Change Detection of Amazonian Alluvial Gold Mining Using Deep Learning and Sentinel-2 Imagery

A global rise in alluvial mining increases sediment load in tropical rivers

Increasing gold and mineral mining activity in rivers across the global tropics has degraded ecosystems and threatened human health1,2. Such river mineral mining involves intensive excavation and sediment processing in river corridors, altering river form and releasing excess sediment downstream2. Increased suspended sediment loads can reduce water clarity and cause siltation to levels that may result in disease and mortality in fish3,4, poor water quality5 and damage to human infrastructure6. Although river mining has been investigated at local scales, no global synthesis of its physical footprint and impacts on hydrologic systems exists, leaving its full environmental consequences unknown. We assemble and analyse a 37-year satellite database showing pervasive, increasing river mineral mining worldwide. We identify 396 mining districts in 49 countries, concentrated in tropical waterways that are almost universally altered by mining-derived sediment. Of 173 mining-affected rivers, 80% have suspended sediment concentrations (SSCs) more than double pre-mining levels. In 30 countries in which mining affects large (>50 m wide) rivers, 23 ± 19% of large river length is altered by mining-derived sediment, a globe-spanning effect representing 35,000 river kilometres, 6% (±1% s.e.) of all large tropical river reaches. Our findings highlight the ubiquity and intensity of mining-associated degradation in tropical river systems.

A lightweight convolutional neural network based on dense connection for open-pit coal mine service identification using the edge-cloud architecture

Remote sensing is an important technical tool for rapid detection of illegal mining behavior. Due to the complex features of open-pit coal mines, there are few studies about automatic extraction of open-pit coal mines. Based on Convolutional Neural Network and Dense Block, we propose a lightweight densely connected network-AD-Net for the extraction of open-pit coal mining areas from Sentinel-2 remote sensing images, and construct three sample libraries of open-pit coal mining areas in north-central Xinzhou City, Shanxi Province. The AD-Net model consists of two convolutional layers, two pooling layers, a channel attention module, and a Dense Block. The two convolutional layers greatly reduce the complexity of the model, and the Dense Block enhances the feature propagation while reducing the parameter computation. The application is designed in different modules that runs independently on different machines and communicate with each other. Furthermore, we create and build a unique remote sensing image service system that connects a remote datacentre and its associated edge networks, employing the edge-cloud architecture. While the datacentre acts as the cloud platform and is in charge of storing and processing the original remote sensing images, the edge network is largely utilised for caching, predicting, and disseminating the processed images. First, we find out the optimal optimizer and the optimal size of the input image by extensive experiments, and then we compare the extraction effect of AD-Net with AlexNet, VGG-16, GoogLeNet, Xception, ResNet50, and DenseNet121 models in the study area. The experimental results show that the combination of NIR, red, green, and blue band synthesis is more suitable for the extraction of the open-pit coal mine, and the OA and Kappa of AD-Net reach 0.959 and 0.918 respectively, which is better than other models and well balances the classification accuracy and running speed. With this design of edge-cloud, the proposed system not only evenly distributes the strain of processing activities across the edges but also achieves data efficiency among them, reducing the cost of data transmission and improving the latency.

Spatio-temporal monitoring of marsh vegetation phenology and its response to hydro-meteorological factors using CCDC algorithm with optical and SAR images: In case of Honghe National Nature Reserve, China

Vegetation phenology is a sensitive indicator which can comprehensively reflect the response of wetland vegetation to external environment changes. However, the time-series monitoring wetland vegetation phenological changes and clarifying its response to hydrology and meteorology still face great challenges. To fill these research gaps, this paper proposed a novel time-series approach for monitoring phenological change of marsh vegetation in Honghe National Nature Reserve (HNNR), Northeast China, using continuous change detection and classification (CCDC) algorithm and Landsat and Sentinel-1 SAR images from 1985 to 2021. We evaluated the spatio-temporal response relationship of phenological characteristics to hydro-meteorological factors by combining CCDC algorithm with partial least squares regression (PLSR). Finally, this study further explored the intra-annual loss and restoration of marsh vegetation in response to hydro-meteorological factors using the transfer entropy (TE) and CCDC-MLSR model constructed by CCDC and Multiple Linear Stepwise Regression (MLSR) algorithms. We found that the bimodal trajectory of phenology reflects two growth processes of marsh vegetation in one year, and high-frequency and high-amplitude loss occurred in shallow-water and deep-water marsh vegetation from April to October, resulting in the loss area within the year was significantly greater than the recovery area. We confirmed that the CCDC algorithm could track the evolution trajectory of time-series phenology of marsh vegetation. We further revealed that precipitation, temperature and frequency of water-level changes are the main driving factors for the spatio-temporal phenological evolution of different marsh vegetation. This study verified the effect of alternative changes of hydrology and climate on loss and recovery of marsh vegetation in each growth stage. The results of this study provide a scientific basis for wetland protection, ecological restoration, and sustainable development.

ELULC-10, a 10 m European Land Use and Land Cover Map Using Sentinel and Landsat Data in Google Earth Engine

Land Use/Land Cover (LULC) maps can be effectively produced by cost-effective and frequent satellite observations. Powerful cloud computing platforms are emerging as a growing trend in the high utilization of freely accessible remotely sensed data for LULC mapping over large-scale regions using big geodata. This study proposes a workflow to generate a 10 m LULC map of Europe with nine classes, ELULC-10, using European Sentinel-1/-2 and Landsat-8 images, as well as the LUCAS reference samples. More than 200 K and 300 K of in situ surveys and images, respectively, were employed as inputs in the Google Earth Engine (GEE) cloud computing platform to perform classification by an object-based segmentation algorithm and an Artificial Neural Network (ANN). A novel ANN-based data preparation was also presented to remove noisy reference samples from the LUCAS dataset. Additionally, the map was improved using several rule-based post-processing steps. The overall accuracy and kappa coefficient of 2021 ELULC-10 were 95.38% and 0.94, respectively. A detailed report of the classification accuracies was also provided, demonstrating an accurate classification of different classes, such as Woodland and Cropland. Furthermore, rule-based post processing improved LULC class identifications when compared with current studies. The workflow could also supply seasonal, yearly, and change maps considering the proposed integration of complex machine learning algorithms and large satellite and survey data.