Search Results (372)

This paper presents a Systematic Mapping Study (SMS) on data-driven approaches in flood forecasting from 2019 to 2024, a period marked by transformative developments in Deep Learning (DL) technologies. Analysing 363 selected studies, this paper provides an overview of the technological evolution in this field, methodological approaches, evaluation practices and geographical distribution of studies. The study revealed that meteorological and hydrological factors constitute approximately 76% of input variables, with rainfall/precipitation and water level measurements forming the core predictive basis. Long Short-Term Memory (LSTM) networks emerged as the dominant algorithm (21% of implementations), whilst hybrid and ensemble approaches showed the most dramatic growth (from 2% in 2019 to 10% in 2024). The study also revealed a threefold increase in publications during this period, with significant geographical concentration in East and Southeast Asia (56% of studies), particularly China (36%). Several research gaps were identified, including limited exploration of graph-based approaches for modelling spatial relationships, underutilisation of transfer learning for data-scarce regions, and insufficient uncertainty quantification. This SMS provides researchers and practitioners with actionable insights into current trends, methodological practices, and future directions in data-driven flood forecasting, thereby advancing this critical field for disaster management. Full article

For effective groundwater resource management, it is essential to model the dynamic behaviour of aquifers in response to rainfall. Here, a methodological approach using a recurrent neural network, specifically a Long Short-Term Memory (LSTM) network, is used to model groundwater levels of the shallow porous aquifer in Southern Italy. This aquifer is recharged by local rainfall, which exhibits minimal variation across the catchment in terms of volume and temporal distribution. To gain a deeper understanding of the complex interactions between precipitation and groundwater levels within the aquifer, we used water level data from six wells. Although these wells were not directly correlated in terms of individual measurements, they were geographically located within the same shallow aquifer and exhibited a similar hydrogeological response. The trained model uses two variables, rainfall and groundwater levels, which are usually easily available. This approach allowed the model, during the training phase, to capture the general relationships and common dynamics present across the different time series of wells. This methodology was employed despite the geographical distinctions between the wells within the aquifer and the variable duration of their observed time series (ranging from 27 to 45 years). The results obtained were significant: the global model, trained with the simultaneous integration of data from all six wells, not only led to superior performance metrics but also highlighted its remarkable generalization capability in representing the hydrogeological system. Full article

Coastal blue carbon ecosystems (traditional types such as mangroves, salt marshes, and seagrass meadows; emerging types such as tidal flats and mariculture) play pivotal roles in capturing and storing atmospheric carbon dioxide. Reliable assessment of the spatial and temporal variation and the carbon storage potential holds immense promise for mitigating climate change. Although previous field surveys and regional assessments have improved the understanding of individual habitats, most studies remain site-specific and short-term; comprehensive, multi-decadal assessments that integrate all major coastal blue carbon systems at the national scale are still scarce for China. In this study, we integrated 30 m Landsat imagery (1992–2022), processed on Google Earth Engine with a random forest classifier; province-specific, literature-derived carbon density data with quantified uncertainty (mean ± standard deviation); and the InVEST model to track coastal China’s mangroves, salt marshes, tidal flats, and mariculture to quantify their associated carbon stocks. Then the GeoDetector was applied to distinguish the natural and anthropogenic drivers of carbon stock change. Results showed rapid and divergent land use change over the past three decades, with mariculture expanded by 44%, becoming the dominant blue carbon land use; whereas tidal flats declined by 39%, mangroves and salt marshes exhibited fluctuating upward trends. National blue carbon stock rose markedly from 74 Mt C in 1992 to 194 Mt C in 2022, with Liaoning, Shandong, and Fujian holding the largest provincial stock; Jiangsu and Guangdong showed higher increasing trends. The Normalized Difference Vegetation Index (NDVI) was the primary driver of spatial variability in carbon stock change (q = 0.63), followed by precipitation and temperature. Synergistic interactions were also detected, e.g., NDVI and precipitation, enhancing the effects beyond those of single factors, which indicates that a wetter climate may boost NDVI’s carbon sequestration. These findings highlight the urgency of strengthening ecological red lines, scaling climate-smart restoration of mangroves and salt marshes, and promoting low-impact mariculture. Our workflow and driver diagnostics provide a transferable template for blue carbon monitoring and evidence-based coastal management frameworks. Full article

