Mangrove sediments are natural carbon sinks that may act as key components for climate change mitigation. To investigate the characteristics and distribution of the carbon-dense muds in the coastal mangrove areas of northern Taiwan, we applied both floating and submerged electrodes for subsurface resistivity imaging. After collecting the apparent resistivity data, we conducted 2D resistivity inversion and 3D modeling. Our results show that the muddy sediments are characterized by low resistivity and are primarily found in the top ten meters below the riverbed. On the other hand, a higher resistivity layer, probably indicating coarse-grained sediments, is situated below the muddy layer. Although the submerged electrodes generally provide the best data quality, the floating electrodes efficiently image the bottom of the muddy sediments. Therefore, we recommend that a combination of floating and submerged electrode methods for resistivity imaging should be an efficient approach to investigate mud distribution in mangrove sediments with shallow water depths.
How to cite: Han, W.-C., Lin, K.-I., Chen, L., and Liu, H.-C.: Electrical resistivity imaging of mangrove sediments, northern Taiwan, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-2366, https://doi.org/10.5194/egusphere-egu25-2366, 2025.
Heavy Rare Earth Elements (HREEs) are critical for the production of most electronic devices. Rapidly increasing demand for these minerals has led to a proliferation of highly polluting makeshift HREE extraction in Myanmar. Monitoring the spread of these mines is important for the preservation of human health and the environment. This paper utilizes a geospatial foundation model pre-trained using self-supervised learning to detect hundreds of rare earth mines using a single template example. This is achieved through the development of a novel method for embedding similarity search-- Cosine Contrast-- which leverages both positive and negative templates to yield more relevant results.
How to cite: Ballinger, O.: Monitoring Illicit Rare Earth Mining in Myanmar via Self-Supervised Learning, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-4496, https://doi.org/10.5194/egusphere-egu25-4496, 2025.
Coalbed methane (CBM) is considered an unconventional gas resource. Accurate determination of CBM content can provide potential disaster warnings and guide exploration and development. Direct measurement and statistical analysis of CBM content are common techniques. Hoverer，direct measurement methods have high accuracy, but they are time consuming, labor intensive, and inefficient; statistical methods have a limited ability to solve complicated nonlinear problems, for example, CBM content prediction commonly used computational methods do not have high enough accuracy due to the small amount of training data and the shallow model structure. 3D seismic exploration has been widely used in CBM exploration and development due to its small grid size and high resolution. It will improve the accuracy of coalbed methane prediction to combine 3D seismic data with coalbed methane content. Machine learning techniques are a set of computational methods that can learn from data and make accurate predictions. In recent years，many applications of machine-learning techniques for CBM content prediction are found to be more reliable，however the results from traditional machine learning models have errors to some extent. A CBM content model based on Deep Belief Network (DBN) has been developed in this paper, with the input as continuous real values and the activation function as the rectified linear unit. Firstly, various seismic attributes of the target coal seam were calculated to highlight its features, then the original attribute features were preprocessed, and finally the performance of the DBN model was developed using the preprocessed features. Different from conventional DBN models, the proposed model uses continuous real values as the input and the rectified linear unit (ReLU) as the activation function. Training process includes pre training and fine-tuning. Pre training gives the model good initial parameters by training with unlabeled data, and fine-tuning uses a standard supervised method with labeled data to optimize the model. This paper successfully applied a DBN model to predict CBM content from a CBM 3D seismic prospecting district. With more layers pre trained, the average error decreased from 3.69% to 2.16% and from 2% to 5.76% for the maximum error. Using a pre training strategy to initialize the model’s parameters can improve the accuracy of the model results. Compared with the typical multilayer perceptron（MLP）and the support vector regression（SVR）models, the DBN model has the smallest error, which means it is more accurate in predicting CBM content than the other two models.
How to cite: Du, W., Peng, S., and Cui, X.: Coalbed methane content prediction based on deep belief network, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-4908, https://doi.org/10.5194/egusphere-egu25-4908, 2025.
