Search Results (4,730)

This study examines whether financial development affects economic growth asymmetrically in Algeria using annual data for 1980–2020. Algeria represents a particularly relevant case because its hydrocarbon dependence, bank-based financial system, strong state involvement in banking, and exposure to oil-price and macroeconomic shocks may shape the finance–growth relationship differently from more diversified economies. The study constructs a composite financial development index using principal component analysis of liquid liabilities, private-sector credit, and bank deposits to GDP (%); the first component explains 81.25% of total variance. A nonlinear autoregressive distributed lag (NARDL) model is estimated with trade openness and government expenditure as controls. Unit-root tests confirm that the variables are I(0) or I(1), and the bounds test supports cointegration (F = 7.455). In the long run, trade openness positively affects growth, whereas government expenditure is statistically insignificant. Positive financial-development shocks are associated with lower long-run growth, although the coefficient is significant only at the 10% level, while negative shocks are insignificant. In the short run, positive shocks generate delayed growth benefits after one and two years. Although the Wald test rejects long-run symmetry within the estimated specification, uncertainty around the individual long-run coefficients warrants interpreting the findings as suggestive evidence of asymmetry that requires further confirmation. Short-run symmetry is not rejected. The error-correction coefficient (−0.674) implies rapid adjustment toward equilibrium. The findings suggest that the quality and productive allocation of financial resources may matter more for growth than financial deepening alone. Full article

Evaluating the structural stability of the relationships between inflation, unemployment, and interest rates is vital for central banks striving to fulfill their dual mandates within an evolving economic landscape. This study evaluates whether the empirical dynamics governing these fundamental macroeconomic relationships have undergone significant parameter shifts over the past quarter century. Utilizing quarterly US macroeconomic data from 2001 to 2025, this study employs a recursive vector autoregression (VAR) framework that incorporates the federal funds rate alongside broader measures of labor market slack: the U6 unemployment rate) and decomposed price pressures (demand-driven inflation). Information criteria definitively show that this model offers a superior empirical fit, while the resulting impulse response functions reveal a weak, unstable response of unemployment to policy shocks that directly contradicts traditional historical benchmarks. Ultimately, this study contributes to the literature by demonstrating that fixed-parameter linear frameworks lack the structural robustness required to capture modern post-pandemic macroeconomic interactions, highlighting a critical need for time-varying analytical models. Full article

The Kelani River is one of the most important freshwater resources in Sri Lanka, supplying drinking water, supporting aquatic ecosystems, and sustaining industrial and agricultural activities. However, increasing anthropogenic pressures have intensified the need for comprehensive long-term assessments of river water quality. This study evaluated the spatio-temporal variability of six physicochemical water-quality parameters (pH, temperature, turbidity, chemical oxygen demand, dissolved oxygen, and chloride) using monthly observations from twelve monitoring locations covering January 2007 to May 2022, representing the most up-to-date long-term CEA monitoring record available to the authors at the time of data acquisition. An integrated statistical framework comprising correlation analysis, Granger causality testing, vector autoregressive (VAR) modelling, ARIMA and seasonal ARIMA (SARIMA) forecasting, K-means time-series clustering, Pettitt temporal homogeneity testing, changepoint detection, and regression kriging was applied to investigate temporal dynamics, predictive relationships, temporal stability, and spatial variability. VAR, ARIMA, and SARIMA were selected as interpretable baseline models for assessing lagged dependence and seasonal temporal structure in the monthly water-quality series; comparison with machine-learning and hybrid models was beyond the scope of this study. The best-performing imputation method varied among parameter–location series; for pH, the minimum RMSE values of the selected methods ranged from 0.07 to 0.27. Temporal forecasting performance was also strongly parameter- and location-dependent. For DO, test-set RMSE ranged from 0.69 to 1.33 mgL⁻¹, whereas chloride forecast RMSE ranged from 5.60 to 838.56 mgL⁻¹, with the largest error observed at Victoria Bridge, reflecting the pronounced variability of the downstream chloride series. After Bonferroni correction across the 360 directional parameter–site comparisons, only two Granger-predictive relationships remained statistically significant: COD → chloride at Maha Oya and temperature → pH at Kaduwela Bridge. The corresponding VAR models had R² values of 0.20 and 0.25, respectively, indicating modest explanatory power, while ARIMA and SARIMA models provided satisfactory forecasting performance for several parameters. The Pettitt homogeneity assessment identified nine candidate shifts at the unadjusted 5% significance level, but none remained statistically significant after Bonferroni correction across the 72 parameter–location tests. Regression kriging provided an exploratory representation of spatial variability in water-quality characteristics along the river. Overall, the proposed framework provides a robust statistical approach for long-term river water-quality assessment and supports evidence-based monitoring, pollution management, and sustainable river basin management, contributing to Sustainable Development Goal 6. Full article