A reliable precipitation dataset with high spatial resolution is essential for climate research in the Tarim Basin. This study evaluated the performances of four models, namely a random forest (RF), a long short-term memory network (LSTM), a support vector machine (SVM), and a feedforward neural network (FNN). FNN, which was found to be superior to the other models, was used to integrate eight precipitation datasets spanning from 1990 to 2022 across the Tarim Basin, resulting in a new monthly high-resolution (0.1°) precipitation dataset named MoHiPr-TB. This dataset was subsequently bias-corrected by the China Land Data Assimilation System version 2.0 (CLDAS2.0). Validation results indicate that the corrected MoHiPr-TB not only accurately reflects the spatial distribution of precipitation but also effectively simulates its intensity and interannual and seasonal variations. Moreover, MoHiPr-TB is capable of detecting the precipitation–elevation relationship in the Pamir Plateau, where precipitation initially increases and then decreases with elevation, as well as the synchronous variation of precipitation and elevation in the Tianshan region. Collectively, this study delivers a high-accuracy precipitation dataset for the Tarim Basin, which is anticipated to have extensive applications in meteorological, hydrological, and ecological research. Full article

Accurate short-term precipitation forecasting (nowcasting) based on weather radar data is essential for managing weather-related risks, particularly in applications such as airport operations, urban flood prevention, and public safety during outdoor events. This study proposes a computationally efficient nowcasting method based on a Lagrangian advection scheme, estimating both the translation and rotation of radar-observed precipitation fields without relying on machine learning or resource-intensive computation. The method was tested on a two-year dataset (2022–2023) over Tuscany, using data collected from the Italian Civil Protection Department’s radar network. Forecast performance was evaluated using the Critical Success Index (CSI) and Mean Absolute Error (MAE) across varying spatial domains (1° × 1° to 2° × 2°) and precipitation regimes. The results show that, for high-intensity events (average rate > 1 mm/h), the method achieved CSI scores exceeding 0.5 for lead times up to 2 h. In the case of low-intensity rainfall (average rate < 0.3 mm/h), its forecasting skill dropped after 20–30 min. Forecast accuracy was shown to be highly sensitive to the temporal stability of precipitation intensity. The method performed well under quasi-stationary stratiform conditions, whereas its skill declined during rapidly evolving convective events. The method has low computational requirements, with forecasts generated in under one minute on standard hardware, and it is well suited for real-time application in regional meteorological centres. Overall, the findings highlight the method’s effective balance between simplicity and performance, making it a practical and scalable option for operational nowcasting in settings with limited computational capacity. Its deployment is currently being planned at the LaMMA Consortium, the official meteorological service of Tuscany. Full article

The Advanced Scatterometer (ASCAT), Soil Moisture Active Passive (SMAP), and European Space Agency-Climate Change Initiative (ESA–CCI) soil moisture (SM) products are widely used in agricultural drought monitoring, water resource management, and climate analysis applications. However, the performance of these SM products varies significantly across regions and environmental conditions, due to in sensor characteristics, retrieval algorithms, and the lack of localized calibration. This study proposes a multi-source long short-term memory (MLSTM) for improving ASCAT, ESA–CCI, and SMAP SM products by combining in-situ SM measurements and four key auxiliary variables: precipitation (PRE), land surface temperature (LST), fractional vegetation cover (FVC), and evapotranspiration (ET). First, the in-situ measured data from four in-situ observation networks were corrected using the LSTM method to match the grid sizes of ASCAT (0.1°), ESA–CCI (0.25°), and SMAP (0.1°) SM products. The RPE, LST, FVC, and ET were used as inputs to the LSTM to obtain loss data against in-situ SM measurements. Second, the ASCAT, ESA–CCI, and SMAP SM datasets were used as inputs to the LSTM to generate loss data, which were subsequently corrected using LSTM-derived loss data based on in-situ SM measurements. When the mean squared error (MSE) loss values were minimized, the improvement for ASCAT, ESA–CCI, and SMAP products was considered the best. Finally, the improved ASCAT, ESA–CCI, and SMAP were produced and evaluated by the correlation coefficient (R), root mean square error (RMSE), and standard deviation (SD). The results showed that the RMSE values of the improved ASCAT, ESA–CCI, and SMAP products against the corrected in-situ SM data in the OZNET network were lower, i.e., 0.014 cm³/cm³, 0.019 cm³/cm³, and 0.034 cm³/cm³, respectively. Compared with the ESA–CCI and SMAP products, the ASCAT product was greatly improved, e.g., in the SNOTEL network, the Root Mean-Square Deviation (RMSD) values of 0.1049 cm³/cm³ (ASCAT) and 0.0662 cm³/cm³ (improved ASCAT). Overall, the MLSTM-based algorithm has the potential to improve the global satellite SM product. Full article

