This study investigates the spatiotemporal evolution and scaling characteristics of the 2025 seismic swarm in the Santorini–Amorgos region, one of the most seismically and volcanically active zonesin the Aegean arc. Using a high-resolution earthquake catalog including 1,601 events (ML≥ 2.9), we analyze variations in key statistical parameters—including b-value, fractal dimension (Dc), scaling exponent (αb), and the q-parameter—within cumulative time windows. Our findings reveal a marked decrease in b-value and simultaneous increase in Dcduring the swarm’s activation phase, consistent with an inverse energy cascade and progressive rupture. The spatial extent of the seismicswarm (L) contracts significantly before expanding during the decay phase, and vertical migration of the seismic barycenter suggests a fluid-triggered nucleation process. 3D visualizations confirm fault-aligned clustering and upward migration patterns, supporting a hybrid rupture mechanism governed by both tectonic faulting and fluid diffusion. These results provide critical insights into the complex dynamics of swarms in extensional volcanic environments and highlight the relevance of fractal and scaling analysis for short-term seismic hazard assessment.

Vehicle mobility across different terrains depends on a multitude of geographical and meteorological variables. Traditional approaches rely on labor-intensive manual field measurements to assess soil conditions for military or humanitarian vehicle passage. In this study, we aimed to explore novel methods for parameterizing the Cone Index, a fundamental metric for assessing passage feasibility, leveraging meteorological data from the global numerical model GFS. Focusing specifically on chernosols, primarily agricultural soils found in flat, open regions conducive to military operations, we utilized machine learning methods to assess how soil conditions affect vehicle mobility. Through rigorous exploratory analysis, we investigated correlations, model performance metrics, and the relative importance of predictors in Cone Index modeling. Our findings highlight the comparative efficacy of different modeling approaches, particularly emphasizing the utility of the Random Forest method. We identified key environmental conditions under which the model reliably predicts the Cone Index. This sets the baseline for spatial modelling in GIS. Despite these insights, our study is constrained by data limitations and the inherent resolution constraints of the GFS model. The obtained solution lays the initial groundwork for implementing the model in a GIS environment to predict the trafficability of chernosols across the broader European region. Future research will aim to expand the dataset, spatial relationships, and employ models with higher resolution for more robust and accurate predictions.

On September 18, 2024, an M4.7 earthquake struck Feidong County in Hefei, Anhui Province. Located at the complex junction of the Tanlu fault zone, this event exhibited a clear foreshock–mainshock–aftershock evolution, attracting widespread attention. To investigate the aftershock activity and seismogenic structure of this earthquake sequence, this study employs deep learning methods applied to continuous waveform data from the Anhui seismic network for automatic detection. Coupled with precise earthquake relocation, a comprehensive and high-precision seismic catalog was constructed. The relocation results indicate that the Earthquake sequence are mainly distributed along a linear trend in the NE–SW direction, extending approximately 12 km, with hypocentral depths concentrated between 8 and 13 km. Notably, both the mainshock and most aftershocks occur at a depth of about 10.5 km. Focal mechanism solutions reveal that the dominant earthquakes in the region exhibit right-lateral strike-slip rupture on steeply dipping faults trending NE, which is highly consistent with the orientation and structural characteristics of the primary Zhuding–Shimenshan fault in the area. In conjunction with observational evidence of fluid intrusion, it is inferred that fluids may have migrated from the southwest into the seismogenic zone through structural pathways such as the Zhuding–Shimenshan fault and accumulated within a depth range of approximately 10 km near the fault. This fluid accumulation likely induced local stress concentration, which may be the primary trigger for the Feidong earthquake sequence.

We demonstrate the use of Facebook's Prophet (usually just called Prophet) model for short-term air quality forecasting at Belgrade-Zeleno brdo monitoring station. To address missing data, we applied minimally-altering data distribution imputation techniques. Linear interpolation proved effective for short-term gaps (1–3 hours), hourly mean method for mid-term gaps (24–26 hours), and Hermite interpolation polynomial for long-term gaps (132–148 hours). The most significant data change was a 3.4% shift in skewness. Partitioning the time series enabled a detailed quality assessment of the Prophet model, with PM2.5 predictions being more precise than PM10. Using the longest time series for forecasting yielded absolute errors of 6.5 μg/m3 for PM10 and 2.7 μg/m3 for PM2.5. Based on 173 forecasts, we anticipate Prophet model root-mean-square values under 6.26 μg/m3 and 9.99 μg/m3 for PM2.5 and PM10 in 50% of cases. The Prophet model demonstrates several advantages and yields satisfactory results. In future research, the results obtained from the Prophet model will serve as benchmark values for other models. Additionally, the Prophet model is capable of providing satisfactory air quality forecasting results and will be utilized in future research.

