP) for LC classification of remotely sensed images. The CHLBP method mainly involves the following steps: first, sign and magnitude information from the HLBP descriptor is extracted, providing an effective alternative to the HLBP complementary contrast measure. Second, for each sign and magnitude function, a new splitting factor δ is used to obtain a depth relationship between the centre and its neighbouring pixels and to enhance noise robustness. Finally, the centre pixels representing the image grey level are also considered to contain discriminative information. The performance of our descriptor is evaluated using four challenging texture databases: Outex (TC10, TC12), Geofen Image Dataset (GID), and large-scale aerial (AID). Extensive experiments were performed using four classifiers: Support Vector Machine (SVM), K-Nearest Neighbours (KNN), Random Forest (RF), and Multi-Layer Perceptron (MLP), demonstrating the effectiveness and robustness of our descriptor against noise and free noise conditions in public remote sensing datasets. [ABSTRACT FROM AUTHOR]

In this paper, based on the Landsat multispectral remote sensing images of 1999, 2008 and 2019 in the oasis area of the Taolai River Basin, a remote sensing image classification method based on ENVI deep learning was constructed to extract and identify the cover information of oasis area on the basis of establishing classification system, interpretation flags and sample data sets, and compared with the classification methods based on backpropagation neural network (BPNN), support vector machine regression (SVM) and random forest (RF). The results show that the overall accuracy of the classification method based on ENVI deep learning is 97.34 %, and the Kappa coefficient is 0.96; Under the same number of samples, compared with the classification method based on BPNN, SVM and RF, the classification method based on ENVI deep learning constructed in this study improves the overall accuracy by 6.80%, 2.04% and 3.03%, and the Kappa coefficient increases by 0.12, 0.07 and 0.09, respectively, and the classification method is the best for extracting surface cover information fin oasis area. This study can provide technical support for rapid and accurate extraction and identification of ground cover information. [ABSTRACT FROM AUTHOR]

Mountain areas in China account for 69% of the total land area; however, it is still an urgent that we grasp the special ecological structure of mountain areas and maximize the resource advantages of mountain areas under the principle of maintaining a certain ecological level. In this paper, Landsat 5, Landsat 8 and Sentinel-2A images were used as data sources to monitor and analyze land development and ecological change in Gui 'an in 2010, 2013, 2016, 2019 and 2021, so as to explore ecological maintenance mechanisms. Firstly, random forest classification based on multi-source remote sensing data was used to classify land, and the five phases of land-use change were assessed using quantitative analysis, in order to analyze the mountain region's land-use-change characteristics at different stages of development. The results show that Gui 'an has the "two-stage", rapid-development, rapid-recovery mode. Each stage includes a development-and-expansion period and a construction-and-protection period. In the construction period, ecological recovery construction will be emphasized, and the change intensity and rate of the second stage are lower than that of the first stage. Secondly, using a remote sensing ecological index, vegetation coverage, and a landscape index, an ecological evaluation model of the study area was constructed to analyze changes in the ecological environment and its protection in the process of land development. The ecological maintenance status of the five stages was quantitatively monitored using the analysis methods of difference change and coefficient of variation. The results showed that in the first stage of land development and expansion, the landscape pattern and ecological quality fluctuated greatly, and the proportion of ecological quality of an excellent grade decreased by 28.46%. However, in the second stage, the change slowed down and remained unchanged, and gradually moved to the middle and upper level. The results show that there is a close relationship between ecological maintenance and the land development mode, and new mountain towns can maintain ecological quality and achieve sustainable development through a reasonable land development mode. At the end of this paper, the factors affecting the ecological maintenance capacity of Gui 'an are discussed, providing effective reference material and development models for the development of mountainous areas. [ABSTRACT FROM AUTHOR]

Impervious surface as an evaluation indicator of urbanization is crucial for urban planning and management. It is necessary to obtain impervious surface information with high accuracy and resolution to meet dynamic monitoring under rapid urban development. At present, the methods of impervious surface extraction are primarily based on medium-low-resolution images. Therefore, it is of theoretical and application value to construct an impervious surface extraction method that applies to high-resolution satellite images and can solve the shadow misclassification problem. This paper builds an impervious surface extraction model by Bayes discriminant analysis (BDA). The Gaussian prior model is incorporated into the Bayes discriminant analysis to establish a new impervious surface extraction model (GBDA) applicable to high-resolution remote sensing images. Using GF-2 and Sentinel-2 remote sensing images as experimental data, we discuss and analyze the applicability of BDA and GBDA in impervious surface extraction of high-resolution remote sensing images. The results showed that the four methods, SVM, RF, BDA and GBDA, had OA values of 91.26%, 94.91%, 94.64% and 97.84% and Kappa values of 0.825, 0.898, 0.893 and 0.957, respectively, in the extraction results of GF-2. In the results of effective Sentinel-2 extraction, the OA values of the four methods were 87.94%, 91.79%, 92.19% and 93.51% and the Kappa values were 0.759, 0.836, 0.844 and 0.870, respectively. Compared with the support vector machine (SVM), random forest (RF) and BDA methods, GBDA has significantly improved the extraction accuracy. GBDA enhances the robustness and generalization ability of the model and can improve the shadow misclassification phenomenon of high-resolution images. The model constructed in this paper is highly reliable for extracting impervious surfaces from high-resolution remote sensing images, exploring the application value of Bayes discriminant analysis in impervious surface extraction and providing technical support for impervious surface information of high spatial resolution and high quality. [ABSTRACT FROM AUTHOR]

This paper presents an in-depth study and analysis of the assessment of English teaching ability using the algorithm of fuzzy mean-shift clustering. The paper proposes an automatic scoring model for Chinese-English sentence-level interpretation based on semantic scoring. The automatic scoring model calculates candidates' Chinese-English sentence interpretation scores by fusing the feature parameters at both the phonological and content levels. Adaptive weights are introduced to fuse the current pixel and the neighborhood mean with adaptive weighting, and the weighted entropy constraint term is embedded in the clustering objective function to solve the selection problem of the weighting parameters. Finally, the graphical fuzzy division information of the neighboring pixels is used to construct the local spatial information constraint term of the current pixel, and the graphical fuzzy division term of the current pixel is adjusted to correct the clustering center obtained from the iteration. Fluency is selected as the feature scoring parameter at the speech level, and the automatic scoring model directly scores the fluency features of the candidates' recordings; two feature scoring parameters, keywords, and sentence semantics are selected at the content level, and the content features are scored after converting the candidates' recordings to text by manual conversion. The zero-energy product method is used to extract speech features to calculate fluency feature scores; the semantic scoring model introduced in this paper is used to calculate keyword and sentence semantic feature scores; finally, the random forest algorithm is used to fuse the above three feature scoring parameters to obtain the total quality score of Chinese-English interpretation. Considering the correlation of neighborhood pixel affiliation, the KL scatter and affiliation space information is used to supervise the current pixel affiliation, to further improve the segmentation accuracy of the algorithm; finally, segmentation tests are conducted on synthetic images, medical images, and remote sensing images. The results show that the proposed algorithm has a stronger noise suppression ability and can obtain more satisfactory segmentation results than other robust fuzzy clustering algorithms. [ABSTRACT FROM AUTHOR]

