Explore the vast range of titles available.

The strong heterogeneity of clastic reservoirs and the phenomenon of similar log responses for different lithologies (i.e., “same spectrum, different rocks”) significantly weaken feature separability. Furthermore, distribution shifts between different wells cause traditional models to suffer from severe generalization bottlenecks in cross-well applications. To address this critical challenge, this paper proposes a dual-driven framework comprising “Multivariate Feature Enhancement + Dynamic Ensemble”. At the feature level, physics-informed enhancement and multi-scale statistics are introduced to construct a Multivariate high-dimensional feature system, thereby strengthening the representation of geological patterns. At the model level, a sample-aware Dynamic Confidence-Weighted Ensemble (DCWE) strategy is designed to achieve sample-wise adaptive decision-making based on prediction uncertainty, fundamentally breaking through the limitations of fixed weights in static ensembles. This method combines the complementary advantages of Gradient Boosting Decision Trees (GBDT) and deep sequence networks, enabling the simultaneous capture of local textural variations and continuous trends across depths. Based on rigorous Leave-One-Group-Out (LOGO) cross-validation, the proposed framework achieves a maximum accuracy of 84.58%. It significantly reduces the misclassification rate in lithology transition zones and for minority class samples, while maintaining the geological continuity of prediction results. These results verify the significant advantages of the proposed method in cross-well generalization scenarios.

Currently, seafloor topography inversion based on satellite altimetry gravity data provides the principal means to predict the global seafloor topography. Researchers often use sea surface geoid height or gravity anomaly to predict sea depth in the space domain. In this paper, a comprehensive discussion on seafloor topography inversion formulas in the space domain is presented using sea surface geoid height, gravity anomaly and introduces an approach that uses vertical gravity gradient. This would be the first study to estimate seafloor topography by vertical gravity gradient in the space domain. Further, a nonlinear iterative least-square inversion process is discussed. Using the search area for the Malaysia Airlines Flight MH370 as study site, we used the DTU17 gravity anomaly model and SIO V29.1 vertical gravity gradient to generate the seafloor topography. The results of the proposed bathymetric models were analyzed and compared with the DTU18 and SIO V20.1 bathymetric models. The experimental results show that the gravity anomaly and vertical gravity gradient in the study area are strongly correlated with the seafloor topography in the 20–200 km wavelength range. The optimal initial iteration values for seafloor topography variance and correlation length are 0.6365 km2 and 10.5′, respectively. Shipborne measurements from SONAR data were used as external checkpoints to evaluate the bathymetric models. The results show that the RMS for BAT_VGG_ILS (inversion model constructed by vertical gravity gradient) is smaller than for BAT_GA_ILS (inversion model constructed by gravity anomaly) and BAT_GA_VGG_ILS (inversion model constructed by gravity anomaly and vertical gravity gradient). The relative accuracy of the DTU18 bathymetry model was 9.27%, while the relative accuracy of the proposed seafloor models was higher than 4%. Within the 200 m difference range, the proportion of checkpoints for BAT_VGG_ILS was close to 95%, about 80% for BAT_GA_ILS and BAT_GA_VGG_ILS, and less than 50% for the DTU18. The results show that the nonlinear iterative least square method in the space domain is feasible.

To address the limitations in global seafloor topography model construction, a scheme is proposed that takes into account the efficiency of seafloor topography prediction, the applicability of inversion methods, the heterogeneity of seafloor environments, and the inversion advantages of sea surface gravity field element. Using the South China Sea as a study area, we analyzed and developed the methodology in modeling the seafloor topography, and then evaluated the feasibility and effectiveness of the modeling strategy. Based on the proposed modeling approach, the STO_IEU2020 global bathymetry model was constructed using various input data, including the SIO V29.1 gravity anomaly (GA) and vertical gravity gradient anomaly (VGG), as well as bathymetric data from multiple sources (single beam, multi-beam, seismic, Electronic Navigation Chart, and radar sensor). Five evaluation areas located in the Atlantic and Indian Oceans were used to assess the performance of the generated model. The results showed that 79%, 89%, 72%, 92% and 93% of the checkpoints were within the ±100 m range for the five evaluation areas, and with average relative accuracy better than 6%. The generated STO_IEU2020 model correlates well with the SIO V20.1 model, indicating that the proposed construction strategy for global seafloor topography is feasible.