Fixed angular separation between LiDAR (Light Detection and Ranging) channels results in different path lengths when the channels observe the same planar target; therefore, geometric effects must be considered when comparing inter-channel range measurements. This study investigated inter-channel range differences in a 2-channel LiDAR with fixed 0° and −3° viewing directions, focusing on short-term temporal dependence and temporal averaging. Repeated static measurements were conducted at 1∼5 m with five independent runs per distance. After the path-length difference expected from the nominal channel geometry was removed as a preprocessing step, the mean inter-channel residual ranged from −0.201 to 1.731 mm. The residual showed short-term positive temporal dependence, with a lag-1 autocorrelation of approximately 0.5 that attenuated toward zero within approximately 0.2∼0.3 s. Temporal averaging reduced residual variability more slowly than the $1/\sqrt{N}$ expectation for independent observations, while an autoregressive model of order 1 (AR(1))-based reference closely reproduced the empirical averaging behavior. These results demonstrate that short-term temporal dependence should be considered when evaluating precision improvement through temporal averaging of repeated inter-channel LiDAR measurements. Full article

Water resources carrying capacity (WRCC) includes the relationships between water resources utilizations and economic development, social demand, and ecological protection. Most previous studies lack WRCC analyses of regions experiencing economic development and industrial development. Therefore, this study takes Jiangsu Province in China as a typical case which includes 13 prefecture-level cities from 2012 to 2021. To provide a comprehensive and nuanced assessment, this study proposes an integrated analytical paradigm for WRCC that explicitly incorporates system coordination, dynamic drivers, and spatiotemporal evolution. First, the entropy-weighted Chebyshev–TOPSIS method is applied to calculate WRCC. Second, the coupling coordination degree model (CCD) is introduced to assess coordination among the four subsystems. Third, the Panel Vector Autoregression model (PVAR) is used to examine dynamic interactions. Fourth, the geographically and temporally weighted regression model (GTWR) is considered to identify spatiotemporal trends with a certain degree of endogeneity. The results indicate that WRCC generally improved across Jiangsu Province over the study period, although marked differences persist among its 13 prefecture-level cities. Suzhou and Nanjing presented relatively high WRCC levels, whereas Huai’an and Suqian showed weaker coordination. The economic subsystem is strongly constrained by water resources and ecological conditions. The impact of the ecological environment on WRCC continues to intensify, which indicates its increasing role in regional water resource management. These findings show that WRCC is jointly shaped by interactions among subsystems and by drivers that vary across space and time. The proposed framework can support water allocation, ecological protection, and development planning tailored to the conditions of individual cities. Full article