The continuous transformation from biodiverse natural forests and mixed-use farms into monoculture rubber plantations may lead to a series of hazards, such as natural forest habitats fragmentation, biodiversity loss, as well as drought and water shortage. Therefore, understanding the spatial distribution of rubber plantations is crucial to regional ecological security and a sustainable economy. However, the spectral characteristics of rubber tree is easily mixed with other vegetation such as natural forests, tea plantations, orchards and shrubs, which brings difficulty and uncertainty to regional scale identification. In this paper, we proposed a classification method combines multi-source phenology characteristics and random forest algorithm. On the basis of optimization of input samples and features, phenological spectrum, brightness, greenness, wetness, fractional vegetation cover, topography and other features of rubber were extracted. Five classification schemes were constructed for comparison, and the one with the highest classification accuracy was used to identify the spatial pattern of rubber plantations in 2014, 2016, 2018 and 2020 in Xishuangbanna. The results show that: (1) the identification results are in consistent with field survey and rubber plantations area generally shows a first increasing and then decreasing trend; (2) the Overall Accuracy (OA) and Kappa coefficient of the proposed method are 90.0% and 0.86, respectively, with a Producer's Accuracy (PA) and User's Accuracy (UA) of 95.2% and 88.8%, respectively; (3) cross-validation was employed to analyze the accuracy evaluation indexes of the identification results: both PA and UA of the rubber plantations stay stable over 85%, with the minimum fluctuation and best stability of UA value. The OA value and Kappa coefficient were stable in the range of 0.88–0.90 and 0.84–0.86, respectively. The method proposed provides reliable results on spatial distribution of rubber, and is potentially transferable to other mountainous areas as a robust approach for rapid monitoring of rubber plantations. [ABSTRACT FROM AUTHOR]

The extraction of large dam candidate regions is critical for broad-scale efforts to rapidly detect large-area dams. The framework proposed in this paper attempts to combine random forest classification models and spatial analysis methods with large dam candidate area extraction methods for large-scale areas. First, we studied the combination of optical, microwave, texture, and topographic features of the dam and constructed a multisource remote sensing and topographic feature vector of the dam. Secondly, we constructed random forest classifiers in different study areas and evaluate their performance. Then we explored the geographic characteristics of the dams and their relationships with other features. Finally, we introduced the spatial analysis method to constrain the large dam candidate area. The proposed framework was tested in a total area of 968,533 km2 in five countries and achieved promising results, which constrained the candidate area to less than 1.06% of the total area. We calculated the completeness rate of large dams using the multi-source dam datasets. The framework achieved a completeness rate of more than 97.62%. Our results show that the entire framework is reliable for automated and fast large dam candidate area acquisition based on data from open remote sensing products. [ABSTRACT FROM AUTHOR]

Grasslands are crucial ecosystems that support and provide a diverse number of ecosystem services. Sown biodiverse pastures rich in legumes (SBP) were developed with the main goal of increasing grassland production while minimizing fertilizers inputs. In this paper, the main properties of SBP in Portugal were estimated using remote sensing and machine learning in six different farms and two production years (spring 2018 and 2019). Four pasture characteristics were considered: aboveground standing biomass, fraction of legumes, plant nitrogen (N) content and plant phosphorus (P) content. Remote sensing data were obtained from Sentinel-2. The spectral bands combined with 5 vegetation indices and 9 covariates were used. Multiple linear regression, LASSO, Ridge, random forests, XGBoost and LightGBM regression models were used. Two cross-validation approaches were used: (1) a random approach with random selection of the folds (RN-CV), and (2) a structured approach where each fold is a unique combination of farm and year, which is subsequently used to assess the performance of the model obtained with the 8 other folds (LLYO-CV). Results showed that the random forest method had the best estimation accuracy for all pasture characteristics. Regarding cross-validation approaches, the algorithms with RN-CV have higher estimation accuracy for all pasture characteristics (on average about 10% lower RMSE and an R2 85% higher), as compared to the algorithms with LLYO-CV. However, LLYO-CV should avoid overfitting and improve generalization of the models because in each fold the model is tested in a farm and year that was not used for training. The RMSE for all variables were significantly low, especially in RN-CV. Plant P is the variable where the choice of CV approach has the least influence (RMSE of test set with RN-CV: 0.71 g P kg− 1; LLYO-CV: 0.72 g P kg− 1). Standing biomass is the variable with the highest difference between CV approaches (RN-CV: 722 kg ha− 1; LLYO-CV: 825 kg ha− 1). The RMSE, of legumes and plant N were moderately affected by the CV approach (legume RN-CV: 0.11; LLYO-CV: 0.12 – plant N RN-CV: 3.96 g N kg− 1; LLYO-CV: 3.99 g N kg− 1). The algorithms developed here were applied for entire parcels in the two farms with the most different climate conditions as demonstration of their potential future use for precision farming. [ABSTRACT FROM AUTHOR]

Affected by geographical location and climatic conditions, crop classification in the Yunnan Plateau of China is greatly restricted by the low utilization rate of annual optical data, complex crop planting structure, and broken cultivated land. This paper combines monthly Sentinel-2 optical remote sensing data with Sentinel-1 radar data to minimize cloud interference to conduct crop classification for plateau areas. However, pixel classification will inevitably produce a "different spectrum of the same object, foreign objects in the same spectrum". A principal component feature synthesis method is developed for multi-source remote sensing data (PCA-MR) to improve classification accuracy. In order to compare and analyze the classification effect of PCA-MR combined with multi-source remote sensing data, we constructed 11 classification scenarios using the Google Earth Engine platform and random forest algorithm (RF). The results show that: (1) the classification accuracy is 79.98% by using Sentinel-1 data and 91.18% when using Sentinel-2 data. When integrating Sentinel-1 and Sentinel-2 data, the accuracy is 92.31%. By analyzing the influence of texture features on classification under different feature combinations, it was found that optical texture features affected the recognition accuracy of rice to a lesser extent. (2) The errors will be reduced if the PCA-MR feature is involved in the classification, and the classification accuracy and Kappa coefficient are improved to 93.47% and 0.92, respectively. [ABSTRACT FROM AUTHOR]

