Improved Method to Detect the Tailings Ponds from Multispectral Remote Sensing Images Based on Faster R-CNN and Transfer Learning

The breaching of tailings pond dams may lead to casualties and environmental pollution; therefore, timely and accurate monitoring is an essential aspect of managing such structures and preventing accidents. Remote sensing technology is suitable for the regular extraction and monitoring of tailings pond information. However, traditional remote sensing is inefficient and unsuitable for the frequent extraction of large volumes of highly precise information. Object detection, based on deep learning, provides a solution to this problem. Most remote sensing imagery applications for tailings pond object detection using deep learning are based on computer vision, utilizing the true-color triple-band data of high spatial resolution imagery for information extraction. The advantage of remote sensing image data is their greater number of spectral bands (more than three), providing more abundant spectral information. There is a lack of research on fully harnessing multispectral band information to improve the detection precision of tailings ponds. Accordingly, using a sample dataset of tailings pond satellite images from the Gaofen-1 high-resolution Earth observation satellite, we improved the Faster R-CNN deep learning object detection model by increasing the inputs from three true-color bands to four multispectral bands. Moreover, we used the attention mechanism to recalibrate the input contributions. Subsequently, we used a step-by-step transfer learning method to improve and gradually train our model. The improved model could fully utilize the near-infrared (NIR) band information of the images to improve the precision of tailings pond detection. Compared with that of the three true-color band input models, the tailings pond detection average precision (AP) and recall notably improved in our model, with the AP increasing from 82.3% to 85.9% and recall increasing from 65.4% to 71.9%. This research could serve as a reference for using multispectral band information from remote sensing images in the construction and application of deep learning models.

Groundwater pollution in the Mesta River terrace caused by Eleshnitsa tailings pond – current status

Even after the implementation of eco-protective measures, the Eleshnitsa tailings pond continues to contaminate the Quaternary aquifer, formed in the alluvial deposits of Mesta River. A three-dimensional model of mass transport of selected key pollutants is developed on the basis of hydrochemical studies, geophysical surveys, and processing and systematization of monitoring data. The spread and degree of contamination in the Quaternary aquifer are estimated.

Fe, Mn and 238U Accumulations in Phragmites australis Naturally Growing at the Mill Tailings Pond; Iron Plaque Formation Possibly Related to Root-Endophytic Bacteria Producing Siderophores

Mine drainage is a vital water problem in the mining industry worldwide because of the heavy metal elements and low pH. Rhizofiltration using wetland plants is an appropriate method to remove heavy metals from the water via accumulation in the rhizosphere. Phragmites australis is one of the candidate plants for this method because of metal accumulation, forming iron plaque around the roots. At the study site, which was the mill tailings pond in the Ningyo-toge uranium mine, P. australis has been naturally growing since 1998. The results showed that P. australis accumulated Fe, Mn, and 238U in the nodal roots without/with iron plaque compared with other plant tissues. Among the 837 bacterial colonies isolated from nodal roots, 88.6% showed siderophore production activities. Considering iron plaque formation around P. australis roots, we hypothesized that microbial siderophores might influence iron plaque formation because bacterial siderophores have catechol-like functional groups. The complex of catechol or other phenolics with Fe was precipitated due to the networks between Fe and phenolic derivatives. The experiment using bacterial products of root endophytes, such as Pseudomonas spp. and Rhizobium spp., showed precipitation with Fe ions, and we confirmed that several Pseudomonas spp. and Rhizobium spp. produced unidentified phenolic compounds. In conclusion, root-endophytic bacteria such as Pseudomonas spp. and Rhizobium spp., isolated from metal-accumulating roots of P. australis, might influence iron plaque formation as the metal accumulation site. Iron plaque formation is related to tolerance in P. australis, and Pseudomonas spp. and Rhizobium spp. might indirectly contribute to tolerance. Although there are many issues to be resolved in this research, we hope that the fundamental analysis of plant-microbe interactions would be helpful for phytoremediation at mine sites.

Beneficiation of fine chromite slurry at Donskoy Mining and Beneficiation Plant JSC on concentration tables to produce hard chromite pellets

Processing industrial products and technogenic waste is an important task in the mining and metallurgical industry. In Kazakhstan, the processing of chrome ore from the Kempirsay group of deposits has produced more than 15 million tonnes of slurry tailings containing up to 30 wt% chrome oxide. The best results in the world for the processing of fine chromium raw materials are shown by Turkish enterprises with the use of the separation of slurries by size classes and beneficiation on concentration tables. The authors conducted researches for beneficiation of chrome slurry from Dubersay tailings pond (Kazakhstan) with the use of similar technological methods that enabled to obtain concentrates with chrome oxide content of 51 wt% and increasing the yield of beneficiated fine-graded chrome concentrates by 14% as compared with the existing beneficiation process. Strong chromium pellets with a crushing resistance of over 5000 N/pellet were produced from the rich chromium concentrates with the use of the ferrofluxing iron-calcium-silica binder technology by roasting the composition consisting of rich chromium concentrate, ferrous diatomite, and intermediate products and wastes of the chromium industry.

The Stability Analysis of Heituwan Tailings Pond Based on Strength Reduction Method

To analyze the stability of Heituwan tailings pond after vacuum well point dewatering treatment, a unit thickness numerical 3D model was built based on field survey data and physical mechanical properties tests; and the model was analyzed by FEM numerical software according to strength reduction method. The properties of stress and strain, the plastic region, and the deformation properties are acquired by numerical stimulation, and the simulation result was compared withe the in-situ monitoring data. The results show that the safety factor does not meet the requirements of the standard; and most of the landfill made of manganese tailing has developed into plastic status; the deformation is more obvious where the tailing store is higher and closer to roller compacted rockfill dam; the manganese tailing landfill near the roller compacted rockfill dam should be grouted by cement to meet the requirements of continued use.

Propagation Laws of Reclamation Risk in Tailings Ponds Using Complex Network Theory

Accidents have occurred periodically in the tailings ponds where mine solid waste is stored in recent years, and thus their safety has become one of the constraints restricting the sustainable development of the mining industry. Reclamation is an important way to treat tailings ponds, but improper reclamation methods and measures not only cannot reduce the accident risk of tailings ponds, but will further increase the pollution to the surrounding environment. The influencing factors of reclamation accidents in tailings ponds are complex, and the existing models cannot characterize them. In order to study the propagation process of tailings pond reclamation risk, this paper proposes a three-dimensional identification framework for accident hazards based on evidence (TDIFAHE) to identify all potential hazards that may occur during the reclamation stage, and obtain a list of hazards. Based on the complex network theory, this paper uses identified hazards as network nodes and the correlation between hazards as the edges of the network. Based on the identified hazard data, the evolution network of reclamation risk in tailings ponds (ENRRTP) is constructed. By analyzing the statistical characteristics of ENRRTP, it can be found that ENRRTP has small world and scale-free characteristics. The above characteristics show that the reclamation risk of tailings ponds is coupled with multiple factors and the disaster path is short. Giving priority to those hub hazards that have a dominant impact on the reclamation risk can significantly reduce the reclamation risk of the tailings pond.