arXiv subjects

Alexandre G. Evsukoff

Publications and source records attributed to Alexandre G. Evsukoff.

7 recordsLinked to original sources

arXiv2026

Towards Fast GNN Surrogates for CO2 Migration in Complex Geological Formations

This chapter discusses how a data-driven machine learning approach can reproduce key aspects of the physical behavior of multiphase flows in complex geological formations. We propose an end-to-end graph neural surrogate tailored to CO$_2$ plume migration forecasting in geological storage. The method is evaluated on the SPE11A benchmark, a well-known industry test case designed to assess CO$_2$ storage scenarios and characterized by sharp gas-water interfaces, strong advective transport, and rapid convective mixing with fingering development. The benchmark is reformulated as a graph in which nodes represent computational cells and edges encode transmissibility-based interactions enriched with geometric attributes. Directional transport arising from grid geometry, permeability contrasts, and geological heterogeneity is captured through an anisotropic message-passing mechanism, where interaction weights are computed via geometry-conditioned edge embeddings, biasing message aggregation toward physically relevant transport directions. Temporal evolution is modeled in latent space using an autoregressive residual formulation trained with multi-step supervision. The proposed model produces competitive forecasts of gas saturation and liquid-phase density, which are key indicators for CO$_2$ storage monitoring, with cumulative errors that remain moderate over extended forecasting horizons.

cs.LG↗

arXiv2026

Deploying Self-Supervised Learning for Real Seismic Data Denoising

Self-supervised learning (SSL) has emerged as a promising approach to seismic data denoising as it does not require clean reference data. In this work, the deployment of the Noisy-as-Clean (NaC) method was evaluated for real seismic data denoising under controlled conditions. Two independent seismic acquisitions, each comprising noisy and filtered data, were organized into four real datasets. The NaC SSL method was adapted to add real noise to the noisy input, controlled by a parameter. An experimental protocol with ten experiments was designed to compare different strategies for deploying the NaC SSL method with the supervised learning baseline, using identical network topology and hyperparameters. The models were evaluated in terms of denoising performance, computational cost, and generalization capability. The results show that the synthetic additive white Gaussian noise (AWGN) is inadequate for the denoising of seismic data within the NaC method, and performance strongly depends on the compatibility between the injected and actual noise characteristics. Furthermore, both the characteristics of the seismic data and the noise level influence the performance of the model. Self-supervised fine-tuning on test data has improved SSL performance, whereas no such gain was observed for fine-tuning of supervised models. Finally, NaC has shown to be a simple, effective, and model-independent method that offers a feasible solution for the denoising of real seismic data.

physics.geo-ph↗

arXiv2025

A multiresolution weather dataset for the Southwestern South Atlantic (2017-2018)

The Southwestern South Atlantic (SWSA) is a key region for climate research and renewable energy assessment, yet high-resolution meteorological data are scarce. We present a multiresolution dataset spanning February 2017--November 2018, combining Weather Research and Forecasting (WRF) simulations with Sentinel-1A/B Synthetic Aperture Radar (SAR) wind fields processed using the CMOD5 model. WRF outputs were generated every 30 minutes for three nested domains (9 km, 3 km, 1 km) through 975 short-term simulations. SAR/CMOD5 wind fields are provided at 500 m and 1 km resolution across 104 acquisition dates. Validation shows strong agreement: daily spatial averages of 10 m wind speed yield RMSE and MAE below 3 m/s on over 93% of acquisition days, while more than 91.5% of pixel-level residuals fall within $\pm$3 m/s. In situ measurements from the Itajaí buoy further confirmed the reliability of both sources. The dataset supports regional climate studies, wind energy resource assessment, and machine-learning applications in forecasting and downscaling, with usage examples included to aid practical adoption.

physics.ao-ph↗

arXiv2025

Validating Urban Scaling Laws through Mobile Phone Data: A Continental-Scale Analysis of Brazil's Largest Cities