Forest logging detection is important for sustainable forest management. The traditional optical satellite images with visible and near-infrared bands showed the ability to identify intensive timber logging. However, less intensive logging is still difficult to detect with coarse spatial resolution such as Landsat or high spatial resolution in fewer spectral bands. Although more high-resolution remote sensing images containing richer spectral bands can be easily obtained nowadays, the questions of whether they facilitate the detection of logging patterns and which spectral bands are more effective in detecting logging patterns, especially in selective logging, remain unresolved. Therefore, this paper aims to evaluate the combinations of visible, near-infrared, red-edge, and short-wave infrared bands in detecting three different logging intensity patterns, including unlogged (control check, CK), selective logging (SL), and clear-cutting (CC), in north subtropical plantation forests with the random forest algorithm using Sentinel-2 multispectral imagery. This study aims to explore the recognition performance of different combinations of spectral bands (visual (VIS) and near-infrared bands (NIR), VIS, NIR combined with red-edge, VIS, NIR combined with short-wave infrared bands (SWIR), and full-spectrum bands combined with VIS, NIR, red edge and SWIR) and to determine the best spectral variables to be used for identifying logging patterns, especially in SL. The study was conducted in Taizishan in Hubei province, China. A total of 213 subcompartments of different logging patterns were collected and the random forest algorithm was used to classify logging patterns. The results showed that full-spectrum bands which contain the red-edge and short-wave infrared bands improve the ability of conventional optical satellites to monitor forest logging patterns and can achieve an overall accuracy of 85%, especially for SL which can achieve 79% and 64% for precision and recall accuracy, respectively. The red-edge band (698–713 nm, B5 in Sentinel-2), short-wave infrared band (2100–2280 nm, B12 in Sentinel-2), and associated vegetation indices (NBR, NDre2, and NDre1) enhance the sensitivity of the spectral information to logging patterns, especially for the SL pattern, and the precision and recall accuracy can improve by 10% and 6%, respectively. Meanwhile, both clear-cutting and unlogged patterns could be well-classified whether adding a red-edge or SWIR band or both in VIS and NIR bands; the best precision and recall accuracies for clear-cutting were enhanced to 97%, 95% and 81%, 91% for unlogged, respectively. Our results demonstrate that the optical images have the potential ability to detect logging patterns especially for the clear-cutting and unlogged patterns, and the selective logging detection accuracy can be improved by adding red-edge and short-wave infrared spectral bands. [ABSTRACT FROM AUTHOR]

Urban forests maintain the ecological balance of cities and are significant in promoting the sustainable development of cities. Therefore, using advanced remote sensing technology to accurately extract forest green space in the city and monitor its change in real-time is very important. Taking Nanjing as the study area, this research extracted 55 vegetation phenological features from Sentinel-2A time series images and formed a feature set containing 81 parameters together with 26 features, including polarimetric- and texture-related information extracted from dual-polarization Sentinel-1A data. On the basis of the improved ABC (ABC-LIBSVM) feature selection method, the optimal feature subset was selected, and the forest coverage areas in the study area were accurately described. To verify the feasibility of the improved feature selection method and explore the potential for the development of multi-source time series remote sensing for urban forest feature extraction, this paper also used the random forest classification model to classify four different feature sets. The results revealed that the classification accuracy based on the feature set obtained by the ABC-LIBSVM algorithm was the highest, with an overall accuracy of 86.80% and a kappa coefficient of 0.8145. The producer accuracy and user accuracy of the urban forest were 93.21% and 82.45%, respectively. Furthermore, by combining the multi-source time series Sentinel-2A optical images with Sentinel-1A dual-polarization SAR images, urban forests can be distinguished from the perspective of phenology, and polarimetric- and texture-related features can contribute to the accurate identification of forests. [ABSTRACT FROM AUTHOR]

The class imbalance problem has been reported to exist in remote sensing and hinders the classification performance of many machine learning algorithms. Several technologies, such as data sampling methods, feature selection-based methods, and ensemble-based methods, have been proposed to solve the class imbalance problem. However, these methods suffer from the loss of useful information or from artificial noise, or result in overfitting. A novel double ensemble algorithm is proposed to deal with the multi-class imbalance problem of the hyperspectral image in this paper. This method first computes the feature importance values of the hyperspectral data via an ensemble model, then produces several balanced data sets based on oversampling and builds a number of classifiers. Finally, the classification results of these diversity classifiers are combined according to a specific ensemble rule. In the experiment, different data-handling methods and classification methods including random undersampling (RUS), random oversampling (ROS), Adaboost, Bagging, and random forest are compared with the proposed double random forest method. The experimental results on three imbalanced hyperspectral data sets demonstrate the effectiveness of the proposed algorithm. [ABSTRACT FROM AUTHOR]

Remote sensing offers a scalable and low cost solution for the production of large-scale crop maps, which can be used to extract relevant crop statistics. However, despite considerable advances in the new generation of satellite sensors and the advent of cloud computing, the use of remote sensing for the production of accurate crop maps and statistics remain dependant on the availability of ground truth data. Such data are necessary for the training of supervised classification algorithms and for the validation of the results. Unfortunately, in-situ data of adequate quality for producing crop statistics are seldom available in many countries. In this paper we compare the performance of two supervised classifiers, the Random Forest (RF) and the Dynamic Time Warping (DTW), the former being a data intensive algorithm and the latter a more data frugal one, in extracting accurate crop type maps from EO and in-situ data. The two classifiers are trained several times using datasets which contain in turn an increasing number in-situ samples gathered in the Kashkadarya region of Uzbekistan in 2018. We finally compare the accuracy of the maps produced by the RF and the DTW classifiers with respect to the different number of training data used. Results show that when using only 5 and 10 training samples per each crop class, the DTW reaches a higher Overall Accuracy than the RF. Only when using five times more training samples, the RF starts to perform slightly better that the DTW. We conclude that the DTW can be used to map crop types using EO data in countries where limited in/situ data are available. We also highlight the critical importance in the choice of the location of the in-situ data and its thematic reliability for the accuracy of the final map, especially when using the DTW. [ABSTRACT FROM AUTHOR]

This paper aims to offer an improved version of the new composite drought monitoring index (CDMI) and to test its applicability in the context of Tensift watershed in Morocco. A synergistic approach incorporating the remote sensing techniques, hydrometeorological data, simulated data and agricultural statistics was used for this purpose. After assessing the performance of CDMI, estimated Soil Moisture Anomaly Indicator (ISMA) was processed, validated and incorporated to the composite model. Random Forest algorithm was used to determine the weight of composite model components. Apart from comparative mapping, Pearson's correlation statistical analysis, linear regression and dependency tests were used to assess the performance of the improved composite model (CDMIa_RF). The result show that CDMIa_RF is better correlated with several indices such as: the Standardized Precipitation Index (SPI), (R2=0.74); Hydrological Drought Index (R2=0.70); grain productivity (R2=0.70), CDMI (R2=0.95), Vegetation Health Index (VHI), (R2=0.87), and Normalized Vegetation Supply Water Index (NVSWI), (R2= 0.85). [ABSTRACT FROM AUTHOR]

Fire susceptibility modeling is crucial for sustaining and managing forests among many other valuable land resources. With 56% of its area covered by forests, Arkansas is known as the "natural state". About 1000 wildfires occurred and burned more than 10,000 acres each year during 1981–2018. In this paper, we use remote-sensing-based machine learning methods to address the natural and anthropogenic factors influencing wildfires and model fire susceptibility in Arkansas. Among the 15 explored variables, potential evapotranspiration, soil moisture, Palmer drought severity index, and dry season precipitation were recognized as the most significant factors contributing to the fire density. The obtained R-squared values are significant, with 0.99 for training the model and 0.92 for the validation. The results show that the Ouachita National Forest and the Ozark Forest, in west-central and west Arkansas, respectively, have the highest susceptibility to wildfires. The southern part of Arkansas has low-to-moderate fire susceptibility, while the eastern part of the state has the lowest fire susceptibility. These new results for Arkansas demonstrate the potency of remote-sensing-based random forest in predicting fire susceptibility at the state level that can be adapted to study fires in other states and help with fire preparedness to reduce loss and save the precious environment. [ABSTRACT FROM AUTHOR]

