Search Results (7,477)

Accurate station-level numerical weather predictions are critical for disaster prevention and mitigation, with error correction playing an essential role. However, existing correction models struggle to effectively handle the high-dimensional features and complex dependencies inherent in meteorological data. To address these challenges, this paper proposes the dual-branch residual-guided multi-view attention fusion network (DRAF-Net), a novel deep learning-based correction model. DRAF-Net introduces two key innovations: (1) a dual-branch residual structure that enhances the spatial sensitivity of deep high-dimensional features and improves output stability by connecting raw data and shallow features to deep features, respectively; and (2) a multi-view attention fusion mechanism that models spatiotemporal influences, temporal dynamics, and spatial associations, significantly improving the representation of complex dependencies. The effectiveness of DRAF-Net was validated on two real-world datasets comprising observations and predictions from Chinese meteorological stations. It achieved an average RMSE reduction of 83.44% and an average MAE reduction of 84.21% across all eight variables, significantly outperforming other methods. Moreover, extensive studies confirmed the critical contributions of each key component, while visualization results highlighted the model’s ability to eliminate anomalous values and improve prediction consistency. The code will be made publicly available to support future research and development. Full article

Accurate and precise prediction of actual evapotranspiration (AET) on a large scale is a fundamental issue in natural sciences such as forestry (especially in species selection and planning), hydrology, and agriculture. With the estimation of AET, controlling dams, agriculture, and irrigation and providing potable and utility water supply for industry would be possible. Gathering reliable AET data is possible only with a sufficient weather station network, which is rarely established in many countries like Turkey. Therefore, climate models must be developed for reliable AET data, especially in countries with complex terrains. This study aimed to generate spatiotemporal AET surfaces using the Australian National University spline (ANUSPLIN) model and compare the results with the maps generated by the inverse distance weighting (IDW) and co-kriging (KRG) interpolation techniques. Findings from the interpolated surfaces were validated in three ways: (1) some diagnostics from the surface fitting model include measures such as signal, mean, root mean square predictive error, root mean square error estimate, root mean square residual of the spline, and the estimated standard deviation of noise in the spline; (2) a comparison of common error statistics between the interpolated surfaces and withheld climate data; and (3) evaluation by comparing model results with other interpolation methods using metrics such as mean absolute error, mean error, root mean square error, and adjusted R² (R²_adj). The correlation between AET and normalized difference vegetation index (NDVI) was also evaluated. ANUSPLIN outperformed the other techniques, accounting for 73 to 94% (RMSE: 3.7 to 26.1%) of the seasonal variation in AET with an annual value of 83% (RMSE: 10.0%). The correlation coefficient between observed and predicted AET based on NDVI ranged from 0.49 to 0.71 for point-based and 0.62 to 0.83 for polygon-based data. The generated maps at a spatial resolution of 0.005° × 0.005° could provide valuable insights to researchers and practitioners in the natural resources management domain. Full article

The aim of this study has been the design and evaluation of a sequence of activities that promotes the development of scientific skills in secondary school. For this purpose, design-based research was conducted using a problem-solving methodology to learn as a tool to engage in scientific inquiry practice. The research was structured around the design, implementation, and evaluation phases, with specific tools created to assess both student learning outcomes and the validity of the TLS. These tools helped identify the performance levels achieved by students in the various stages of scientific inquiry, from formulating hypotheses to interpreting data, and also allowed for the evaluation of the teaching methodology’s effectiveness. The results indicated that the TLS significantly enhanced students’ scientific competence by promoting skills related to scientific inquiry, such as hypothesis formulation, variable identification, observation, data collection, and interpretation. Additionally, the use of a weather station as the central topic provided a context closely tied to the students’ local environment, which facilitated deeper engagement and understanding. The evaluation revealed that students progressed in their scientific inquiry skills, moving from “pre-scientific” to “uncertain inquirer” performance levels. While challenges such as initial disorientation and difficulties in representing experimental data were observed, the overall performance of students demonstrated the success of the TLS. Furthermore, the students worked collaboratively, contributing their individual skills and experiences to achieve group goals. This study provides valuable insights into the potential of TLS as an alternative to traditional teaching methods, offering an innovative way to assess and enhance students’ scientific skills. It also highlights the importance of teacher guidance in inquiry-based activities and suggests that future projects could benefit from allowing students to choose the topic, further enhancing their motivation and engagement. Full article