The São Domingos Mine (Mértola, Portugal) is an abandoned sulphide mine whose exploitation has had a long-term impact on its soil and water contamination problem covering an area of approximately a length of 20 km and a width of 2 km. The mining heaps are spread along a watercourse that flows into the Chança River dam, which merges with the Guadiana River, both international rivers. This constitutes a serious environmental problem leading to contamination by heavy metals (HMs). Contamination assessment is a slow process that involves collecting soil samples for HMs analysis.
The study of mining heaps using geophysical and geospatial methods allows us to determine their depth and the volume of accumulated materials, as well as their characterization in relation to soils contaminated by HMs. We propose the use of electromagnetic induction, electrical resistivity tomography and GNSS methods to carry on the analysis.
This work is part of an interdisciplinary study that is being carried out within the scope of the INCOME Project (Inputs for a more sustainable region – Instruments for managing metal-contaminated areas). The aim is to combine data from Geophysics, Chemistry and Remote Sensing to create a tool, using Artificial Intelligence, that allows the calculation of contamination maps using less data than standard methodologies.
This is a sustainable management model that will increase optimization and reduce resources spent in the sampling and analysis phases. Moreover, the model aims to provide important real-time information for decision-making subjected to monitoring and managing pollution. It also has a high replication potential for other contaminated environments, such as landfills, industry or even intensive agriculture.
Funding: The work was supported by the Promove Program of the “la Caixa” Foundation, in partnership with BPI and the Foundation for Science and Technology (FCT), in the scope of the project INCOME – Inputs para uma região mais sustentável: Instrumentos para a gestão de zonas contaminadas por metais (Inputs for a more sustainable region: Instruments for managing metal-contaminated areas), PD23-00013. Acknowledgment: CREATE Project (R&D Unit ID 6107).
How to cite: Oliveira, R. J., Caldeira, B., Borges, J. F., Teixeira, P., Rodrigues, G., Costa, M. J., Palma, P., Custódio, M., Catarino, A., and Semedo, N.: Geophysical and geospatial characterization of mining heaps of the São Domingos Mine (Mértola, Portugal), EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-7010, https://doi.org/10.5194/egusphere-egu25-7010, 2025.
Many endocrine-disrupting chemicals (EDCs) are discharged into the aquatic environment mainly due to their incomplete removal during biological treatment at municipal wastewater treatment plants. For this reason, advanced oxidation processes (AOPs), using ozone with other oxidant agents, like hydrogen peroxide, are effective in removing EDCs. Furthermore, to reduce the risk of drinking water contamination by EDCs, it is necessary to ensure a real-time monitoring of wastewater treatment processes. Fluorescence spectroscopy could be used for wastewater quality monitoring to control the fate of EDCs in water systems. However, the complex physical, biological and chemical process involved in wastewater treatment process exhibit non-linear behaviors, which are difficult to describe by linear mathematical models. The artificial neural networks (ANNs) have been applied with remarkable success in several modeling studies including the highly non-linear ones.
The main objective of the present work was to use fluorescence data and ANN to monitor two EDCs, namely a pesticide (Diuron) and a pharmaceulical and corrosion inhibitor (Benzotriazole) during advanced wastewater treatments.
The data used were obtained from the pilot plant installed and operated by AquaSoil at the municipal wastewater reclamation plant of Fasano (Brindisi, Italy). The influent wastewater was obtained from tertiary treatment consisting of a coagulation stage by aluminum polychloride, sedimentation stage in lamella clarifiers and disinfection stage by sodium hypochlorite. An aliquot of the tertiary effluent was redirected to the pilot plant employing the O3/H2O2 advanced oxidation process. This process was operated in the patented technology commercialized by AquaSoil as MITO3X.
Diuron and Benzotriazole were analyzed using standard methos. Fluorescence data were collected using a Shimadzu RF-5301PC fluorescence spectrophotometer at different excitation emission wavelengths, while ANN model has been developed using Matlab software with ANN toolbox to match the measured and the predicted concentrations of EDCs.