One of the key constraints for the accelerometer of GRACE-type gravity satellites to accurately measure the non-gravitational accelerations acting on the satellite is that the center of mass of the satellite and the proof mass of the accelerometer should maintain a coincidence. In addition, the accuracy requirement is that the center of mass offset (CM-offset) in the three directions is less than 100 microns. Since the center of mass (CoM) of the satellite will change with the consumption of cold-gas fuel in the tanks, it is necessary to regularly carry out the CoM calibration maneuver. Firstly, the observation equations consisting of the accelerometer linear acceleration, angular acceleration, and the CM-offset vector are established in order to estimate the amount of CM-offset. Then, according to the estimated CM-offset, the satellite mass trim mechanisms are used to change the satellite’s CoM, so that the satellite’s CoM always approaches the proof mass of the accelerometer, with an accuracy of 100 μm per axis. The CM-offset of the satellite of GRACE-FO is estimated by using the accelerometer, star camera, magnetic torquer, magnetometer, and the precision orbit data during the GRACE-C CM-offset calibration period on 1 February 2020. Four kinds of CM-offset results are obtained by four different angular accelerations as follows: the angular acceleration based on the attitude dynamics (“MTQ angular acceleration”), the accelerometer angular acceleration calibrated by MTQ, the accelerometer angular acceleration, and the angular acceleration calculated by the star camera. By comparing the four kinds of CM-offset results that are estimated by the four different methods, all four of the results are shown to have the same level of accuracy. Based on the accelerometer (calibrated) angular acceleration, the difference with the JPL result is 0.5 μm, while the difference between the conventional method and the JPL result is 6.0 μm. All four of the methods can achieve the requirement of 50 μm accuracy and using four CM-offset estimation methods simultaneously can improve the integrity of the calibration results. Subsequently, the CM-offset results of GRACE-C since its launch are estimated here. The calibration algorithm that is proposed in this paper can be used as a reference in the calibration of gravity satellites carrying an accelerometer payload.

Total organic carbon (TOC) is a critical parameter for evaluating shale source rock quality and hydrocarbon generation potential. However, accurate TOC estimation from conventional well logs remains challenging, especially in data-limited geological settings. This study proposes an optimized XGBoost model for TOC prediction using conventional logging data from the Shahejie Formation in the Dongying Depression, Bohai Bay Basin, China. We systematically transform four standard logs—resistivity, acoustic transit time, density, and neutron porosity—into 165 candidate features through multi-scale smoothing, statistical derivation, interaction term creation, and spectral transformation. A two-stage feature selection process, combining univariate filtering and recursive feature elimination and further refined by principal component analysis, identifies ten optimal predictors. The model hyperparameters are optimized via Bayesian search within the Optuna framework to minimize cross-validation error. The optimized model achieves an R2 of 0.9395, with a Mean Absolute Error (MAE) of 0.3392, a Root Mean Squared Error (RMSE) of 0.4259, and a Normalized Root Mean Squared Error (NRMSE) of 0.0604 on the test set, demonstrating excellent predictive accuracy and generalization capability. This study provides a reliable and interpretable methodology for TOC characterization, offering a valuable reference for source rock evaluation in analogous shale formations and sedimentary basins.

In practical well-logging datasets, severe missing values, anomalous disturbances, and highly imbalanced lithology classes are pervasive. To address these challenges, this study proposes a well-logging lithology identification framework that combines Robust Feature Engineering (RFE) with quality-aware XGBoost. Instead of relying on interpolation-based data cleaning, RFE uses sentinel values and a meta-information tensor to explicitly encode patterns of missingness and anomalies, and incorporates sliding-window context to transform data defects into discriminative auxiliary features. In parallel, a quality-aware sample-weighting strategy is introduced that jointly accounts for formation boundary locations and label confidence, thereby mitigating training bias induced by long-tailed class distributions. Experiments on the FORCE 2020 lithology prediction dataset demonstrate that, relative to baseline models, the proposed method improves the weighted F1 score from 0.66 to 0.73, while Boundary F1 and the geological penalty score are also consistently enhanced. These results indicate that, compared with traditional workflows that rely solely on data cleaning, explicit modeling of data incompleteness provides more pronounced advantages in terms of robustness and engineering applicability.