Surface longwave net radiation (SLNR) is the key component of the surface radiation budget. SLNR is estimated by using the conventional method, in which surface longwave downward radiation (SLDR) and surface longwave upward radiation (SLUR) are estimated separately, and then SLNR is calculated by subtracting SLUR from SLDR. However, uncertainties in SLDR and SLUR may accumulate and propagate into the calculated SLNR, affecting the SLNR estimate accuracy. This paper proposes a direct algorithm that links clear-sky SLNR with the top-of-atmosphere (TOA) radiances in MODIS channels 29, 31 and 32 and column water vapour (CWV). The multivariate regression (MVR) and extreme gradient boosting (×GBoost) were employed to establish the relationships sequentially with collocated ground measurements and satellite data for high-latitude, mid-latitude, and low-latitude areas. According to the training and validation results, the GBoost model performs slightly better than the MVR model, which indicates that weak nonlinearity exists between clear-sky SLNR and its predictors. The absolute biases of the XGBoost models are approximately 1.0 Wm−2, and the RMSEs of the XGBoost model are 17.77, 27.33, and 19.60 Wm−2 in high-latitude, mid-latitude, and low-latitude areas. These results indicate that nonlinear relationships exist between clear-sky SLNR and predictors, but this nonlinearity is not strong. The superiority of the direct algorithm over the conventional method was also demonstrated. In the future, we will collect more data to improve the performance of the proposed direct algorithm especially in high-latitude and mid-latitude areas. [ABSTRACT FROM AUTHOR]

The mining industry production is an important pillar industry in China, while its extensive production activities have led to several ecological and environmental problems. Earth observation technology using high-resolution satellite imagery can help us efficiently obtain information on surface elements, surveying and monitoring various land occupation issues arising from open-pit mining production activities. Conventional pixel-based interpretation methods for high-resolution remote sensing images are restricted by "salt and pepper" noise caused by environmental factors, making it difficult to meet increasing requirements for monitoring accuracy. With the Jingxiang phosphorus mining area in Jingmen Hubei Province as the studied area, this paper uses a multi-scale segmentation algorithm to extract large-scale main characteristic information using a layered mask method based on the hierarchical structure of the image object. The remaining characteristic elements were classified and extracted in combination with the random forest model and characteristic factors to obtain land occupation information related mining industry production, which was compared with the results of the Classification and Regression Tree model. 23 characteristic factors in three aspects were selected, including spectral, geometric and texture characteristics. The methods employed in this study achieved 86% and 0.78 respectively in overall extraction accuracy analysis and the Kappa coefficient analysis, compared to 79% and 0.68 using the conventional method. [ABSTRACT FROM AUTHOR]

Combining optical and polarimetric synthetic aperture radar (PolSAR) earth observations offers a complementary data set with a significant number of spectral, textural, and polarimetric features for crop mapping and monitoring. Moreover, a temporal combination of both sources of information may lead to obtaining more reliable results compared to the use of single-time observations. In this paper, an operational framework based on the stacked generalization of random forest (RF), which efficiently employed bi-temporal observations of optical and radar data, was proposed for crop mapping. In the first step, various spectral, vegetation index, textural, and polarimetric features were extracted from both data sources and placed into several groups. Each group was classified separately using a single RF classifier. Then, several additional classification tasks were accomplished by another RF classifier. The earth observations used in this paper were collected by RapidEye satellites and the Unmanned Aerial Vehicle Synthetic Aperture Radar (UAVSAR) system over an agricultural region near Winnipeg, Manitoba, Canada. The results confirmed that the proposed methodology was able to provide a higher overall accuracy and kappa coefficient than traditional stacking method, and also than all the individual RFs using each group. These accuracy metrics were also better than those of the RFs using the stacked features. Moreover, only the proposed methodology could achieve standard accuracy (F-score ≥85%) for all crop types in the study area. The visual comparison also demonstrated that the crop maps produced by the proposed methodology had more homogeneous, uniform appearances. Moreover, the mixed pixels of crop types, which abundantly existed in the traditional stacking and individual RFs̕ maps, were significantly eliminated. [ABSTRACT FROM AUTHOR]

Airborne dust has broad adverse effects on human activity, including aviation, human health, and agriculture. Remote sensing observations are used to detect dust and aerosols in the atmosphere using long established techniques. False color Red‐Green‐Blue (RGB) imagery using band differences sensitive to dust absorption (Dust RGB) is currently used operationally to assist forecasters and decision‐makers in identifying dust at night, but there are still limitations, subjectivity, and nuances to image interpretation making night‐time dust identification difficult even for experts. This study applies machine learning to the problem of night‐time dust detection with a simple random forest (RF) model using Geostationary Operational Environmental Satellite‐16 (GOES‐16) Advanced Baseline Imager (ABI) infrared imagery, band differences sensitive to dust absorption, and Dust RGB color components as inputs to the model. The RF model achieves an Area‐Under‐Curve (AUC) of 0.97 with a standard deviation of 0.04 for dust cases. For images with dust present, the model correctly labels 85% of dust pixels and 99.96% of no‐dust pixels for all dust images in the validation data set. The addition of a single null case to the training data set drastically reduces error in labeling no‐dust pixels as dust from 45% to 14.5%. Application of the machine learning model to the April 13–14, 2019 dust event demonstrates the ability of the model to identify dust during night‐time hours when visual dust detection is limited by the cooling ground surface characteristics. Plain Language Summary: Airborne dust adversely affects many facets of human activity, including aviation, human health, and agriculture. While many observational systems can detect airborne dust during the day, nighttime observations are limited, making airborne dust difficult to identify and track at night. This paper combines satellite infrared imagery with machine learning to enable improved dust detection at night when visual dust detection is hindered. The relatively simple model correctly labels 85% of dust pixels in the satellite imagery, supporting operational forecasters in airborne dust identification and hazard detection at night. Key Points: Detection of hazardous dust events in nighttime satellite imagery is limited as dust is difficult to distinguish from the cooling groundA physically based approach to developing a machine learning model with satellite imagery inputs correctly labels 85% of dust pixelsApplication of the machine learning model to a dust event enhances identification of dust, distinguishing the boundaries of the plume [ABSTRACT FROM AUTHOR]

Using remote sensing data to monitor large area drought is one of the important methods of drought monitoring at present. However, the traditional remote sensing drought monitoring methods mainly focus on monitoring single drought response factors such as soil moisture or vegetation status, and the research on comprehensive multi-factor drought monitoring is limited. In order to improve the ability to resist drought events, this paper takes Henan Province of China as an example, takes multi-source remote sensing data as data sources, considers various disaster-causing factors, adopts random forest method to model, and explores the method of regional remote sensing comprehensive drought monitoring using various remote sensing data sources. Compared with neural network, classification regression tree and linear regression, the performance of random forest is more stable and tolerant to noise and outliers. In order to provide a new method for comprehensive assessment of regional drought, a comprehensive drought monitoring model was established based on multi-source remote sensing data, which comprehensively considered the drought factors such as soil water stress, vegetation growth status and meteorological precipitation profit and loss in the process of drought occurrence and development. [ABSTRACT FROM AUTHOR]