The concentrations of Diuron and Benzotriazole were well correlated with selected fluorescence indexes. The combination of differente fluorescence peaks enhanced the determination coefficients of the single and multiple linear regressions. The developed ANN model that incorporated as input parameters the values of the fluorescence indices strongly enhanced the prediction of the fate of Diuron and Benzotriazole during AOPs. Therefore, the ANN-based model have been found to provide an efficient and robust tool in predicting the fate of EDCs removal. The comparison between ANN predicted data and experimental data shows the ability of artificial intelligence tools to predict EDCs concentrations with high accuracy and precision. Moreover, this model requires no additional information on the mechanism and the kinetics of chemical degradation of target contaminants. Since ANN have valuable advantages such as learning ability, dealing with imprecise, noisy and highly complex non-linear data, and parallel processing ability and due to the high sensitivity of fluorescence, it is expected that the developed fluorescence-ANN based model can be successfully applied for real-time control of AOPs employed for EDCs removal. This may also lead to AOPs optimization and cost savings.
How to cite: Roccaro, P., Fazzino, F., Spadaro, M. R., Gagliano, E., and Santoro, D.: Modelling the fate of endocrine-disrupting chemicals during wastewater ozonation by fluorescence and artificial neural network, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-7399, https://doi.org/10.5194/egusphere-egu25-7399, 2025.
According to the European Union Soil Strategy for 2030, it is crucial to address contamination from mining areas, and its impact on watercourses. to achieve these goals, it is essential developing methodologies for identifying and monitoring contaminated areas, and implementing sustainable solutions for their recovery to protect soil health and ensure sustainable land use. In Portugal, soil and water contamination in former mining areas, is a significant environmental challenge, especially due to the presence of potentially toxic metals that can affect human health and ecosystems. São Domingos mine, located in the Iberian Pyrite Belt, is an open-pit mine, submerged in acidic drainage water, resulting from mining extraction activities carried out until the middle of the 20th century. In this sense, the objective of this study was to analyze the chemical and biological characterization of the soils of the São Domingos Mine, contributing to the development of an environmental management model for abandoned mining areas. To achieve this purpose, 11 topsoil (0-20cm) samples (A2 to A12) were collected in São Domingos mine, and the following parameters were analyzed: (i) chemical: pH (deionized water suspension of 1:2.5 (w/v)); electrical conductivity (EC) (deionized water suspension of 1:2 (w/v)); phosphorus (P) and potassium (K) (Egner-Riehm Method); total nitrogen (N) (Kjeldah method); organic matter (OM) (Walkley & Black method); (ii) biological(enzymatic parameters): dehydrogenase activity, acid phosphatase activity and β-glucosidase activity. The results evidenced pH ranged from 3 to 4 (very acidic). The EC, ranging from 115 to 5043 µS/cm, with most of the samples classified as non-saline. The percentage of OM was generally low (0.2 to 2.5%). Regarding macronutrients, the results were equally limiting, with the samples showing low levels of N (0.05 to 0.17%), P (1 to 6 mg P2O5 kg-1) and K (3 to 30 mg K2O kg-1). Analysis of enzyme parameters revealed low enzymatic activity frequently lower than the detection limit of the technique. An exception to β-glucosidase that generally had low values, (0.01 to 0.40 µmol PNP g-1 DM h-1), and phosphatase showing values among 0.27 to 0.96 µmol PNP g-1 DM h-1. This can be mainly related to the low values of pH, low percentage of organic matter and nutrients, and high amount of potentially toxic metals. These results will be extremely important in the development of the environmental management model proposed in INCOME project, as they provide essential information on the variability of the contamination in the mine area, essential information for validate the rest of the methodologies applied. Further, this type of model will be applicable to other regions of contamination, contributing to economic and tourist development, public health, and protection of local ecosystems, in line with the Sustainable Development Goals.