Pangu is an AI-based model designed for rapid and accurate numerical weather forecasting. To evaluate Pangu’s short- to medium-term weather forecasting skill over various meteorological parameters, this paper validated its performance in predicting temperature, wind speed, wind direction, and barometric pressure using data from over 2000 weather stations in China. Pangu’s performance was compared with ECMWF-HRES and GFS to assess its effectiveness relative to traditional high-precision NWP models under real meteorological conditions. Furthermore, the more recent FuXi and FengWu models were included in the analysis to further validate Pangu’s forecasting skill. The study examined Pangu’s forecast performance from spatial perspectives, evaluated the dispersion of forecast deviations, and analyzed its performance at different lead times and with various initial fields. The iteration precision of Pangu’s four forecast models with lead times of 1 h, 3 h, 6 h, and 24 h was also assessed. Finally, a case study on typhoon track forecasting was conducted to evaluate Pangu’s performance in predicting typhoon paths. The results indicate that Pangu surpasses traditional NWP systems in temperature forecasting, while its performance in predicting wind direction, wind speed and pressure is comparable to them. Additionally, the forecast skill of Pangu diminishes as the lead time extends, but it tends to surpass traditional NWP systems with longer lead times. Moreover, FuXi and FengWu demonstrate even higher accuracy compared to Pangu. Pangu’s performance is also dependent on initial fields, and the temperature forecasting of Pangu is more sensitive to the initial field compared with other meteorological parameters. Furthermore, the iteration precision of Pangu’s 1 h forecast model is significantly lower than that of the other models, but this discrepancy in precision may not be prominently reflected in Pangu’s actual forecasting process due to the greedy algorithm employed. In the case study on typhoon forecasting, Pangu, along with FuXi and FengWu, demonstrates comparable performance in predicting Bebinca’s track compared to ECMWF and outperforms GFS in its track predictions. This study demonstrated Pangu’s applicability in short- to medium-term forecasting of meteorological parameters, showcasing the significant potential of AI-based numerical weather models in enhancing forecast performance. Full article

Electricity price forecasting has been a topic of significant interest since the deregulation of electricity markets worldwide. The New Zealand electricity market is run primarily on renewable fuels, and so weather metrics have a significant impact on electricity price and volatility. In this paper, we employ a mixed-frequency vector autoregression (MF-VAR) framework where we propose a VAR specification to the reverse unrestricted mixed-data sampling (RU-MIDAS) model, called RU-MIDAS-VAR, to provide point forecasts of half-hourly electricity prices using several weather variables and electricity demand. A key focus of this study is the use of variational Bayes as an estimation technique and its comparison with other well-known Bayesian estimation methods. We separate forecasts for peak and off-peak periods in a day since we are primarily concerned with forecasts for peak periods. Our forecasts, which include peak and off-peak data, show that weather variables and demand as regressors can replicate some key characteristics of electricity prices. We also find the MF-VAR and RU-MIDAS-VAR models achieve similar forecast results. Using the LASSO, adaptive LASSO, and random subspace regression as dimension-reduction and variable selection methods helps to improve forecasts where random subspace methods perform well for large parameter sets while the LASSO significantly improves our forecasting results in all scenarios. Full article

To mitigate the prevalence of highway accidents in Southwest China during adverse weather conditions, this study introduces a novel method for identifying accident-prone sections in complex meteorological circumstances. The technique, anchored in data mining’s support index, pioneers the concept of meteorological responsiveness, which includes the elucidation of its mechanisms and the development of computational methodologies. Historical meteorological data and accident records from mountainous highways were meticulously analyzed to quantify the spectrum of adverse weather impacts on driving risks. By integrating road geometry, weather data, and accident site information, meteorological events were identified, categorized, and assigned a meteorological responsiveness score. Outlier sections were processed for preliminary screening, enabling the identification of high-risk segments. The Meteorological Response Ratio Index was instrumental in highlighting and quantifying the influence of adverse weather on traffic safety, facilitating the prioritization of critical sections. The case study of the SC2 highway in Southwest China validated the method’s feasibility, successfully pinpointing eight high-risk sections significantly affected by adverse weather, which constituted approximately 19.05% of the total highway length. Detailed analysis of these sections, especially those impacted by rain, fog, and snow, revealed specific zones prone to accidents. The meteorological responsiveness method’s efficacy was further substantiated by correlating accident mechanisms under adverse weather with the road geometry of key sections. This approach stands to significantly enhance the safety management of operational highways. Full article

