Artificial Intelligence in Geosciences最新文献

Magnitude estimation is a critical task in seismology, and conventional methods usually require dense seismic station arrays to provide data with sufficient spatiotemporal distribution. In this context, we propose the Earthquake Graph Network (EQGraphNet) to enhance the performance of single-station magnitude estimation. The backbone of the proposed model consists of eleven convolutional neural network layers and ten RCGL modules, where a RCGL combines a Residual Connection and a Graph convolutional Layer capable of mitigating the over-smoothing problem and simultaneously extracting temporal features of seismic signals. Our work uses the STanford EArthquake Dataset for model training and performance testing. Compared with three existing deep learning models, EQGraphNet demonstrates improved accuracy for both local magnitude and duration magnitude scales. To evaluate the robustness, we add natural background noise to the model input and find that EQGraphNet achieves the best results, particularly for signals with lower signal-to-noise ratios. Additionally, by replacing various network components and comparing their estimation performances, we illustrate the contribution of each part of EQGraphNet, validating the rationality of our approach. We also demonstrate the generalization capability of our model across different earthquakes occurring environments, achieving mean errors of $\pm$0.1 units. Furthermore, by demonstrating the effectiveness of deeper architectures, this work encourages further exploration of deeper GNN models for both multi-station and single-station magnitude estimation.

This study develops a method to use deep learning models to predict the mineral assemblages and their relative abundances in paleolake cores using high-resolution XRF core scan elemental data and X-ray diffraction (XRD) mineralogical results from the same core taken at coarser resolution. It uses the XRF core scan data along with published mineralogical information from the Olduvai Gorge Coring Project (OGCP) 2014 sediment cores 1A, 2A, and 3A from Paleolake Olduvai, Tanzania. Both regression and classification models were developed using a Keras deep learning framework to assess the predictability of mineral assemblages with their relative abundances (in regression models) or at least the mineral assemblages (in classification models) using XRF core scan data. Models were created using the Sequential class and Functional API with different model architectures. The correlation matrix of element ratios calculated from XRF element intensity records from the cores and XRD-derived mineralogical information was used to select the most useful features to train the models. 1057 training data records were used for the models. Lithological classes were also used for some models using Wide & Deep neural networks since those combine the benefits of memorization and generalization for mineral prediction. The results were validated using 265 validation data records unseen by the model and discuss the accuracy of models using six test records. The optimized Deep Neural Network (DNN) classification model achieved over 86% binary accuracy while the regression models were also able to predict the relative mineral abundances of samples with high accuracies. Overall, the study shows the efficacy of a carefully crafted Deep Learning (DL) model for predicting mineral assemblages and abundances using high-resolution XRF core scan data.

In this study, we present an artificial neural network (ANN)-based approach for travel-time tomography of a volcanic edifice under sparse-ray coverage. We employ ray tracing to simulate the propagation of seismic waves through the heterogeneous medium of a volcanic edifice, and an inverse modeling algorithm that uses an ANN to estimate the velocity structure from the “observed” travel-time data. The performance of the approach is evaluated through a 2-dimensional numerical study that simulates i) an active source seismic experiment with a few (explosive) sources placed on one side of the edifice and a dense line of receivers placed on the other side, and ii) earthquakes located inside the edifice with receivers placed on both sides of the edifice. The results are compared with those obtained from conventional damped linear inversion. The average Root Mean Square Error (RMSE) between the input and output models is approximately 0.03 km/s for the ANN inversions, whereas it is about 0.4 km/s for the linear inversions, demonstrating that the ANN-based approach outperforms the classical approach, particularly in situations with sparse ray coverage. Our study emphasizes the advantages of employing a relatively simple ANN architecture in conjunction with second-order optimizers to minimize the loss function. Compared to using first-order optimizers, our ANN architecture shows a ∼25% reduction in RMSE. The ANN-based approach is computationally efficient. We observed that even though the ANN is trained based on completely random velocity models, it is still capable of resolving previously unseen anomalous structures within the edifice with about 5% anomalous discrepancies, making it a potentially valuable tool for the detection of low velocity anomalies related to magmatic intrusions or mush.

Water prediction plays a crucial role in modern-day water resource management, encompassing both hydrological patterns and demand forecasts. To gain insights into its current focus, status, and emerging themes, this study analyzed 876 articles published between 2015 and 2022, retrieved from the Web of Science database. Leveraging CiteSpace visualization software, bibliometric techniques, and literature review methodologies, the investigation identified essential literature related to water prediction using machine learning and deep learning approaches. Through a comprehensive analysis, the study identified significant countries, institutions, authors, journals, and keywords in this field. By exploring this data, the research mapped out prevailing trends and cutting-edge areas, providing valuable insights for researchers and practitioners involved in water prediction through machine learning and deep learning. The study aims to guide future inquiries by highlighting key research domains and emerging areas of interest.