Funding: The work was supported by the Promove Program of the “La Caixa” Foundation, in partnership with BPI and the Foundation for Science and Technology (FCT), in the scope of the project INCOME - Inputs for a more sustainable region: Instruments for managing metal-contaminated areas, PD23-00013. Acknowledgment: CREATE Project (R&D Unit ID 6107).
How to cite: Semedo, N., Custódio, M., Catarino, A., Rodrigues, G., Teixeira, P., Oliveira, R. J., Palma, P., Caldeira, B., and Costa, M. J.: Soils chemical and biological characterization tools for managing metal-contaminated areas: case-study São Domingos mine (South of Portugal), EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-10841, https://doi.org/10.5194/egusphere-egu25-10841, 2025.
The INCOME project (Instruments for Managing Areas Contaminated by Metals) proposes the development of an environmental management model for mining soils contaminated by metals. This study presents preliminary results obtained using the Multispectral Imager (MSI) aboard the European Space Agency's (ESA) Sentinel-2 satellite to identify contaminated soils in the São Domingos Mine, located in southeastern Portugal.
The MSI instrument offers significant advantages, including high spatial resolution (10, 20, or 60 m depending on the spectral band), open access for rapid image download, and frequent revisitation of the study area. The preliminary analysis focuses on identifying areas with fully exposed soil using spectral indices, which combine spectral measurements at different wavelengths to improve classification accuracy. Additionally, the Random Forest (RF) method, a widely recognised approach to general-purpose classification, was tested. Contaminated soils characteristically exhibit discrepancies in optical properties, such as distinct colouration, which can also be detected in the visible region bands of the MSI instrument. The Shortwave Infrared (SWIR) bands are particularly efficacious for identifying heavy metals.
The designated soil areas will be subject to monitoring for metal contamination utilizing the MSI instrument, with the prospective incorporation of hyperspectral data from satellites such as the Environmental Mapping and Analysis Program (EnMAP).
Funding: The work was supported by the Promove Program of the “la Caixa” Foundation, in partnership with BPI and the Foundation for Science and Technology (FCT), in the scope of the project INCOME – Inputs para uma região mais sustentável: Instrumentos para a gestão de zonas contaminadas por metais (Inputs for a more sustainable region: Instruments for managing metal-contaminated areas), PD23-00013. Acknowledgment: CREATE Project (R&D Unit ID 6107).
How to cite: Rodrigues, G., Teixeira, P., Oliveira, R. J., Costa, M. J., Palma, P., Caldeira, B., Custódio, M., Catarino, A., and Semedo, N.: Preliminary Assessment of Metal Contamination in Mining Soils Using Sentinel-2 MSI: A Case Study of São Domingos Mine, Portugal, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-11957, https://doi.org/10.5194/egusphere-egu25-11957, 2025.
Traveltime tomography recovers the background velocity field by minimizing the difference between observed and theoretical traveltime. Due to its computational efficiency and robustness, this method has been widely applied in studies of Earth's internal structure, oil and gas exploration, and other fields. However, most existing studies rely on regular rectangular grids for tomography, which exhibit limited adaptability when dealing with irregular topography and subsurface interfaces. The utilization of unstructured triangular meshes are more suitable for handling such complex study areas, and the development of traveltime tomography based on triangular meshes is necessary.
Compared with rectangular grids, the inversion method based on triangular meshes faces more complex gradient computation formulas, which has, to some extent, hindered the development of traveltime tomography. To address this challenge, we introduce automatic differentiation (AD) method to calculate the gradients more automatically, enabling the implementation of traveltime tomography based on triangular meshes. After building the forward computational graph, AD method can compute the gradient using the chain rule, thereby saving a lot of manpower in theoretical derivation, coding, and other processes. In this study, we used a finite difference method based on triangular meshes to solve the eikonal equation, accurately and efficiently calculating the traveltime in complex structural areas. Then, we integrate the eikonal solver into the current deep learning framework (e.g. pytorch), and update the velocity model with its built-in optimization algorithm after calculating the gradient in AD method. The process of traveltime tomography is completed on GPU, which can fully utilize the computing power of GPU and efficiently calculate inversion. Numerical tests indicate that the method can achieve promising inversion results and provide a suitable initial model for the full-waveform inversion. Our research provides a new approach for seismic inversion with unstructured grids, which is helpful for high-precision imaging of complex structural areas.