Last-mile logistics is one of the most complicated operations in the whole logistic process. This concept describes the final leg of a product travel from a warehouse or hub to specific customers. One of the last-mile logistics challenges that courier delivery companies face is route planning. Ineffective route planning can cause operational delays that cascade and affect several last-mile deliveries. Thus, numerous factors need to be considered to plan and optimize effective delivery routes. These involve many extraordinary and unpredictable events, including weather, traffic conditions, and traffic regulations. A lack of accessible data hinders dynamic, efficient, and reliable route planning, leading to these factors being overlooked. In this paper, we propose the use of open data (OD) to overcome these limitations. OD are information available for anyone to access, reuse, and distribute for free with minimal attribution and sharing restrictions. Therefore, the aim of this work is to examine the impact of incorporating specific open data elements on the performance of the Clarke and Wright algorithm, particularly in calculating savings, and identifying optimal routes. The results we obtained showed the effect of considering OD with an increase rate of approximately 2% on the total distance compared to not considering them. Full article

This study aimed at systematically exploring the seasonalities of bacterial identifications from 1 February 2014 to 31 January 2020 in hospitalized patients, considering the infectious site and the community-acquired or hospital-associated origin. Bacterial identifications were extracted from the data warehouse of the Institut Hospitalo-Universitaire Mediterranée Infection surveillance system, along with their epidemiological characteristics. Each species’ series was processed using a scientific workflow based on the TBATS time series model. Possible co-seasonalities were researched using seasonal peak clustering and series cross-correlations. In this study, only the 15 most frequent species were described in detail. The three most frequent species were Escherichia coli, Staphylococcus aureus, and Staphylococcus epidermidis, with median weekly incidences of 145, 74, and 39 cases, respectively. Samplings of S. aureus and E. coli follow the same seasonal dynamics. S. aureus hospital-associated infections exhibited a significant association with temperature, humidity, and pressure change, whereas community-acquired infections were only associated with precipitations. More seasonal peaks were observed during the winter season. Among the 15 peaks of this seasonal maximum, 6.7% came from blood (Klebsiellia oxytoca) and 13.3% from respiratory specimens (E. coli and S aureus). Our results showed significant associations of periodicity between pathogens, origin of infection, type of sampling, and weather drivers. Full article

Accurate photovoltaic (PV) power forecasting is crucial for effective smart grid management, given the intermittent nature of PV generation. To address these challenges, this paper proposes the Temporal Bottleneck-enhanced Bidirectional Temporal Convolutional Network with Multi-Head Attention and Autoregressive (TB-BTCGA) model. It introduces a temporal bottleneck structure and Deep Residual Shrinkage Network (DRSN) into the Temporal Convolutional Network (TCN), improving feature extraction and reducing redundancy. Additionally, the model transforms the traditional TCN into a bidirectional TCN (BiTCN), allowing it to capture both past and future dependencies while expanding the receptive field with fewer layers. The integration of an autoregressive (AR) model optimizes the linear extraction of features, while the inclusion of multi-head attention and the Bidirectional Gated Recurrent Unit (BiGRU) further strengthens the model’s ability to capture both short-term and long-term dependencies in the data. Experiments on complex datasets, including weather forecast data, station meteorological data, and power data, demonstrate that the proposed TB-BTCGA model outperforms several state-of-the-art deep learning models in prediction accuracy. Specifically, in single-step forecasting using data from three PV stations in Hebei, China, the model reduces Mean Absolute Error (MAE) by 38.53% and Root Mean Square Error (RMSE) by 33.12% and increases the coefficient of determination (R2) by 7.01% compared to the baseline TCN model. Additionally, in multi-step forecasting, the model achieves a reduction of 54.26% in the best MAE and 52.64% in the best RMSE across various time horizons. These results underscore the TB-BTCGA model’s effectiveness and its strong potential for real-time photovoltaic power forecasting in smart grids. Full article