There is growing global consensus that assessments of land degradation be conducted at regional or smaller scales. Working at this scale allows for locally relevant environmental and land use conditions to be incorporated into the assessment methodology. In this paper, a recently developed regional approach to assessing land degradation in the Hardeveld bioregion of the Succulent Karoo is applied to the Little Karoo region of this biome. The methodology uses fuzzy classification statistical techniques to combine field data with multiple Sentinel-2A and Landsat vegetation indices, as well as regionally modelled soil variables. The resultant habitat condition archetype map values show strong correlation with field observations of perennial plant and bare soil cover in 96 ground-truthed plots. The archetype map indicates that heavily degraded hotspots of high bare ground cover occur throughout the project region, although there is an overall lower average habitat condition in the western half of the Little Karoo. The mean habitat condition archetype value for the entire project area is 0.54 (standard deviation = 0.13), on a continuous scale where 0 and 1 represent the most degraded and pristine extremes, respectively. Random forest regression analysis of various environmental covariates of degradation indicates a strong relationship between habitat condition and topographic as well as rainfall variables, although the limited accuracy of modelled livestock data may obscure the negative impacts of overgrazing. The 30 m resolution habitat condition archetype map builds upon previous degradation research in the Little Karoo and has the potential to inform future conservation, restoration, and rangeland management decisions. The methodology was successfully transferred to a new region and provides an opportunity to improve reporting on the extent of land degradation across South Africa. • Effective transfer of methodology for regional scale assessment of degradation. • Western Little Karoo has lower habitat condition score than the eastern part. • Topography, rainfall, and land use history influence habitat condition score. • 30 × 30 m habitat condition map can inform sustainable land management decisions. [ABSTRACT FROM AUTHOR]

Ocean satellites provide accurate and precise data across various scales, making them a vital tool for investigating the association between global change and ocean processes. However, low-quality data creates unavoidable gaps in satellite data, diminishing its usefulness and continuity. These deficiencies can be resolved by implementing machine learning techniques as valuable tools. This paper details a new satellite data prediction tool titled "SatelliteFixer". The SatelliteFixer model, utilizing a custom-built random forest structure, can generate dependable data with enhanced temporal-spatial continuity. This model has demonstrated feasibility with diverse satellite data sources and light bands, and outperforms the basic machine learning approach. The juxtaposition of model data with in-situ cruise sampling results allows for widespread analysis of the movement and dispersion of suspended sediment. The above entails the inversion of long-term events and the observation of short-term events, which enables accurate seasonal analysis using continuous data without the influence of uneven data volume distribution and outliers, and is also the first-time satellite data has tracked the entire process of pulsed artificial flooding. SatelliteFixer provides a fresh outlook for detailing the varying trends on consecutive timescales and successional spaces among ocean processes. • A data-driven model for ocean satellite missing data prediction is established based on a specially designed machine learning structure. • The feasibility and availability of model result are demonstrated with MODIS Aqua and GOCI visible light data. • As an example of application, SatelliteFixer data are used with in-situ cruise data to analyze the transportation and distribution of suspended sediment. • SatelliteFixer data enables the accurate seasonal analysis and the first tracking of the entire process on pulsed artificial flooding. [ABSTRACT FROM AUTHOR]

The Global Positioning System (GPS) is not only widely used in navigation, measurement and other services, but also an indispensable key equipment for the military. With the increasing complexity of the communication environment and the increasing number of interference factors, the recognition of GPS interference signal types is a prerequisite for the development of efficient anti-interference means. This paper focuses on three typical GPS interference signals, by extracting four different entropy features including power spectral entropy, establishing a hybrid entropy dataset and then using support vector machine (SVM) and random forest (RF) methods so as to classify and identify the dataset. The results show that the RF has a high recognition rate for the interference signal, and the average accuracy is above 90%, which greatly exceeds the SVM. Also, in the three kinds of interference signals, the noise FM interference is the least concealed and the most easily recognized. [ABSTRACT FROM AUTHOR]

Cloud detection is a necessary phase in satellite images processing to retrieve the atmospheric and lithospheric parameters. Currently, some cloud detection methods based on Random Forest (RF) model have been proposed but they do not consider both spectral and textural characteristics of the image. Furthermore, they have not been tested in the presence of snow/ice. In this paper, we introduce two RF based algorithms, Feature Level Fusion Random Forest (FLFRF) and Decision Level Fusion Random Forest (DLFRF) to incorporate visible, infrared (IR) and thermal spectral and textural features (FLFRF) including Gray Level Co-occurrence Matrix (GLCM) and Robust Extended Local Binary Pattern (RELBP_CI) or visible, IR and thermal classifiers (DLFRF) for highly accurate cloud detection on remote sensing images. FLFRF first fuses visible, IR and thermal features. Thereafter, it uses the RF model to classify pixels to cloud, snow/ice and background or thick cloud, thin cloud and background. DLFRF considers visible, IR and thermal features (both spectral and textural) separately and inserts each set of features to RF model. Then, it holds vote matrix of each run of the model. Finally, it fuses the classifiers using the majority vote method. To demonstrate the effectiveness of the proposed algorithms, 10 Terra MODIS and 15 Landsat 8 OLI/TIRS images with different spatial resolutions are used in this paper. Quantitative analyses are based on manually selected ground truth data. Results show that after adding RELBP_CI to input feature set cloud detection accuracy improves. Also, the average cloud kappa values of FLFRF and DLFRF on MODIS images (1 and 0.99) are higher than other machine learning methods, Linear Discriminate Analysis (LDA), Classification And Regression Tree (CART), K Nearest Neighbor (KNN) and Support Vector Machine (SVM) (0.96). The average snow/ice kappa values of FLFRF and DLFRF on MODIS images (1 and 0.85) are higher than other traditional methods. The quantitative values on Landsat 8 images show similar trend. Consequently, while SVM and K-nearest neighbor show overestimation in predicting cloud and snow/ice pixels, our Random Forest (RF) based models can achieve higher cloud, snow/ice kappa values on MODIS and thin cloud, thick cloud and snow/ice kappa values on Landsat 8 images. Our algorithms predict both thin and thick cloud on Landsat 8 images while the existing cloud detection algorithm, Fmask cannot discriminate them. Compared to the state-of-the-art methods, our algorithms have acquired higher average cloud and snow/ice kappa values for different spatial resolutions. [ABSTRACT FROM AUTHOR]