Offshore wind turbines equipped with full-power converters (FPCs) require coordinated control of power tracking, electrical safety, and mechanical load mitigation. Nonlinear model predictive control (NMPC) provides a systematic framework for this multi-objective problem; however, standard implementations solve the nonlinear program to full convergence at every control step, incurring computational times incompatible with the 10–100 ms control periods of commercial turbines. Existing formulations also commonly omit electrical hard constraints and rely on expensive lidar for wind speed feedforward. To address these limitations, this paper proposes a real-time iteration NMPC (RTI-NMPC) framework in which each control step performs a truncated real-time iteration, i.e., a single inexact SQP step realized by at most five interior-point (IPOPT) Newton iterations on the parametric NLP, combined with warm-start initialization and a solution-shift strategy. Torque command amplitude bounds and a DC-link voltage deadband are embedded as hard bounds on the predicted trajectory within the OCP, while the torque rate limit is enforced through a dual mechanism consisting of a quadratic penalty in the cost function and a ±15 kN·m/s hard saturation at the solver output. A second-order autoregressive predictor supplies wind speed feedforward over a 2 s horizon, and the power-tracking weight is adaptively adjusted according to the prediction confidence. Ablation experiments confirm that the warm-start solution-shift mechanism is decisive for closed-loop performance: removing it degrades the power-tracking RMSE by up to 370% under grid load-drop transients, collapsing to the level of the higher-iteration-budget standard NMPC, while the adaptive weighting is shown to be intrinsically coupled to the predictor and remains inactive under persistence forecasting. Comparative simulations under four operating scenarios show that the proposed controller achieves an average per-step computation time of 3.796–4.313 ms, approximately one order of magnitude faster than standard NMPC, thereby satisfying the real-time requirement within the present simulation setting. Under a grid load-drop scenario, the root-mean-square power-tracking error is reduced by 78.1% and 78.5% relative to a PI controller and standard NMPC, respectively; under model mismatch, the corresponding reductions are 18.0% and 17.5%. This advantage is scenario- and model mismatch-dependent and partly arises from reduced commitment to an imperfect internal prediction model, rather than the general superiority of truncated optimization over a more fully solved NMPC. The torque rate remains within ±15 kN·m/s in all scenarios. It is noted that the DC-link hard bounds apply to the predicted trajectory within the OCP, while the actual plant voltage is subject to the simplified model dynamics. The results demonstrate the computational feasibility and engineering potential of the electrically constrained, lidar-free RTI-NMPC for coordinated power-tracking and electrical constraint management of offshore FPC wind turbines. Full article

Combining phase change materials (PCM) with a solar chimney is a natural ventilation strategy that extends ventilation duration and reduces energy use linked to global warming. However, its integration with thin fabric window curtains remains understudied compared with building envelope components such as walls or roofs. This study evaluated the thermal performance of a curtain-type PCM solar chimney with PCM thicknesses of 0.001 to 0.004 m and varying collector absorptance, using EnergyPlus simulations across 1533 hourly observations under tropical conditions. An autoregressive model, informed by sensitivity analysis, explained 59.3% of airflow variance (adjusted R² = 0.593, p < 0.001), with no significant residual autocorrelation. The collector-to-room temperature differential was the strongest predictor (10.91 m³/h per 1 °C). A significant PCM–absorptance interaction showed that thin PCM (0.001 m) responds rapidly to solar gain, whereas thick PCM (0.004 m) sustains evening ventilation through latent heat storage. The optimal configuration, consisting of 0.004 m PCM with 0.95 absorptance, increased airflow by 23% and extended ventilation beyond 20 h/day, while low-to-medium absorptance combined with thin PCM maintained 38.4–55.6 m³/h for up to 23 h/day. These findings support a flexible, climate-responsive strategy for sustainable passive cooling in tropical buildings. Full article

Many studies focus on the location of bicycle stations rather than network design, often overlooking the cyclist’s experience, which depends on individual characteristics and the type of bicycle used. This study proposes a methodology for designing bicycle corridors using GPS data and autoregressive models with exogenous variables (ARX). Two experiments were conducted: (1) a fixed-route scenario varying cyclists and bicycles, and (2) a variable-route scenario using the cyclists’ own bicycles. Both experiments evaluated the model’s accuracy and “transferability” using historical datasets to predict subsequent speeds. Results show an R-value of 0.94 for fixed routes and 0.93 for variable routes, with lags of 255 and 139 historical data points, respectively. The best-performing model for fixed routes achieved an MAE of 2.01 km/h, an RMSE of 2.55 km/h, a MAPE of 0.13%, and an R-value of 0.92 with 238 lags; with only 2 lags, the R-value was 0.82. Public and six-speed bicycles demonstrated greater transfer compatibility than electric ones. Finally, the model (using two lags) was applied to the road network in Querétaro to create a preliminary corridor design. The study concludes that while the model is useful—and could better represent cyclist behavior with additional data—it should not be viewed as a universal solution. Full article