Tornadoes are severe convective weather exhibiting localized and sudden occurrences. Weather radar is widely regarded as the most effective tool for monitoring tornadoes and issuing early warnings. Dual-polarization updating has significantly improved tornado prediction and forecasting abilities. This article proposes an innovative tornado detection algorithm based on XGBoost which is suitable for dual-polarization radar data, was upgraded and has been used in China since 2019, and has been applied in the Tornado Key Open Laboratory of the China Meteorological Administration. The characteristics associated with the velocity attributes, reflectivity, velocity spectrum width, differential reflectivity, and correlation coefficient are used in the algorithm to achieve real-time tornado detection. Experimental evaluations have demonstrated that the proposed algorithm significantly improves tornado detection rates and leading times. Compared with the traditional TDA-RF algorithm based on Doppler weather radar data, the TDA-XGB algorithm introduces dual polarization parameters (such as differential reflectivity and the correlation coefficient), which effectively improves tornado identification performance. In addition, the TDA-XGB algorithm combines artificial intelligence-assisted learning to optimize the traditional algorithm based on the tornado vortex signature (TVS) and tornado debris signature (TDS), further improving the detection effect. Furthermore, the algorithm provides classification probabilities in the genesis and evolution of tornadoes, thereby supporting forecasters in their efforts to anticipate and issue timely tornado warnings. Full article

The unpredictable nature of renewable energy sources, such as wind and solar, makes them unreliable sources of energy for the power system. Nevertheless, with the advancement in the field of artificial intelligence (AI), one can predict the availability of solar and wind energy in the short, medium, and long term with fairly high accuracy. As such, this research work aims to develop a machine-learning-based framework for forecasting global horizontal irradiance (GHI) for Muscat, Oman. The proposed framework includes a data preprocessing stage, where the missing entries in the acquired data are imputed using the mean value imputation method. Afterward, data scaling is carried out to avoid the overfitting/underfitting of the model. Features such as the GHI cloudy sky index, the GHI clear sky index, global normal irradiance (GNI) for a cloudy sky, GNI for a clear sky, direct normal irradiance (DNI) for a cloudy sky, and DNI for a clear sky are extracted. After analyzing the correlation between the abovementioned features, model training and the testing procedure are initiated. In this research, different models, named Linear Regression (LR), Support Vector Machine (SVR), KNN Regressor, Decision Forest Regressor, XGBoost Regressor, Neural Network (NN), Autoregressive Integrated Moving Average (ARIMA), Long Short-Term Memory (LSTM), Random Forest Regressor, Categorical Boosting (CatBoost), Deep Autoregressive (DeepAR), and Facebook Prophet, are trained and tested under both identical features and a training–testing ratio. The model evaluation metrics used in this study include the mean absolute error (MAE), the root mean squared error (RMSE), R², and mean bias deviation (MBD). Based on the outcomes of this study, it is concluded that the Facebook Prophet model outperforms all of the other utilized conventional machine learning models, with MAE, RMSE, and R² values of 9.876, 18.762, and 0.991 for the cloudy conditions and 11.613, 19.951 and 0.988 for the clean weather conditions, respectively. The mentioned error values are the lowest among all of the studied models, which makes Facebook Prophet the most accurate solar irradiance forecasting model for Muscat, Oman. Full article

The conservation of the olive ridley turtle (Lepidochelys olivacea) is increasingly critical due to declining global populations. This study investigates the influence of hydrometeorological conditions on the nesting season and annual hatchling sex ratios conducted at the Playón de Mismaloya Federal Reserve in Tomatlán, Jalisco, Mexico. The research specifically examines variations in sand temperature at both the beach surface and nesting depths, with extended measurements taken at multiple depths (20, 40, 60, 80, and 100 cm) to analyze the vertical temperature gradient along the beach. Atmospheric parameters were modeled using Newton’s Cooling Law and solved with the finite difference method to estimate heat loss rates from beach sand to its surroundings, shedding light on microclimatic effects on incubation and embryonic development. Meteorological data were gathered from an automatic weather station, while sand temperatures were monitored with thermographs. During the warm period (approximately 32 °C), sand temperature showed a negative correlation with depth (20–100 cm), indicating cooler temperatures at greater depths. These conditions were associated with female-biased hatchling production. Conversely, the cold period (approximately 28 °C) led to male-biased hatchling production, with a positive correlation between sand and air temperatures. This study emphasizes the importance of monitoring in situ environmental conditions and extending the protection season until February to avoid the loss of male hatchlings. Full article