Drought is an important factor that limits economic and social development due to its frequent occurrence and profound influence. Therefore, it is of great significance to make accurate predictions of drought for earlywarning and disaster alleviation. In this paper, SPEI-1was confirmed to classify drought grades in the Guanzhong Area, and the autoregressive integrated moving average (ARIMA), random forest (RF) and support vectormachine (SVM)modelwere established. Meteorological data and remote sensing datawere used to derive the predictionmodels. The results showed the following. (1) The SVM model performed the best when themodelswere developed using meteorological data, remote sensing data and a combination of meteorological and remote sensing data, but the model's corresponding kernel functions are different and include linear, polynomial and Gaussian radial basis kernel functions, respectively. (2) The RF model driven by the remote sensing data and the SVMmodel driven by the combinedmeteorological and remote sensing data were found to perform better than the model driven by the corresponding other data in the Guanzhong Area. It is difficult to accurately measure drought with the singlemeteorological data. Only by considering the combined factors can we more accurately monitor and predict drought. This study can provide an important scientific basis for regional drought warnings and predictions. [ABSTRACT FROM AUTHOR]

Remote sensing scene classification is gaining much more interest in the recent few years for many strategic fields such as security, land cover and land use monitoring. Several methods have been proposed in the literature and they can be divided into three main classes based on the features used: handcrafted features, features obtained by unsupervised learning and those obtained from deep learning. Handcrafted features are generally time consuming and suboptimal. Unsupervised learning based features which have been proposed later gave better results but their performances are still limited because they mainly rely on shallow networks and are not able to extract powerful features. Deep learning based features are recently investigated and gave interesting results. But, they cannot be usually used because of the scarcity of labelled remote sensing images and are also computationally expensive. Most importantly, whatever kind of feature is used, the neighbourhood information of them is ignored. In this paper, we propose a novel remote sensing scene representation and classification approach called Bag of Visual SubGraphs (BoVSG). First, each image is segmented into superpixels in order to summarize the image content while retaining relevant information. Then, the superpixels from all images are clustered according to their colour and texture features and a random label is assigned to each cluster that probably corresponds to some material or land cover type. Thus superpixels belonging to the same cluster have the same label. Afterwards, each image is modelled with a graph where nodes correspond to labelled superpixels and edges model spatial neighbourhoods. Finally, each image is represented by a histogram of the most frequent subgraphs corresponding to land cover adjacency patterns. This way, local spatial relations between the nodes are also taken into account. Resultant feature vectors are classified using standard classification algorithms. The proposed approach is tested on three popular datasets and its performance outperforms state-of-the-art methods, including deep learning methods. Besides its accuracy, the proposed approach is computationally much less expensive than deep learning methods. [ABSTRACT FROM AUTHOR]

Cloud detection of remote sensing imagery is quite challenging due to the influence of complicated underlying surfaces and the variety of cloud types. Currently, most of the methods mainly rely on prior knowledge to extract features artificially for cloud detection. However, these features may not be able to accurately represent the cloud characteristics under complex environment. In this paper, we adopt an innovative model named Fuzzy Autoencode Model (FAEM) to integrate the feature learning ability of stacked autoencode networks and the detection ability of fuzzy function for highly accurate cloud detection on remote sensing imagery. Our proposed method begins by selecting and fusing spectral, texture, and structure information. Thereafter, the proposed technique established a FAEM to learn the deep discriminative features from a great deal of selected information. Finally, the learned features are mapped to the corresponding cloud density map with a fuzzy function. To demonstrate the effectiveness of the proposed method, 172 Landsat ETM+ images and 25 GF-1 images with different spatial resolutions are used in this paper. For the convenience of accuracy assessment, ground truth data are manually outlined. Results show that the average RER (ratio of right rate and error rate) on Landsat images is greater than 29, while the average RER of Support Vector Machine (SVM) is 21.8 and Random Forest (RF) is 23. The results on GF-1 images exhibit similar performance as Landsat images with the average RER of 25.9, which is much higher than the results of SVM and RF. Compared to traditional methods, our technique has attained higher average cloud detection accuracy for either different spatial resolutions or various land surfaces. [ABSTRACT FROM AUTHOR]

Peatlands are important reserves of terrestrial carbon and biodiversity, and given that many peatlands across the UK and Europe exist in a degraded state, their conservation is a major area of concern and a focus of considerable research. Aerial surveys are valuable tools for habitat mapping and conservation and provide useful insights into their condition. We investigate how SfM photogrammetry-derived topography and habitat classes may be used to construct an estimate of carbon loss from erosion features in a remote blanket bog habitat. An autonomous, unmanned, aerial, fixed-wing remote sensing platform (Quest UAV 300™) collected imagery over Moor House, in the Upper Teesdale National Nature Reserve, a site with a high degree of peatland erosion. The images were used to generate point clouds into orthomosaics and digital surface models using SfM photogrammetry techniques, georeferenced and subsequently used to classify vegetation and peatland features. A classification of peatbog feature types was developed using a random forest classification model trained on field survey data and applied to UAV-captured products including the orthomosaic, digital surface model and derived surfaces such as topographic index, slope and aspect maps. Using the area classified as eroded peat and the derived digital surface model, we estimated a loss of 438 tonnes of carbon from a single gully. The UAV system was relatively straightforward to deploy in such a remote and unimproved area. SfM photogrammetry, imagery and random forest modelling obtained classification accuracies of between 42% and 100%, and was able to discern between bare peat, saturated bog and sphagnum habitats. This paper shows what can be achieved with low-cost UAVs equipped with consumer grade camera equipment and relatively straightforward ground control, and demonstrates their potential for the carbon and peatland conservation research community. [ABSTRACT FROM AUTHOR]

• A method to evaluate the quality of pansharpening remote sensing images by combining deep-level features and low-level features. • A new preprocessing method is proposed in the low-level features extraction process, and a new preprocessing method is also proposed in the deep-level features extraction process, both of which can be extended to high-dimensional images. • For both types of features extraction is performed in the fused image, without involving any information other than the fused image. Pansharpening is an important fusion method in remote sensing image processing, aiming at obtaining a multispectral image with high spatial resolution by fusing a multispectral image of low spatial resolution with a panchromatic image of high spatial resolution. A persistent problem with this approach concerns how the quality of the fused images obtained by different methods can be assessed. We address this by constructing a full-resolution quality assessment method of pansharpened remote sensing image based on the combination of low-level features and deep-level features. For the former, different distribution parameters are estimated based on three types of image features prior. On the other hand, deep-level features are designed by employing network. Finally, the quality scores are obtained by combining the two types of features with random forest for regression. The experimental results showed that, compared with the state-of-the-art methods, the FCBM method in this paper can improve the PLCC metric by 16.2%, the SROCC metric by 25%, and the RMSE metric by 42.9%. [ABSTRACT FROM AUTHOR]