This paper examines the expected future heat load in five Croatian cities: Dubrovnik, Zadar, Rijeka, Zagreb, and Osijek. The heat load is estimated by temperature-related climate indices and mean, maximum and minimum daily temperatures obtained by climate simulations using two different regional climate models (DHMZ-RegCM4 and SMHI-RCA4) with a horizontal resolution of 12.5 km, forced with two global climate models (EC-EARTH and MPI-ESM-MR/LR) for two different greenhouse gas concentration scenarios (RCP4.5 and RCP8.5). By comparing these variables for the period 2041–2070 with respect to the current climate (defined as that of the period 1991–2020), a significant increase in mean, maximum and minimum temperatures was observed in all analysed combinations of regional and global climate models for all analysed cities. Although there is a difference in results depending on the combination of regional and global models, the largest increase is mainly found in the warm part of the year (April-October), with the strongest warming of Dubrovnik and Rijeka. Due to similar trends in minimum and maximum temperatures, the trend in the daily temperature range is weak. Under warmer climate conditions, the number of days with a maximum air temperature above 25 °C increases in all considered cities (especially in Dubrovnik), as does the number of days with a minimum air temperature exceeding 20 °C (especially in Rijeka and Zadar). Furthermore, a reduction in the number of days with maximum and minimum temperatures below 0 °C is projected for all cities. Nevertheless, some differences are found between coastal and inland cities caused by local factors.

This paper presents two novel spectral soil area indices to identify bare soil area and distinguish it more accurately from the urban impervious surface area (ISA). This study designs these indices based on medium spatial resolution remote sensing data from Landsat 8 OLI dataset. Extracting bare soil or urban ISA is more challenging than extracting water bodies or vegetation in multispectral Remote Sensing (RS). Bare soil and the urban ISA area often were mixed because of their spectral similarity in multispectral sensors. This study proposes Normalized Soil Area Index 1 (NSAI1) and Normalized Soil Area Index 2 (NSAI2) using typical multispectral bands. Experiments show that these two indices have an overall accuracy of around 90%. The spectral similarity index (SDI) shows these two indices have higher separability between soil area and ISA than previous indices. The result shows that percentile thresholds can effectively classify bare soil areas from the background. The combined use of both indices measured the soil area of the study area over 71 km2. Most importantly, proposed soil indices can refine urban ISA measurement accuracy in spatiotemporal studies.

Accurate visibility forecasting is essential for safe aircraft operations. This study examines how various configurations of the Random Forest model can enhance visibility predictions. Preprocessing techniques are employed, including correlation analysis to identify fundamental relationships in weather observations. Time-series data is transformed into a regular Data Frame to facilitate analysis. This study proposes a classification framework for organizing visibility data and phenomena, which is then used to develop a visibility forecast using the Random Forest method. The study also presents procedures for hyperparameter tuning, feature selection, data balancing, and accuracy evaluation for this dataset. The main outcomes are the Random Forest model parameters for a three-hour visibility forecast, along with an analysis of errors in low visibility forecasts. Additionally, models for one-hour forecasts and visibility forecasting under precipitation are also examined. The resulting models demonstrate a deterministic forecast accuracy of approximately 78%, with a false alarm rate of around 6%, providing a comprehensive overview of the capabilities of the Random Forest model for visibility forecasting. As anticipated, the model demonstrated limitations in accurately simulating fast radiative cooling or abrupt decreases in visibility caused by precipitation. Specifically, in relation to precipitation, the model achieved an accuracy of 79%, yet exhibited a false alarm rate of 19%. Additionally, this method sets a foundation for enhancing prediction accuracy through the inclusion of supplementary forecast data, while its implementation on real-world datasets expands the reach of machine learning techniques to the members of the meteorological community.