This study analyzes the relationships between the onset, the duration, and the maximum extent of the Atlantic Cold Tongue (ACT) using ERA5 reanalysis data from the European Centre for Medium-Range Weather Forecasts (ECMWF) over the period 1979–2019. After calculating the start and end dates of the ACT each year, this study investigates potential relationships between early or late onset that may be linked to the maximum duration and extent of the ACT, which is known to influence weather patterns and precipitation in surrounding regions and the West African Monsoon System. Results show that 68% of years with a short ACT duration are associated with a late-onset ACT, while 70% of years with a long ACT duration are associated with early ACT onset years. In addition, 63% of years with a short duration of ACT have a cold tongue with a low maximum extent, while 83% of years with a long duration of ACT have a cold tongue with a greater maximum extent. Finally, 78% of early ACT onset years are associated with the coldest SST tongue in the eastern equatorial Atlantic Ocean. A comparison of the last 20 years (1999–2019) with the previous 20 years (1979–1998) shows a cooling trend in SST, with ACT occurring and ending earlier in recent years than in the past. However, as the changes in the end date are greater than those in the onset date, the duration of the ACT has been 5–12 days shorter in the last 20 years than in the previous 20 years. Knowledge of these ACT characteristics and their interrelations and drivers is crucial for understanding the West African Monsoon System and for improving climate models and seasonal forecasts. Full article

Traffic accidents occur frequently, causing significant losses to people’s lives and property safety. Accurately predicting the severity level of traffic accidents is of great significance. Based on traffic accident data, this study comprehensively considers various influencing factors such as the geographical location, road conditions, and environment. The data are divided into accident-related categories, weather-related categories, and road- and environment-related categories. The machine learning method is improved through integration for the accident level prediction. In the experiment, effective preprocessing measures were taken for problems such as data imbalance, missing values, the encoding of categorical variables, and the standardization of numerical features. The unbalanced distribution of “Severity” was improved through under-sampling and over-sampling techniques. Firstly, we adopted a multi-stage fusion strategy. A multi-layer perceptron (MLP) was used for the preliminary prediction, and then its result was combined with the original features to form a new feature. Decision tree, XGBoost, and random forest algorithms, respectively, were applied for the secondary prediction. The analysis results show that the improved machine learning model is significantly superior to a single model in the overall performance. The “MLP + random forest” model performs well in evaluation indicators such as the accuracy, recall rate, and F1 value. The accuracy rate is as high as 94%. In the prediction of different traffic accident severity levels (minor, moderate, and severe), the improved machine learning model also generally shows better performance and stability. The research results of this study have broad prospects in the field of intelligent driving. It can realize real-time accident prediction and early warnings, and provide decision support for drivers and autonomous driving systems. The research also provides a scientific basis for traffic planning and management departments to improve driving conditions and reduce the probability and losses of traffic accidents. Full article

Building optimisation techniques provide a rigorous framework for exploring new optimal design solutions. In this study, a genetic algorithm (GA) was used to investigate the energy efficiency of a vernacular architectural element (Rawshan) in Saudi Arabia. Two objectives were optimised using a GA simulation enhanced: energy consumption optimisation and useful daylight illuminance (UDI) optimisation. A calibrated simulation model of a typical house in Saudi Arabia was used in the study. Several metrics, such as light interference from shadows or other windows, were considered to indicate the importance of the Rawshan. Computational studies were performed using different climatic conditions, and the results were compared with and without a Rawshan element using the weather data of Mecca, Jeddah, Riyadh, and Al-Baha. In this study, the blind thicknesses on the front and sides of the Rawshan were used as optimisation variables. The results showed that using a GA with energy consumption as an objective can reduce energy consumption. One of the methods proposed in the paper can reduce energy consumption by 3.6%, 3.6%, and 16.6% for Mecca, Riyadh, and Al-Baha, respectively. The single-objective optimisation method demonstrated that Rawshan provided sufficient UDI in four cities: Mecca, Jeddah, Riyadh, and Al-Baha. The research provided optimised values for Rawshan blind thicknesses on the front and lateral sides under different optimisation constraints. The results showed that using Rawshans in modern building architecture can reduce energy consumption and improve useful daylight illuminance. Full article