The Standardized Precipitation Index (SPI) is widely used to determine drought severity worldwide. However, inconsistencies exist regarding its application in warm, humid tropical climatic zones. Originally developed for temperate regions with a continental climate, the index may not adequately reflect drought conditions in tropical environments where rainfall regimes differ substantially. This study identifies the following two principal reasons why the traditional calculation method fails to characterize drought severity in tropical domains: first, the marked humidity contrast between the consistently humid rainy season and the rest of the year, and second, the diverse drought types in tropical regions, which include both long-term and short-term events. Using data from meteorological stations in Mexico’s humid tropics and comparing them with temperate regions, the study demonstrates significant discrepancies between SPI-based drought classifications and actual precipitation patterns. Our analysis shows that the abundant precipitation during the rainy season causes biases in longer time scales integrated into multivariate drought indices. Considerations are established for adapting the SPI for decision makers who monitor drought in humid tropics, with specific recommendations on time scale limits to avoid biases. This work contributes to more accurate drought monitoring in tropical regions by addressing the unique climatic characteristics of these environments. Full article

Accurately forecasting karst spring discharge remains a significant challenge due to the inherent nonstationarity and multi-scale hydrological dynamics of karst hydrological systems. This study presents a physics-informed variational mode decomposition long short-term memory (VMD-LSTM) model, enhanced with an attention mechanism and Monte Carlo dropout for uncertainty quantification. Hourly discharge data (2013–2018) from the Zhaidi karst spring in southern China were decomposed using VMD to extract physically interpretable temporal modes. These decomposed modes, alongside precipitation data, were input into an attention-augmented LSTM incorporating physics-informed constraints. The model was rigorously evaluated against a baseline standalone LSTM using an 80% training, 15% validation, and 5% testing data partitioning strategy. The results demonstrate substantial improvements in prediction accuracy for the proposed framework compared to the standard LSTM model. Compared to the baseline LSTM, the RMSE during testing decreased dramatically from 0.726 to 0.220, and the NSE improved from 0.867 to 0.988. The performance gains were most significant during periods of rapid conduit flow (the peak RMSE decreased by 67%) and prolonged recession phases. Additionally, Monte Carlo dropout, using 100 stochastic realizations, effectively quantified predictive uncertainty, achieving over 96% coverage in the 95% confidence interval (CI). The developed framework provides robust, accurate, and reliable predictions under complex hydrological conditions, highlighting substantial potential for supporting karst groundwater resource management and enhancing flood early-warning capabilities. Full article

Short-term precipitation forecasting is a core task in meteorological science, aiming to achieve accurate predictions by modeling the spatiotemporal evolution of radar echo sequences, thereby supporting meteorological services and disaster warning systems. However, existing spatiotemporal sequence prediction methods still struggle to disentangle complex spatiotemporal dependencies effectively and fail to capture the nonlinear chaotic characteristics of precipitation systems. This often results in ambiguous predictions, attenuation of echo intensity, and spatial localization errors. To address these challenges, this paper proposes a unified spatiotemporal sequence prediction framework based on spatiotemporal masking, which comprises two stages: self-supervised pre-training and task-oriented fine-tuning. During pre-training, the model learns global structural features of meteorological systems from sparse contexts by randomly masking local spatiotemporal regions of radar images. In the fine-tuning stage, considering the importance of the temporal dimension in short-term precipitation forecasting and the complex long-range dependencies in spatiotemporal evolution of precipitation systems, we design an RNN-based cyclic temporal mask self-encoder model (MAE-RNN) and a transformer-based spatiotemporal attention model (STMT). The former focuses on capturing short-term temporal dynamics, while the latter simultaneously models long-range dependencies across space and time via a self-attention mechanism, thereby avoiding the smoothing effects on high-frequency details that are typical of conventional convolutional or recurrent structures. The experimental results show that STMT improves 3.73% and 2.39% in CSI and HSS key indexes compared with the existing advanced models, and generates radar echo sequences that are closer to the real data in terms of air mass morphology evolution and reflection intensity grading. Full article