This study analyzes the performance of several forecasting models applied to Standardized Precipitation Index (SPI) time series for the 1967–2021 period, using data from six weather stations in southeastern Romania. Four temporal aggregation scales (SPI3, SPI6, SPI12, and SPI24) were considered for one-step-ahead (one-month) forecasting. The evaluated models included Random Forest, XGBoost, Support Vector Regression (SVR), N-BEATS (generic variant), and SARIMA (Seasonal Autoregressive Integrated Moving Average), covering machine learning, deep learning, and statistical approaches. To ensure a rigorous out-of-sample evaluation, SPI distribution parameters were estimated solely based on the pre-test calibration period and held constant when calculating SPI values for the independent test period, thereby preventing information leakage from the preprocessing stage. Forecast performance was assessed using RMSE, MAE, NSE, and a skill score relative to a persistence benchmark, while differences in forecast accuracy were statistically evaluated using the Diebold–Mariano test. The results revealed a clear scale-dependent pattern: absolute forecast accuracy increased with the SPI aggregation scale—alongside more pronounced temporal persistence—though this did not necessarily translate into superior skill compared to the persistence model. Among the data-driven models, SVR demonstrated the most robust performance, particularly at shorter timescales, while N-BEATS also provided competitive forecasts without consistently outperforming SVR. SARIMA showed the clearest gains over the persistence model at the SPI12 and SPI24 scales. The maximum NSE increased from 0.61 for SPI3 and 0.81 for SPI6, both obtained with SVR, to 0.94 for SPI12 (obtained with SARIMA) and 0.97 for SPI24 (obtained with SARIMA, N-BEATS and SVR). In conclusion, the results demonstrate that the SPI accumulation scale, temporal structure, and performance relative to persistence must be considered simultaneously when selecting forecasting models for drought prediction. Full article

Background. Most evidence on inflation–growth nonlinearity relies on cross-country panels or short samples, leaving a gap for a single transition economy on high-frequency data. We address this gap for Vietnam over 1996–2025. Methods. We combine threshold regression with the NARDL framework on quarterly data complemented by a linear ARDL benchmark and a Granger causality pre-test. Results. The endogenous inflation threshold is estimated at 5.7% (95% CI: 4.8–7.2%). Above it, inflation exerts a sharply negative effect on real GDP growth; below it, the effect is statistically indistinguishable from zero. Long-run responses are asymmetric: positive inflation shocks are significantly contractionary (coefficient ≈ −0.26, p < 0.001), while negative shocks show no detectable effect. The threshold is stable across subsamples. Implications. The breakpoint is a conditional-mean estimate, not a welfare optimum; any operational inflation target derived from it should be framed as risk-management guidance. The results support keeping inflation below the estimated threshold band. Full article

Air quality degradation in rapidly urbanizing regions involves complex, non-linear dynamics that frequently elude traditional deterministic models. To tackle this limitation, this research introduces a probabilistic forecasting and transfer station methodology. By reliably extracting patterns from official sensor data, our framework predicts the Air Quality Index (AQI) within the Guadalajara Metropolitan Area, Mexico. The approach utilizes NeuralProphet, a hybrid architecture that integrates deep Auto-Regressive Networks with seasonal decomposition. Based on hourly concentrations of six criteria pollutants (${\mathrm{PM}}_{10}$, ${\mathrm{PM}}_{2.5}$, ${\mathrm{O}}_3$, ${\mathrm{NO}}_2$, ${\mathrm{SO}}_2$, $\mathrm{CO}$) collected from four official monitoring stations between 2022 and 2024, a preprocessing pipeline aligned with US EPA standards was implemented. Optimization identified a 12-h lag window to forecast a 1-h future horizon. To evaluate the model’s generalization, the cross-station framework was trained on an aggregated signal from peripheral stations and tested on the industrial core of Miravalle. The framework was benchmarked against standard deep learning architectures (LSTM, GRU, RNN). While recurrent models exhibited high deterministic accuracy, NeuralProphet maintained robust predictive performance ($R^2\approx 0.995$ on the held-out industrial test set) while additionally providing explicit seasonal interpretability and 80% prediction intervals via quantile regression. This quantification of uncertainty helps distinguish stable diurnal trends from sporadic hazardous events. By providing probabilistic forecasting ranges rather than deterministic point estimates, this transparent, data-driven tool offers a practical option for proactive public health interventions alongside black-box alternatives. Full article