[Display omitted] • The ecological maintenance was studied based on ecological environment quality. • The impacts of factors on RSEI were studied based on geographical detectors. • We utilized the minimum cumulative resistance to construct ecological security. • The study can contribute to global ecological security and sustainable development. The world's ecological security has been significantly threatened because of the rapid growth of urbanization. Recently, the worldwide studies of ecological maintenance and ecological security have become of crucial scientific value as it ensures the balance between human needs and resources. In this paper, the ecological maintenance status of Pingtan Island, Songming County (China), and Yoshkar-Ola (Russia) was studied from the perspective of ecological environment quality and ecological security pattern. The remote sensing-based ecological index (RSEI) was used to study the ecological environment quality, and the geographic detector was used to analyze the impact of six factors on RSEI. The random forest algorithm was used to classify land use from remote sensing images, and an ecological security pattern was constructed based on the minimum cumulative resistance (MCR) model. The results show that the overall ecological maintenance of Pingtan Island is moderate, which accounts for 32.46% of the total area. The explanatory power of the dryness factor, which is the most influential factor on ecological environmental quality, increased by 3.90% during the five years, and the ecological corridor increased by 8.67 km. Songming County accounted for 33.66% of the area with moderate ecological maintenance. The dryness factor which affects the ecological environment quality is changed into the greenness factor, and its explanatory power has decreased by 3.59%, and the ecological corridor has increased by 57.95 km. Yoshkar-Ola accounted for 28.47% of the area with moderate ecological maintenance. The explanatory power of the dryness factor, which is the biggest factor affecting ecological environment quality, increased by 2.69% during the five-year period, and the length of the ecological corridor increased by 10.64 km. In this paper, remote sensing and GIS methods were used to integrate the ecological security status of inland and coastal areas and different natural zones and to conduct a comparative study on the ecological maintenance status of the three case studies, which can provide decision-making support for regional planning and contribute to global ecological security and sustainable development. [ABSTRACT FROM AUTHOR]

Maps that categorise the landscape into discrete units are a cornerstone of many scientific, management and conservation activities. The accuracy of these maps is often the primary piece of information used to make decisions about the mapping process or judge the quality of the final map. Variance is critical information when considering map accuracy, yet commonly reported accuracy metrics often do not provide that information. Various resampling frameworks have been proposed and shown to reconcile this issue, but have had limited uptake. In this paper, we compare the traditional approach of a single split of data into a training set (for classification) and test set (for accuracy assessment), to a resampling framework where the classification and accuracy assessment are repeated many times. Using a relatively simple vegetation mapping example and two common classifiers (maximum likelihood and random forest), we compare variance in mapped area estimates and accuracy assessment metrics (overall accuracy, kappa, user, producer, entropy, purity, quantity/allocation disagreement). Input field data points were repeatedly split into training and test sets via bootstrapping, Monte Carlo cross-validation (67:33 and 80:20 split ratios) and k- fold (5-fold) cross-validation. Additionally, within the cross-validation, we tested four designs: simple random, block hold-out, stratification by class, and stratification by both class and space. A classification was performed for every split of every methodological combination (100’s iterations each), creating sampling distributions for the mapped area of each class and the accuracy metrics. We found that regardless of resampling design, a single split of data into training and test sets results in a large variance in estimates of accuracy and mapped area. In the worst case, overall accuracy varied between ~40–80% in one resampling design, due only to random variation in partitioning into training and test sets. On the other hand, we found that all resampling procedures provided accurate estimates of error, and that they can also provide confidence intervals that are informative about the performance and uncertainty of the classifier. Importantly, we show that these confidence intervals commonly encompassed the magnitudes of increase or decrease in accuracy that are often cited in literature as justification for methodological or sampling design choices. We also show how a resampling approach enables generation of spatially continuous maps of classification uncertainty. Based on our results, we make recommendations about which resampling design to use and how it could be implemented. We also provide a fully worked mapping example, which includes traditional inference of uncertainty from the error matrix and provides examples for presenting the final map and its accuracy. [ABSTRACT FROM AUTHOR]

Although past work has noted that contrasts in turbidity often are detectable on remotely sensed images of rivers downstream from confluences, no systematic methodology has been developed for assessing mixing over distance of confluent flows with differing surficial suspended sediment concentrations (SSSC). In contrast to field measurements of mixing below confluences, satellite remote-sensing can provide detailed information on spatial distributions of SSSC over long distances. This paper presents a methodology that uses remote-sensing data to estimate spatial patterns of SSSC downstream of confluences along large rivers and to determine changes in the amount of mixing over distance from confluences. The method develops a calibrated Random Forest (RF) model by relating training SSSC data from river gaging stations to derived spectral indices for the pixels corresponding to gaging-station locations. The calibrated model is then used to predict SSSC values for every river pixel in a remotely sensed image, which provides the basis for mapping of spatial variability in SSSCs along the river. The pixel data are used to estimate average surficial values of SSSC at cross sections spaced uniformly along the river. Based on the cross-section data, a mixing metric is computed for each cross section. The spatial pattern of change in this metric over distance can be used to define rates and length scales of surficial mixing of suspended sediment downstream of a confluence. This type of information is useful for exploring the potential influence of various controlling factors on mixing downstream of confluences, for evaluating how mixing in a river system varies over time and space, and for determining how these variations influence water quality and ecological conditions along the river. [ABSTRACT FROM AUTHOR]

Plastic mulching is an important technology in agricultural production both in China and the rest of the world. In spite of its benefit of increasing crop yields, the booming expansion of the plastic mulching area has been changing the landscape patterns and affecting the environment. Accurate and effective mapping of Plastic-Mulched Farmland (PMF) can provide useful information for leveraging its advantages and disadvantages. However, mapping the PMF with remote sensing is still challenging owing to its varying spectral characteristics with the crop growth and geographic spatial division. In this paper, we investigated the potential of Radarsat-2 data for mapping PMF.We obtained the backscattering intensity of different polarizations and multiple polarimetric decomposition descriptors. These remotely-sensed information was used as input features for Random Forest (RF) and Support Vector Machine (SVM) classifiers. The results indicated that the features from Radarsat-2 data have great potential for mapping PMF. The overall accuracies of PMF mapping with Radarsat-2 data were close to 75%. Although the classification accuracy with the back-scattering intensity information alone was relatively lower owing to the inherent speckle noise in SAR data, it has been improved significantly by introducing the polarimetric decomposition descriptors. The accuracy was nearly 75%. In addition, the features derived from the Entropy/Anisotropy/Alpha (H/A/Alpha) polarimetric decomposition, such as Alpha, entropy, and so on, made a greater contribution to PMF mapping than the Freeman decomposition, Krogager decomposition and the Yamaguchi4 decomposition. The performances of different classifiers were also compared. In this study, the RF classifier performed better than the SVM classifier. However, it is expected that the classification accuracy of PMF with SAR remote sensing data can be improved by combining SAR remote sensing data with optical remote sensing data. [ABSTRACT FROM AUTHOR]

This paper describes a workflow utilizing detailed canopy height information derived from digital airphotos combined with ground inventory information gathered in state-owned forests and regression modelling techniques to quantify forest-growing stocks in private woodlands, for which little information is generally available. Random forest models were trained to predict three different variables at the plot level: quadratic mean diameter of the 100 largest trees (d100), basal area weighted mean height of the 100 largest trees (h100), and gross volume (V). Two separate models were created – one for a spruce- and one for a beech-dominated test site. We examined the spatial portability of the models by using them to predict the aforementioned variables at actual inventory plots in nearby forests, in which simultaneous ground sampling took place. When data from the full set of available plots were used for training, the predictions ford100,h100, andVachieved out-of-bag model accuracies (scaled RMSEs) of 15.1%, 10.1%, and 35.3% for the spruce- and 15.9%, 9.7%, and 32.1% for the beech-dominated forest, respectively. The corresponding independent RMSEs for the nearby forests were 15.2%, 10.5%, and 33.6% for the spruce- and 15.5%, 8.9%, and 33.7% for the beech-dominated test site, respectively. [ABSTRACT FROM PUBLISHER]