The study presents a scenario analysis of the long-term dynamics of the water level of Lake Balkhash, one of the largest closed lakes in Central Asia, taking into account climate change according to CMIP6 scenarios (SSP2-4.5 and SSP5-8.5) and socio-economic factors of water use. Based on historical data (1947–2021) and a water balance model, the contribution of surface runoff, precipitation and evaporation to the formation of the lake’s hydrological regime was assessed. It was established that the main source of water resources for the lake is the flow of the Ile River, which feeds the western part of the reservoir. The eastern part is characterized by extremely limited water inflow, while evaporation remains the main element of water consumption, having increased significantly in recent decades due to rising air temperatures. Increasing intra-seasonal and interannual fluctuations in water levels have been recorded: The amplitude of short-term fluctuations reached 0.7–0.8 m, which exceeds previously characteristic values. The results of water balance modeling up to 2050 show a trend towards a 30% reduction in surface inflow and an increase in evaporation by 25% compared to the 1981–2010 climate norm, which highlights the high sensitivity of the lake’s hydrological regime to climatic and anthropogenic influences. The results obtained justify the need for the comprehensive and adaptive management of water resources in the Balkhash Lake basin, taking into account the transboundary nature of water use and changing climatic conditions. Full article

This study compares three Standardized Precipitation and Evapotranspiration Index (SPEI) prediction models at different time scales: (1) Kalman filter with exogenous variables (DKF-ARX-Pt, FK), (2) gated recurrent unit (GRU), and (3) autoregressive neural networks with external input (NARX). Using observed data from meteorological stations in the State of Mexico and Mexico City, considering performance metrics, such as mean absolute error (MAE), mean square error (MSE), root mean square error (RMSE), coefficient of determination (R²), Nash–Sutcliffe efficiency coefficient (NSE) and Kling–Gupta efficiency (KGE). The results indicate that the FK model with exogenous variables is the most accurate model for SPEI prediction at different time scales, standing out in terms of stability and low variance in prediction error. GRU networks showed acceptable performance on long time scales (SPEI12 and SPEI24), but with lower stability on short scales. In contrast, NARX presented the worst performance, with high errors and negative efficiency coefficients at several time scales. Models based on Kalman filters can be key tools to improve drought mitigation strategies in vulnerable regions, as it has an improved average predictive accuracy by reducing the MAE by up to 68% and achieving higher consistency in KGE values at longer time scales (SPEI12 and SPEI24). Full article

Road traffic accidents pose a substantial global public health burden, resulting in significant fatalities and economic costs. This study employs Bayesian Poisson regression to model traffic accident rates in Poland, focusing on the intricate relationships between weather conditions and socioeconomic factors. Analyzing both yearly county-level and weekly nationwide data from 2020 to 2023, we created four distinct models examining the relationships between accident occurrence and predictors including temperature, humidity, precipitation, population density, passenger car registrations, and road infrastructure. Model evaluation, based on WAIC and PSIS-LOO criteria, demonstrated that integrating both weather and socioeconomic variables enhanced predictive accuracy. Results showed that socioeconomic variables—especially passenger car registrations—were strong predictors of accident rates over longer timeframes and across localized regions. In contrast, weather variables, particularly temperature and humidity, were more influential in explaining short-term fluctuations in nationwide accident counts. These findings provide a statistical foundation for identifying high-risk conditions and guiding targeted interventions. The study supports Poland’s national road safety goals by offering evidence-based strategies to reduce accident-related fatalities and injuries. Full article