Accurate mapping of building damage after destructive natural disasters is essential for emergency response and recovery planning. However, annotated post-disaster samples are often scarce, and conventional supervised models usually generalize poorly across scenes and hazard types. These constraints limit the operational use of remote sensing-based damage assessment during emergency response. To address this problem, this study proposes SCTSC-VLM, a training-free framework for damage grading of pre-localized building instances, implemented using a locally deployed Q4_K_M 4-bit quantized version of the Qwen3-VL-30B vision–language model. The framework has three main components. First, an instance-level visual input construction strategy integrates independent assessment of building instances localized by the supplied footprint data without a dedicated instance segmentation model. Second, a hierarchical reasoning mechanism decomposes the assessment into four progressive stages: multi-view structured feature extraction, independent visual description, cross-validation, and rule-constrained classification. This design is intended to reduce hallucination and visually ungrounded reasoning. Third, a multi-sampling ensemble decision mechanism mitigates the randomness of autoregressive generation. The framework was evaluated using post-disaster optical imagery from the 2025 Dingri earthquake and the 2025 Eaton Fire in California. On the full Dingri evaluation dataset comprising Areas A–C (n = 398), SCTSC-VLM achieved a Macro-F1 score of 88.2% and a Kappa coefficient of 0.806. On the separate full California evaluation dataset (n = 386), it achieved a Macro-F1 score of 83.8% and a Kappa coefficient of 0.800. Separate ablation experiments were conducted on explicitly defined analysis populations to examine the contributions of the framework components. In both full-dataset evaluations, the fully damaged and undamaged classes obtained higher F1-scores than the partially damaged class. These results provide case-level evidence that structured reasoning can support training-free damage grading of pre-localized buildings, although broader generalization and large-area operational efficiency remain to be established. Full article

The availability of multiple remote sensing-based vegetation index products provides unprecedented power to monitor vegetation dynamics. Accurately quantifying global vegetation trends remains challenging due to inconsistencies among products and the widespread violation of statistical independence caused by spatiotemporal autocorrelation. Here, we identified and statistically tested global patterns of vegetation change trends by comparing multiple representative long-term remote sensing-based vegetation products and using a powerful statistical approach, Partitioned Autoregressive Time Series (PARTS). All remote sensing-based products exhibited strong temporal and spatial autocorrelation. The autoregressive model accounting for temporal autocorrelation detected, on average, 34.4% of pixels with significant trends, markedly lower than linear regression (46.9%) and the Mann–Kendall test (45.2%), which did not account for autocorrelation. To aggregate pixel-level trends to derive overall trends at regional levels, the estimates from linear regression or ANOVA always gave a highly significant test with a p value less than 10⁻¹⁵. This suggests that failing to consider spatial and temporal autocorrelation tends to inflate Type I error rates and give a spurious pattern. Our results demonstrated a significant global greening trend (p < 0.05), with strong consistency across multiple data products. Areas exhibiting high spatial coherence further identified several greening regions, including China, India, Sahel, and so forth. In contrast, trends in the ecosystems of mid to high latitudes exhibited large fluctuations, which were closely related with the differences in the ecological meaning of vegetation indices and the uncertainty of remote-sensing observations in these regions. Our research provides a more robust and spatially coherent characterization of global greening trends, enhancing our understanding of large-scale ecosystem dynamics under environmental change and informing the selection of vegetation indices for ecosystem change assessment. Full article