Emeralds – the green colored variety of beryl – occur as gem-quality specimens in over fifty deposits globally. While digital traceability methods for emerald have limitations, sample-based approaches offer robust alternatives, particularly for determining the geographic origin of emerald. Three factors make emerald suitable for provenance studies and hence for developing models for origin determination. First, the diverse elemental chemistry of emerald at minor (<1 wt%) and trace levels (<1 to 100’s ppmw) exhibits unique inter-element fractionations between global deposits. Second, minimally destructive techniques, including laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS), enable measurement of these diagnostic elemental signatures. Third, when applied to extensive datasets, machine learning (ML) techniques enable the creation of predictive models and statistical discrimination with adequate characterization of the deposits. This study employs a carefully selected dataset comprising more than 1000 LA-ICP-MS analyses of gem-quality emeralds, enriched with new analyses. This dataset represents the largest available for global emerald deposits. We conducted unsupervised exploratory analysis using Principal Component Analysis (PCA). For machine learning-based classification, we employed Support Vector Machine Classification (SVM-C), achieving an initial accuracy rate of 79%. This was enhanced to 96.8% through the use of hierarchical SVM-C with PCA filters as our modeling approach. The ML models were trained using the concentrations of eight statistically significant elements (Li, V, Cr, Fe, Sc, Ga, Rb, Cs). By leveraging high-quality LA-ICP-MS data and ML techniques, accurate identification of the geographical origin of emerald becomes possible. These models are important for accurate provenance of emerald, and from a geochemical perspective, for understanding the formation environments of beryl-bearing pegmatites and shales.

Seismic inversion can be divided into time-domain inversion and frequency-domain inversion based on different transform domains. Time-domain inversion has stronger stability and noise resistance compared to frequency-domain inversion. Frequency domain inversion has stronger ability to identify small-scale bodies and higher inversion resolution. Therefore, the research on the joint inversion method in the time-frequency domain is of great significance for improving the inversion resolution, stability, and noise resistance. The introduction of prior information constraints can effectively reduce ambiguity in the inversion process. However, the existing model-driven time-frequency joint inversion assumes a specific prior distribution of the reservoir. These methods do not consider the original features of the data and are difficult to describe the relationship between time-domain features and frequency-domain features. Therefore, this paper proposes a high-resolution seismic inversion method based on joint data-driven in the time-frequency domain. The method is based on the impedance and reflectivity samples from logging, using joint dictionary learning to obtain adaptive feature information of the reservoir, and using sparse coefficients to capture the intrinsic relationship between impedance and reflectivity. The optimization result of the inversion is achieved through the regularization term of the joint dictionary sparse representation. We have finally achieved an inversion method that combines constraints on time-domain features and frequency features. By testing the model data and field data, the method has higher resolution in the inversion results and good noise resistance.

Fluctuations in oil prices adversely affect decision making situations in which performance forecasting must be combined with realistic price forecasts. In periods of significant price drops, companies may consider extended duration of well shut-ins (i.e. temporarily stopping oil production) for economic reasons. For example, prices during the early days of the Covid-19 pandemic forced operators to consider shutting in all or some of their active wells. In the case of partial shut-in, selection of candidate wells may evolve as a challenging decision problem considering the uncertainties involved. In this study, a mature oil field with a long (50+ years) production history with 170+ wells is considered. Reservoirs with similar conditions face many challenges related to economic sustainability such as frequent maintenance requirements and low production rates. We aimed to solve this decision-making problem through unsupervised machine learning. Average reservoir characteristics at well locations, well production performance statistics and well locations are used as potential features that could characterize similarities and differences among wells. While reservoir characteristics are measured at well locations for the purpose of describing the subsurface reservoir, well performance consists of volumetric rates and pressures, which are frequently measured during oil production. After a multivariate data analysis that explored correlations among parameters, clustering algorithms were used to identify groups of wells that are similar with respect to aforementioned features. Using the field’s reservoir simulation model, scenarios of shutting in different groups of wells were simulated. Forecasted reservoir performance for three years was used for economic evaluation that assumed an oil price drop to $30/bbl for 6, 12 or 18 months. Results of economic analysis were analyzed to identify which group(s) of wells should have been shut-in by also considering the sensitivity to different price levels. It was observed that wells can be characterized in the 3-cluster case as low, medium and high performance wells. Analyzing the forecasting scenarios showed that shutting in all or high- and medium-performance wells altogether results in better economic outcomes. The results were most sensitive to the number of active wells and the oil price during the high-price period. This study demonstrated the effectiveness of unsupervised machine learning in well classification for operational decision making purposes. Operating companies may use this approach for improved decision making to select wells for extended shut-in during low oil-price periods. This approach would lead to cost savings especially in mature fields with low-profit margins.