Land cover information is necessary for a large range of environmental applications related to climate impacts and adaption, emergency response, wildlife habitat, etc. In Canada, a 2008 user survey indicated that the most practical land cover data is provided in a nationwide 30 m spatial resolution format, with an update frequency of five years. In response to this need, the Canada Centre for Remote Sensing (CCRS) has generated a 30 m land cover map of Canada for the base year 2010, as the first of a planned series of maps to be updated every five years, or more frequently. This land cover dataset is also the Canadian contribution to the 30 m spatial resolution 2010 Land Cover Map of North America, which is produced by Mexican, American and Canadian government institutions under a collaboration called the North American Land Change Monitoring System (NALCMS). This paper describes the mapping approach used for generating this land cover dataset for Canada from Thematic Mapper (TM) and Enhanced Thematic Mapper (ETM+) Landsat sensor observations. The innovative part of the mapping approach is the local optimization of the land cover classifier, which has resulted in increased spatial consistency and accuracy. Training and classifying with locally confined reference samples over a large number of partially overlapping areas (i.e., moving windows) ensures the optimization of the classifier to a local land cover distribution, and decreases the negative effect of signature extension. A weighted combination of labels, which is determined by the classifier in overlapping windows, defines the final label for each pixel. Since the approach requires extensive computation, it has been developed and deployed using the Government of Canada's High-Performance Computing Center (HPC). An accuracy assessment based on 2811 randomly distributed samples shows that land cover data produced with this new approach has achieved 76.60% accuracy with no marked spatial disparities. [ABSTRACT FROM AUTHOR]

Soil spatial variations are scale dependent and can be controlled by many environmental factors. Numerous factors have been related to variations of soil salinity, some of them combined empirical mode decomposition (EMD) method and correlation analysis. However, environmental factors that essentially affect soil-water balance are not given enough attention. In addition, further analysis is needed in exploring how well the environmental factors can interpret the variations in soil salinity at different scales, especially in arid oasis areas and at large scales. This paper explores the potential of modeling variations in soil salinity via the EMD and Random Forest modeling of remote sensing based environmental factors. A case study is presented for Tarim basin, Xinjiang, China, using land surface temperature (LST), evapotranspiration (ET), TRMM precipitation (TRM) and digital elevation model (DEM) products. Soil salinity and its decompositions were first correlated with environmental factors for feature selection. Then, those selected environmental factors and their decompositions were correlated and coupled with their counterparts of soil salinity to evaluate their synchronization. Finally, those IMF components of environmental factors that had high correlation coefficients and were coupled well with corresponding IMF components of soil salinity were identified and divided into different feature sets for modeling. Mean absolute error and mean bias error were adopted for accuracy assessment of the models. Our results indicate that soil salinity series can be separated into eight scales ranging from 170 km to 480 km. IMF components 5–7 account for most of the variation and can be modeled using the corresponding IMF components of different combinations of DEM, ET, LST and TRM. IMF components 6–7 are well coupled with LST and ET at approximately 475 km scale. Overall, regional-scale modeling of variations in soil salinity based on remote sensing products is possible. Reasonably accurate results can be obtained in arid oasis areas where researchers and policy makers must focus on preventing the loss of agricultural productivity and ecosystem health. [ABSTRACT FROM AUTHOR]

Land cover information extraction through object-based image analysis (OBIA) has become an important trend in remote sensing, thanks to the increasing availability of high-resolution imagery. Segmented objects have a large number of features that cause high-dimension and low-sample size problems in the classification process. In this study, on the basis of a partial least squares generalized linear regression (PLSGLR), we propose a group corrected PLSGLR, known as G-PLSGLR, that aims to reduce the redundancy of object features for land cover identifications. Using Gaofen-2 images, the area of interest was segmented and sampled to generate small sample-size training datasets with 51 object features. The features selected by G-PLSGLR were compared against a guided regularized random forest (GRRF) in metrics of reduction rate, feature redundancy, and accuracy assessment of classification. Three indicators of overall accuracy (OA), user's accuracy (UA), and producer's accuracy (PA) were applied for accuracy assessment in this paper. The result shows that the G-PLSGLR achieved a reduction rate of 9.27 with a feature redundancy of 0.29, and a value of OA 90.63%. The GRRF achieved a reduction rate of 1.61 with a feature redundancy of 0.42, and a value of OA 85.56%. The PA of each land cover category was more than 95% using features selected by G-PLSGLR, while the PA ranged from 77 to 96% using features selected by GRRF. The UA of G-PLSGLR-selected features ranged from 70 to 80% except for grass land and bare land, which achieved 10% higher UA than GRRF-selected features. The G-PLSGLR method we proposed has the advantages of a large reduction rate, low feature redundancy, and high classification performance, which can be applied in OBIA-based land cover classification. [ABSTRACT FROM AUTHOR]

Slum identification in urban settlements is a crucial step in the process of formulation of pro-poor policies. However, the use of conventional methods for slum detection such as field surveys can be time-consuming and costly. This paper explores the possibility of implementing a low-cost standardized method for slum detection. We use spectral, texture and structural features extracted from very high spatial resolution imagery as input data and evaluate the capability of three machine learning algorithms (Logistic Regression, Support Vector Machine and Random Forest) to classify urban areas as slum or no-slum. Using data from Buenos Aires (Argentina), Medellin (Colombia) and Recife (Brazil), we found that Support Vector Machine with radial basis kernel delivers the best performance (with F2-scores over 0.81). We also found that singularities within cities preclude the use of a unified classification model. [ABSTRACT FROM AUTHOR]

In light of the need for fine-grained, accurate, and timely urban land use information, a per-field classification approach was proposed in this paper to automatically map fine-grained urban land use in a study area within Haidian District, Beijing, China, in 2016. High-resolution remote sensing imagery and multi-source social sensing data were used to provide both physical and socioeconomic information. Four categories of attributes were derived from both data sources for urban land use parcels segmented by the OpenStreetMap road network, including spectral/texture attributes, landscape metrics, Baidu Point-Of-Interest (POI) attributes, and Weibo attributes. The random forests technique was adopted to conduct the classification. The importance of each attribute, attribute category, and data source was evaluated for the classification as a whole and the classification of individual land use types. The results showed that a testing accuracy of 77.83% can be achieved. The approach is relatively good at classifying open space and residential parcels, and poor at classifying institutional parcels. While using solely remote sensing data or social sensing data can achieve equally high overall accuracy, their importance varies in terms of the classification of individual classes. Landscape metrics are the most important for open space parcels. Spectral/texture attributes are more important in identifying institutional and residential parcels. The classification of business parcels relies more on landscape metrics and social sensing data, and less on spectral/texture attributes. The classification accuracy can be potentially improved upon the acquisition of purer parcels and the addition of new attributes. It is expected that the proposed approach will be useful for the routine update of urban land use information and large-scale urban land use mapping. [ABSTRACT FROM AUTHOR]