Tuberculosis (TB) is one of the world’s leading causes of death from a single infectious agent, with Mycobacterium tuberculosis (Mtb) primarily infecting the lungs via the respiratory tract. Following infection, immune cells such as macrophages and neutrophils phagocytose Mtb and initiate complex inflammatory and immune responses, driving the formation of granulomas and cavities within the lungs, ultimately leading to structural damage. In this intricate cascade, the MAPK signaling emerges as a critical regulator, orchestrating various cellular processes including inflammatory signaling, autophagy, apoptosis, and immune cell differentiation. Emerging evidence indicates that MAPK signaling critically shapes anti-TB immunity predominantly within macrophages, neutrophils, T cells and dendritic cells. Through extensive crosstalk among immune cells, MAPK signaling influences both host defense and disease progression. This review systematically summarizes current advances in understanding MAPK-mediated immune regulation during TB infection, with particular emphasis on the distinct roles of p38, ERK, and JNK signaling pathways. Furthermore, we discuss emerging therapeutic strategies to enhance anti-mycobacterial immunity by targeting MAPK signaling, thereby providing a valuable theoretical framework for the development of novel TB treatments.

In this paper, selective catalytic reduction (SCR) of N2O by CO was investigated over Fe-beta zeolites catalysts. The catalysts were prepared by wet ion-exchange (IE), impregnation (IM) and solid state ion-exchange (SSIE) methods. These catalysts were characterized by XRD, UV–vis DR spectroscopy, H2-TPR, TPD and in-situ DRIFTS. At 350 °C, more than 90% N2O conversion could be obtained over the Fe-beta-IE catalyst. The activity for N2O removal of Fe-beta-IE was higher than Fe-beta-IM and Fe-beta-SSIE catalysts. The UV–vis spectra showed that 84.2% of isolated Fe(III) ion appeared on Fe-beta-IE catalyst. It indicated that the isolated Fe(III) ions might be considered as the active sites for N2O reduction. Besides, in the presence of H2O, the activities in CO-SCR for N2O removal over Fe-beta catalysts were inhibited, which might be due to the hydroxylation deactivation of iron species and excess accumulation of carbonates.Graphic Abstract

The coordinates of the KBR (K-band ranging system) antenna phase center of GRACE-type gravity satellites in the satellite Science Reference Frame should be precisely known, and the determination accuracy should reach 0.3 mrad in the Y (pitch) and Z (yaw) directions. Due to the precision limitation of ground measurement and the change of space environment during orbit, the KBR antenna phase center changes. In order to obtain more accurate KBR antenna phase center coordinates, it is necessary to maneuver the satellite to achieve the on-orbit calibration of the KBR antenna phase center. Based on the in-orbit calibration data of KBR of GRACE-FO satellites, a new method is proposed to estimate the antenna phase center of KBR using the inter-satellite range acceleration as the observation value. The antenna phase center of KBR is solved by the robust estimation method, and the obtained calibration results are better than 72 μm in the Y and Z directions and better than 1.3 mm in the X direction, which is 50% better than the least squares estimation algorithm. The accuracy of KBR calibration results obtained by using the data of positive maneuvers or mirror (negative) maneuvers, respectively, does not meet 0.3 mrad. It is shown that mirror maneuvers are required for KBR calibration of a GRACE-type gravity satellite to obtain antenna phase center estimation results that meet the accuracy requirements. The calibration algorithm proposed in this paper can provide reference for KBR antenna phase center calibration of Chinese GRACE-type gravity satellites.

Calcrete-type uranium prospectivity prediction is challenged by the strong heterogeneity of multi-source geoscientific raster datasets, weak anomaly responses, and the lack of explicit heterogeneous information organization in conventional deep learning models. In this study, the Yilgarn Craton of Western Australia was selected as the study area, and a geology-guided fixed-group fusion ResUNet model (GGF-ResUNet) was developed based on 12-channel multi-source geoscientific raster datasets. At the input stage, the evidence layers were divided into four fixed geoscientific proxy groups according to their data modality and geological interpretation, namely gravity, aeromagnetic, radiometric, and geochemical groups, and intra-group channel weighting together with inter-group gating was introduced to enhance the hierarchical representation and adaptive fusion of heterogeneous information. Ablation results showed that GGF-ResUNet achieved better performance than the baseline ResUNet, with AUC increasing from 0.9340 to 0.9740 and F1-score improving from 0.7264 to 0.8356. Further comparative experiments with Attention U-Net, U-Net, SegNet, and FCN showed that GGF-ResUNet achieved comparatively better quantitative performance and more spatially coherent prediction results under the current experimental setting. Without substantially increasing model complexity, the proposed method improves the representation and integration of heterogeneous geoscientific information and provides a feasible technical pathway for calcrete-type uranium prospectivity prediction under weak-anomaly conditions.