Knowing the soil moisture distribution in the unsaturated zone can improve understanding the water flow through the unsaturated zone and thereby enable the calculation of aquifer recharge, which occurs through precipitation. One part of the Zagreb aquifer recharge occurs through infiltration from precipitation. In order to observe and model infiltration from precipitation through the unsaturated zone, the research polygon was constructed at the Velika Gorica well field, located in the southern part of the Zagreb aquifer, Croatia, where hourly measurements of electric conductivity (EC) and soil moisture content were carried out. EC and soil moisture data are measured by Time Domain Reflectometry (TDR) probes which are placed at different depths in the unsaturated zone. Furthermore, electrical resistivity tomography (ERT) measurements were conducted. Geophysical data, along with moisture and EC data from TDR probes, were used as input data for MoisturEC software, in order to obtain soil moisture distribution along a 2D profile. MoisturEC program offers three options for translating EC data to moisture content data which are all tested in this research. We obtained eight moisture content distributions along the observed profile and concluded that MoisturEC provides reasonable results with input data from geophysical measurements and TDR probe measurements. Soil moisture distribution in the unsaturated zone represents the initial conditions for further unsaturated flow modeling. Understanding the flow in the unsaturated zone enables the quantification of effective infiltration and can improve groundwater management.

Renewable energy is one of the most important factors for developed and sustainable societies. However, its utilization in electrical power grid systems can be very challenging regarding rates predictably. Renewable energy depends mainly on environmental conditions such as rainfall-runoff ratios and temperature. Because of that, the expected power production heavily fluctuates, which makes the prediction and calculation of feed-in into the power grid very challenging. The accurate forecasting of energy production is a very crucial issue for power management process. This paper presents the results of deploying Machine Learning Techniques in short-term forecasting of the amount of energy produced of General Circulation Models (GCMs) Data by Almus Dam and Hydroelectric Power Plant in Tokat, Turkey. The study demonstrates the use of modeling techniques in hydropower forecasting process using the predicted monthly hydroelectric power generation data of GCMs from 2018 to 2080. Decision Tree, Deep Learning, Generalized Linear, Gradient Boosted Trees and Random Forest models are utilized to forecast the hydropower production. The results show that the correlation value of the gradient boosted trees model equals 0.717, which means that the gradient boosted trees model is the most successful model for the present data. The gradient boosted trees model used in the prediction process for each GCM in each scenario is 4.5 and 8.5. The results show that there are small differences between the models, which means that the predictions are going in similar directions for all these models.

Global flood hazard is gradually increasing. Though it is impossible to avoid them, losses and damage of hazards (e.g., floods, cyclones, and earthquakes) could be efficiently reduced by reducing household vulnerability with appropriate measures. This study aims to quantitatively measure the household vulnerability of flood hazards as a mitigation tool. It also proposed a unique approach to quantify flood-hazard household vulnerability, and shows its application in the flood prone city of Dhaka as an example case. Data were collected from both slum and non-slum areas to cover the entire urban habitat, and to compare their level of flood vulnerability. A total of 300 households were surveyed by structured questionnaire on the basis of five factors (economic, social, environmental, structural, and institutional) of flood vulnerability. The analytical hierarchy process (AHP) was applied to measure individual household vulnerability scores by using the relative weightage of variables and indicators with proper standardisation. Analytical results demonstrated that 63.06% slum and 20.02% non-slum households were highly vulnerable to floods. In addition, this paper determined and assessed responsible factors for household flood vulnerability in Dhaka. For structural vulnerability, results exhibited that 82% of slum households were highly vulnerable, and 95.3% of non-slum households were moderately vulnerable. Socially, 67.3% of slum and 78.7% of non-slum households were moderately and low-vulnerable. The majority of slum and non-slum households (84% and 59.3%, respectively) showed high and moderate vulnerability with respect to economic vulnerability. Moreover, 69.3% of slum and 65.3% of nonslum household institutional vulnerability levels were high. Of slum inhabitants, 63.3% were environmentally at high risk, and 78% of non-slum habitats were in the low-vulnerability category. However, as an effective tool to measure location-specific vulnerability, it is applicable for the measuring vulnerability of other cities in the world with proper customisation. On the basis of this study, future research could be conducted with more factors, variables, and indicators of human vulnerability to natural or artificial hazards/disasters. Future work may provide a better reflection of the vulnerability status of single/multiple hazard(s)/disaster(s).