Atmospheric water vapor is a key driver of weather variability, and accurate forecasts of precipitable water vapor (PWV) can support quantitative precipitation forecasting. Existing PWV forecasting studies have relied mainly on Global Navigation Satellite System (GNSS) observations, which provide high temporal resolution but limited spatial coverage. This study develops a regional multi-lead PWV forecasting framework for Xinjiang, China, by combining Fengyun-4B (FY-4B) Advanced Geosynchronous Radiation Imager (AGRI) observations with ground-based GNSS PWV measurements. FY-4B gridded observations were collocated with 34 GNSS stations using a KD-tree search followed by a Haversine distance calculation, yielding 332,115 valid hourly samples from June 2023 to August 2024. Predictors included geographic and calendar variables, AGRI thermal-infrared brightness temperatures, Level-2 cloud products, physically derived water-vapor features, and multiscale lagged variables. Twenty-four independent LightGBM regressors were trained for lead times from 1 to 72 h, avoiding the error accumulation inherent in autoregressive forecasting. At T + 1 h, the model achieved a mean absolute error (MAE) of 0.72 mm and a coefficient of determination (R²) of 0.92. At T + 12 h, MAE increased to 2.45 mm while R² remained 0.47. Forecast skill declined further at T + 24 h (MAE = 5.10 mm; R² = 0.23) and was limited at T + 72 h (MAE = 6.22 mm; R² = 0.04). The results demonstrate that satellite-driven machine learning can provide useful very-short-range PWV guidance, particularly within the first 12 h. Full article

Particulate matter (PM_2.5 and PM₁₀) forecasting is increasingly framed as an applied machine-learning problem operating on real-world environmental sensor infrastructure, yet most benchmarking studies evaluate models on a single, curated station rather than on the heterogeneous, imperfect data that operational networks actually produce, and on a single chronological test split whose representativeness is rarely questioned. This study benchmarks four supervised model families, Random Forest (RF), Support Vector Regression with an RBF kernel (SVM), a feed-forward Neural Network (NN), and a Long Short-Term Memory network (LSTM), against a naïve persistence baseline and a conventional autoregressive baseline (SARIMA), across five of six monitoring stations of the Greek National Air Pollution Monitoring Network (EDPAR), a sixth, short-record rural station retained for exploratory analysis only, selected to span contrasting emission regimes and record lengths of 4 to 24 years. Using a univariate, past-only feature set (autoregressive lags, trailing rolling statistics, and cyclical calendar encodings), we forecast both next-day concentration and next-week maximum concentration for PM_2.5 and PM₁₀ independently at each station, under a two-stage evaluation design: a 40-window sliding model-selection stage, which selects each model family’s hyperparameters by mean R² across many chronological cutoffs, and a single held-out 20% model-assessment stage on data never used for selection. SVM is the most broadly reliable model family under both stages, winning 11 of 20 station/pollutant/horizon combinations under model selection and 13 of 20 under model assessment, with its advantage most pronounced at the seven-day-maximum horizon; RF, NN and LSTM are each competitive at specific stations, but none is reliably dominant. The two evaluation stages agree on the winning model in only 10 of 20 combinations, illustrating that model rankings from a single chronological split can depend materially on the specific evaluation window chosen. SARIMA underperforms the best machine-learning model in all 20 of 20 combinations and underperforms naïve persistence itself at several stations, indicating that the machine-learning models’ advantage reflects genuine predictive skill rather than merely the seasonal and autoregressive information already available to any conventional time-series method. Achievable R² is generally, though not universally, lower for PM₁₀ than for PM_2.5, reflecting PM₁₀’s larger coarse-mode, episodically-driven component; station-specific exceptions to this pattern are attributable to long-term trend and test-set variance-compression effects rather than to any intrinsic reversal of pollutant forecastability. These results indicate that model selection for PM forecasting should be validated across multiple evaluation windows rather than a single split, that conventional statistical baselines remain a necessary comparison point, and that reported gains from more complex architectures should be interpreted cautiously in light of test-set non-stationarity and the sensitivity of model rankings to the choice of evaluation window. Full article