\abstract{Urban scaling theories posit that larger cities exhibit disproportionately higher levels of socioeconomic activity and human interactions. Yet, evidence from developing contexts (especially those marked by stark socioeconomic disparities) remains limited. To address this gap, we analyse a month-long dataset of 3.1~billion voice-call records from Brazil's 100 most populous cities, providing a continental-scale test of urban scaling laws. We measure interactions using two complementary proxies: the number of phone-based contacts (voice-call degrees) and the number of trips inferred from consecutive calls in distinct locations. Our findings reveal clear superlinear relationships in both metrics, indicating that larger urban centres exhibit intensified remote communication and physical mobility. We further observe that gross domestic product (GDP) also scales superlinearly with population, consistent with broader claims that economic output grows faster than city size. Conversely, the number of antennas required per user scales sublinearly, suggesting economies of scale in telecommunications infrastructure. Although the dataset covers a single provider, its widespread coverage in major cities supports the robustness of the results. We nonetheless discuss potential biases, including city-specific marketing campaigns and predominantly prepaid users, as well as the open question of whether higher interaction drives wealth or vice versa. Overall, this study enriches our understanding of urban scaling, emphasising how communication and mobility jointly shape the socioeconomic landscapes of rapidly growing cities.

physics.soc-ph↗

arXiv2024

A Real Benchmark Swell Noise Dataset for Performing Seismic Data Denoising via Deep Learning

The recent development of deep learning (DL) methods for computer vision has been driven by the creation of open benchmark datasets on which new algorithms can be tested and compared with reproducible results. Although DL methods have many applications in geophysics, few real seismic datasets are available for benchmarking DL models, especially for denoising real data, which is one of the main problems in seismic data processing scenarios in the oil and gas industry. This article presents a benchmark dataset composed of synthetic seismic data corrupted with noise extracted from a filtering process implemented on real data. In this work, a comparison between two well-known DL-based denoising models is conducted on this dataset, which is proposed as a benchmark for accelerating the development of new solutions for seismic data denoising. This work also introduces a new evaluation metric that can capture small variations in model results. The results show that DL models are effective at denoising seismic data, but some issues remain to be solved.

physics.geo-ph↗

arXiv2020

Deep learning for lithological classification of carbonate rock micro-CT images

In addition to the ongoing development, pre-salt carbonate reservoir characterization remains a challenge, primarily due to inherent geological particularities. These challenges stimulate the use of well-established technologies, such as artificial intelligence algorithms, for image classification tasks. Therefore, this work intends to present an application of deep learning techniques to identify patterns in Brazilian pre-salt carbonate rock microtomographic images, thus making possible lithological classification. Four convolutional neural network models were proposed. The first model includes three convolutional layers followed by fully connected layers and is used as a base model for the following proposals. In the next two models, we replace the max pooling layer with a spatial pyramid pooling and a global average pooling layer. The last model uses a combination of spatial pyramid pooling followed by global average pooling in place of the last pooling layer. All models are compared using original images, when possible, as well as resized images. The dataset consists of 6,000 images from three different classes. The model performances were evaluated by each image individually, as well as by the most frequently predicted class for each sample. According to accuracy, Model 2 trained on resized images achieved the best results, reaching an average of 75.54% for the first evaluation approach and an average of 81.33% for the second. We developed a workflow to automate and accelerate the lithology classification of Brazilian pre-salt carbonate samples by categorizing microtomographic images using deep learning algorithms in a non-destructive way.

cs.CV↗

arXiv2020

Entropy as a measure of attractiveness and socioeconomic complexity in Rio de Janeiro metropolitan area

Defining and measuring spatial inequalities across the urban environment remains a complex and elusive task that has been facilitated by the increasing availability of large geolocated databases. In this study, we rely on a mobile phone dataset and an entropy-based metric to measure the attractiveness of a location in the Rio de Janeiro Metropolitan Area (Brazil) as the diversity of visitors' location of residence. The results show that the attractiveness of a given location measured by entropy is an important descriptor of the socioeconomic status of the location, and can thus be used as a proxy for complex socioeconomic indicators.

physics.soc-ph↗