Machine learning methods dealing with the spatial auto-correlation of the response variable have garnered significant attention in the context of spatial prediction. Nonetheless, under these methods, the relationship between the response variable and explanatory variables is assumed to be homogeneous throughout the entire study area. This assumption, known as spatial stationarity, is very questionable in real-world situations due to the influence of contextual factors. Therefore, allowing the relationship between the target variable and predictor variables to vary spatially within the study region is more reasonable. However, existing machine learning techniques accounting for the spatially varying relationship between the dependent variable and the predictor variables do not capture the spatial auto-correlation of the dependent variable itself. Moreover, under these techniques, local machine learning models are effectively built using only fewer observations, which can lead to well-known issues such as over-fitting and the curse of dimensionality. This paper introduces a novel geostatistical machine learning approach where both the spatial auto-correlation of the response variable and the spatial non-stationarity of the regression relationship between the response and predictor variables are explicitly considered. The basic idea consists of relying on the local stationarity assumption to build a collection of local machine learning models while leveraging on the local spatial auto-correlation of the response variable to locally augment the training dataset. The proposed method’s effectiveness is showcased via experiments conducted on synthetic spatial data with known characteristics as well as real-world spatial data. In the synthetic (resp. real) case study, the proposed method’s predictive accuracy, as indicated by the Root Mean Square Error (RMSE) on the test set, is 17% (resp. 7%) better than that of popular machine learning methods dealing with the response variable’s spatial auto-correlation. Additionally, this method is not only valuable for spatial prediction but also offers a deeper understanding of how the relationship between the target and predictor variables varies across space, and it can even be used to investigate the local significance of predictor variables.

Microfossil classification is an important discipline in subsurface exploration, for both oil & gas and Carbon Capture and Storage (CCS). The abundance and distribution of species found in sedimentary rocks provide valuable information about the age and depositional environment. However, the analysis is difficult and time-consuming, as it is based on manual work by human experts. Attempts to automate this process face two key challenges: (1) the input data are very large - our dataset is projected to grow to 3 billion microfossils, and (2) there are not enough labeled data to use the standard procedure of training a deep learning classifier. We propose an efficient pipeline for processing and grouping fossils by genus, or even species, from microscope slides using self-supervised learning. First we show how to efficiently extract crops from whole slide images by adapting previously trained object detection algorithms. Second, we provide a comparison of a range of self-supervised learning methods to classify and identify microfossils from very few labels. We obtain excellent results with both convolutional neural networks and vision transformers fine-tuned by self-supervision. Our approach is fast and computationally light, providing a handy tool for geologists working with microfossils.

Since its arrival in late November 2022, ChatGPT-3.5 has rapidly gained popularity and significantly impacted how research is planned, conducted, and published using a generative artificial intelligence approach. ChatGPT-4 was released four months later and became more popular in November 2023. However, there is little study about the perception of scientists of these chatbots, especially in soil science. This article presents the new findings of a brief research investigating soil scientists' responses and perceptions towards chatbots in Indonesia. This artificial intelligence application facilitates conversation-based interactions in text format. The study evaluated ten ChatGPT answers to fundamental questions in soil science, which has developed into a normal science with a mutually agreed-upon paradigm. The evaluation was carried out by seven soil scientists recognized for their expertise in Indonesia, using a scale of 1–100. In addition, a questionnaire was distributed to soil scientists at the National Research and Innovation Agency of the Republic of Indonesia (BRIN), universities, and Indonesian Soil Science Society (HITI) members to gauge their perception of ChatGPT's presence in the research field. The study results indicate that the scores of ChatGPT answers range from 82.99 to 92.24. ChatGPT-4 is better than both the paid and free versions of ChatGPT-3.5. There is no significant difference between the English and Indonesian versions of ChatGPT-4.0. However, the perception of general soil scientists about the level of trust is only 55%. Furthermore, 80% of soil scientists believe that chatbots can only be used as digital tools to assist in soil science research and cannot be used without the involvement of soil scientists.