How to cite: Chen, X., Zhang, T., Cao, D., and Liao, W.: Traveltime tomography on the triangular mesh based on automatic differentiation, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-14686, https://doi.org/10.5194/egusphere-egu25-14686, 2025.
In recent years, due to the enhanced interference-resistance of airborne gravimeters and the advanced gravity anomaly calculation techniques, the China Geological Survey has carried out numerous airborne gravity survey missions in mid-high mountainous and deeply incised regions, including Tibet, Xinjiang, and Gansu. In practical applications, the measured free-air gravity anomalies need to have local topographic corrections and intermediate layer corrections to obtain Bouguer gravity anomalies for geological interpretation. Currently, commercial airborne gravity terrain correction software adopts the Nagy flat-topped prism method for near-field areas and the rod formula for far-field areas. This approach results in poor continuity between different terrain correction zones and fails to effectively eliminate terrain effects in deeply incised areas. This paper presents a novel method. By utilizing the coordinate surfaces in the spherical coordinate system, namely conical surfaces and semi-planes, the area is divided into rings (m rings) and blocks (n blocks), forming m×n "sectorial spherical shell blocks". A terrain correction calculation formula for these "sectorial spherical shell blocks" in the circular domain is derived, unifying the terrain correction formulas for both near and far regions. This unification allows for seamless connection among various terrain correction areas and obviates the need for intermediate layer corrections. The method has been validated by theoretical models, showing reliable accuracy in terrain correction value calculations. It has also been successfully applied in the West Kunlun airborne gravity survey. When compared with commercial software, it effectively eliminates terrain effects and achieves better terrain correction results.
How to cite: Wang, L., Xue, D., Wang, G., Teng, D., and Zheng, J.: Aerogravity terrain correction method based on spherical coordinate system, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-15987, https://doi.org/10.5194/egusphere-egu25-15987, 2025.
Posters virtual: Mon, 28 Apr, 14:00–15:45 | vPoster spot 4
EGU25-2102 | ECS | Posters virtual | VPS16
Atmospheric correction of geostationary ocean color imager data over turbid coastal waters under high solar zenith anglesMon, 28 Apr, 14:00–15:45 (CEST) | vP4.2
The traditional atmospheric correction models employed with the near-infrared iterative schemes inaccurately estimate aerosol radiance at high solar zenith angles (SZAs), leading to a substantial loss of valid products for dawn or dusk observations by the geostationary satellite ocean color sensor. To overcome this issue, we previously developed an atmospheric correction model suitable for open ocean waters observed by the first geostationary satellite ocean color imager (GOCI) under high SZAs. This model was constructed based on a dataset from stable open ocean waters, which makes it less suitable for coastal waters. In this study, we developed a specialized atmospheric correction model (GOCI-II-NN) capable of accurately retrieving the water-leaving radiance from GOCI-II observations in coastal oceans under high SZAs. We utilized multiple observations from GOCI-II throughout the day to develop the selection criteria for extracting the stable coastal water pixels and created a new training dataset for the proposed model. The performance of the GOCI-II-NN model was validated by in-situ data collected from coastal/shelf waters. The results showed an Average Percentage Difference (APD) of less than 23% across the entire visible spectrum. In terms of the valid data and retrieval accuracy, the GOCI-II-NN model was superior to the traditional near-infrared and ultraviolet atmospheric correction models in terms of accurately retrieving the ocean color products for various applications, such as tracking/monitoring of algal blooms, sediment dynamics, and water quality among other applications.
How to cite: Li, H., He, X., Palanisamy, S., Bai, Y., and Xuchen, J.: Atmospheric correction of geostationary ocean color imager data over turbid coastal waters under high solar zenith angles, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-2102, https://doi.org/10.5194/egusphere-egu25-2102, 2025.
