The main objective of this study is to estimate the changes in land use and land cover in the Peshawar basin, Pakistan, from 2000 to 2020. This will greatly improve the selection of areas designated as the agricultural, industrial, and/or urban sectors of the region and will help in overcoming future problems. With the help of an advanced geographic information system (GIS), land-use and topographic changes were identified. Based on data of the 20 years from 2000 to 2020, the total runoff volume in the Peshawar basin from 2000 to 2010 was calculated to be 13.9 km³ and from 2010 to 2020 was 19.4 km³. This volume estimation will assist in quantifying the total infiltration rate. We inferred that the built-up area increased the most from 2010 to 2020 as compared to other classes. Results showed that from 2000 to 2020, there was a significant increase in urbanization and a significant decrease in vegetation. This study will help the farmer community and environmentalists to manage range land, agricultural land, populations, and water bodies.

Well logging is a significant procedure that assists geophysicists and geologists with making predictions regarding boreholes and efficiently utilizing and optimizing the drilling process. The current study area is positioned in the Punjab Territory of Pakistan, and the geographic coordinates are 30020′10 N and 70043′30 E. The objective of the current research work was to interpret the subsurface structure and reservoir characteristics of the Kabirwala area Tola (01) well, which is located in the Punjab platform, Central Indus Basin, utilizing 2D seismic and well log data. Formation evaluation for hydrocarbon potential using the reservoir properties is performed in this study. For the marked zone of interest, the study also focuses on evaluating the average water saturation, average total porosity, average effective porosity, and net pay thickness. The results of the study show a spotted horizon stone with respect to time and depth as follows: Dunghan formation, 0.9 s and 1080.46 m; Cretaceous Samana Suk formation, 0.96 s and 1174.05 m; Datta formation, 1.08 s and 1400 m; and Warcha formation, 1.24 s and 1810 m. Based on the interpretation of well logs, the purpose of petrophysical analysis was to identify hydrocarbon-bearing zones in the study area. Gamma ray, spontaneous potential, resistivity, neutron, and density log data were utilized. The high zone present in the east–west part of the contour maps may be a possible location of hydrocarbon entrapment, which is further confirmed by the presence of the Tola-01 well.

The present study interprets the subsurface structure of the Rajian area using seismic sections and the identification of hydrocarbon-bearing zones using petrophysical analysis. The Rajian area lies within the Upper Indus Basin in the southeast (SE) of the Salt Range Potwar Foreland Basin. The marked horizons are identified using formation tops from two vertical wells. Seismic interpretation of the given 2D seismic data reveals that the study area has undergone severe distortion illustrated by thrusts and back thrusts, forming a triangular zone within the subsurface. The final trend of those structures is northwest–southeast (NW–SE), indicating that the area is part of the compressional regime. The zones interpreted by the study of hydrocarbon potential include Sakessar limestone and Khewra sandstone. Due to the unavailability of a petrophysics log within the desired investigation depths, lithology cross-plots were used for the identification of two potential hydrocarbon-bearing zones in one well at depths of 3740–3835 m (zone 1) and 4015–4100 m (zone 2). The results show that zone 2 is almost devoid of hydrocarbons, while zone 1 has an average hydrocarbon saturation of about 11%.

Reliable detection of seismic events from continuous waveform data remains a major challenge due to variations in sensor performance and recording quality across seismic networks. Conventional deep learning models assume uniformly high quality input, which often leads to degraded performance and unstable confidence estimates under noisy conditions. This paper proposes a data quality aware deep learning framework that integrates waveform reliability directly into the detection process. Each waveform is assigned to a scalar quality score derived from measurable indicators such as signal to noise ratio, glitch fraction, and clipping fraction. The score is used as both an auxiliary input feature and a weighting factor in the loss function, allowing the network to emphasize reliable data while maintaining robustness to noise. The model is implemented using a compact one-dimensional convolutional neural network and evaluated on onset waveforms from the Southern California Seismic Network. The proposed framework achieves an average precision and ROC–AUC of 0.989 and an F1 score of 0.949, outperforming baseline CNNs that disregard waveform quality. Stratified analysis shows stable performance across varying quality levels, and calibration evaluation demonstrates well-aligned predicted probabilities. The approach provides a transparent and interpretable method for improving reliability in automated seismic monitoring and can be readily extended to other sensor-based detection applications.

The authors would like to make the following corrections to the published paper