Accurately predicting streamflow using process-based models remains challenging due to uncertainties in model parameters and the complex nature of streamflow generation. Data-driven approaches, however, offer feasible alternatives, avoiding the need for physical process representation. This study introduces a hybrid deep learning framework, CNN-GRU-BiLSTM, for daily streamflow prediction. This model integrates convolutional neural networks (CNN), gated recurrent units (GRU), and bidirectional long short-term memory (BiLSTM) networks to leverage their complementary strengths. When applied to the Neuse River Basin (NRB) (North Carolina, USA), the proposed model achieved strong predictive performance, yielding a root mean square (RMSE) of 11.8 m³/s (compared to an average streamflow of 132.7 m³/s), and a mean absolute error (MAE) of 8.7 m³/s, and a Nash–Sutcliffe efficiency (NSE) of 0.994 for the testing dataset. Similar performance trends were observed in the training and validation phases. A comparative analysis against seven other deep learning and hybrid models of similar complexity highlighted the outstanding performance of the CNN-GRU-BiLSTM model across all flood events. Furthermore, its stability, robustness, and transferability were evaluated in a seasonal dataset, peak floods, and different locations along the river. These findings underscore the potential of hybrid deep learning models and reinforce the effectiveness of integrating multiple data-driven techniques for streamflow prediction in regions where precipitation is the dominant driver of streamflow. Full article

Climate change is a pressing global issue with severe consequences for the planet and human health. The Earth’s temperature has risen by 2 °C from 1901 to 2023, and this warming trend is expected to continue, causing potentially dangerous shifts in climate. Climate change impacts are already visible, with more frequent and severe heat waves, droughts, intense rain, and floods becoming increasingly common. Therefore, hydropower can contribute to addressing the global climate change issue and help to achieve global energy transition and stabilize global energy security. A Long Short-Term Memory (LSTM)-based model implemented in Python for global and regional hydropower forecasting was developed for a study period of 2023 to 2060 by taking the input data from 1980 to 2022. The results revealed that Asian countries have greater hydropower potential, which is expected to increase from 1926.51 TWh in 2023 to 2318.78 TWh in 2030, 2772.06 TWh in 2040, 2811.41 TWh in 2050, and 3195.79 TWh in 2060, as compared with the other regions of the world like the Middle East, Africa, Asia, Common Wealth of Independent State (CIS), Europe, North America, and South and Central America. The global hydropower potential is also expected to increase from 4350.12 TWh in 2023 to 4806.26 TWh in 2030, 5393.80 TWh in 2040, 6003.53 TWh in 2050, and 6644.06 TWh in 2060, which is sufficient for achieving energy transition and energy security goals. Furthermore, the performance and accuracy of the LSTM-based model were found to be 98%. This study will help in the efficient scheduling and management of hydropower resources, reducing uncertainties caused by environmental variability such as precipitation and runoff. The proposed model contributes to the energy transition and security that is needed to meet the global climate targets. Full article

Sustained reductions in terrestrial water storage (TWS) have been observed globally using Gravity Recovery and Climate Experiment (GRACE) satellite data since 2002. However, the underlying mechanisms remain incompletely understood due to limited record lengths and data discontinuity. Recently, explainable artificial intelligence (XAI) has provided robust tools for unveiling dynamics in complex Earth systems. In this study, we employed a deep learning technique (Long Short-Term Memory network, LSTM) to reconstruct global TWS dynamics, filling gaps in the GRACE record. We then utilized the Local Interpretable Model-agnostic Explanations (LIME) method to uncover the underlying mechanisms driving observed TWS reductions. Our results reveal a consistent decline in the global mean TWS over the past 22 years (2002–2024), primarily influenced by precipitation (17.7%), temperature (16.0%), and evapotranspiration (10.8%). Seasonally, the global average of TWS peaks in April and reaches a minimum in October, mirroring the pattern of snow water equivalent with approximately a one-month lag. Furthermore, TWS variations exhibit significant differences across latitudes and are driven by distinct factors. The largest declines in TWS occur predominantly in high latitudes, driven by rising temperatures and significant snow/ice variability. Mid-latitude regions have experienced considerable TWS losses, influenced by a combination of precipitation, temperature, air pressure, and runoff. In contrast, most low-latitude regions show an increase in TWS, which the model attributes mainly to increased precipitation. Notably, TWS losses are concentrated in coastal areas, snow- and ice-covered regions, and areas experiencing rapid temperature increases, highlighting climate change impacts. This study offers a comprehensive framework for exploring TWS variations using XAI and provides valuable insights into the mechanisms driving TWS changes on a global scale. Full article