Sürdürülebilir Kalkınma Amaçları
Yayınlar
- ArticleYalnız künye2026Air quality index modeling using ensemble and data-intelligence modeling
Pande, Chaitanya Baliram; Katipoğlu, Okan Mert; Radwan, Neyara; Sidek, Lariyah Mohd; Pal, Subodh Chandra; Yaseen, Zaher Mundher; Ali, Tarig; Chakrabortty, Rabin; Refadah, Samyah Salem; Khan, Mohd Yawar Ali
Urban population growths, expansion of the industrial economy, burning stubble, and other major sources have significantly contributed to the increasing levels of air pollution in urban areas. Therefore, accurate prediction of the Air Quality Index (AQI) is essential for effective monitoring and planning within cities. In this research, an ensemble machine learning (ML) called Gradient Boosting (GB) model with 5- and 10-folds’ cross-validation was developed to achieve more precise and operative prediction outcomes based on important air quality related input variables. The results demonstrated that the suggested hybrid model attained the best accuracy in comparison with other benchmark ML models for predicting AQI values of Delhi city. Statistically, GB model reported second best model based on statistical metrics (R2 = 0.96, R = 0.98, RMSE = 23.03, MSE = 530.71, MAE = 15.16 and MARE = 11.96). Whereas Light Gradient Boosting Machine (LightGBM) model third best accuracy reported (R2 = 0.96, R = 0.98, RMSE = 23.21, MSE = 538.74, MAE = 14.68, and MARE = 11.33). The AQI prediction values are presented as spatial thematic maps for better understanding, highlighting city areas with high AQI levels to support air quality planning. The daily AQI prediction maps were prepared using various ML models. Further, the current research forecasted five days of AQI values and compared with the observed values. The results of five-day forecasting of GB model again indicated the best match with observed data of AQI values. Finally, forecasting accuracy results of GB (Boosting) models was attained 90% match with the observed values of AQI as compared with hybrid model. In conclusion, the reported results provided better understanding of AQI modeling for values and classes of spatial maps.
RDL Kalıcı Kaydıİncele - ArticleAçık Erişim2026Hybrid VMD-EMD Enhanced Machine Learning Models for Daily Reference Evapotranspiration Forecasting
Acar, Reşat; Katipoğlu, Okan Mert; Çırağ, Burak
Getting accurate daily estimates of reference evapotranspiration (ET₀) is essential for hydrological modeling, irrigation planning, and climate research. This study evaluated the performance of four machine learning models: Adaptive Boosting (AdaBoost), Categorical Boosting (CatBoost), Deep Neural Network (DNN), and Light Gradient Boosting Machine (LGBM) for daily ET₀ prediction. Feature selection was implemented using Partial Autocorrelation (PACF) and Autocorrelation (ACF) analysis on the basis of an 8-day lag structure, while the CatBoost feature importance framework was implemented to identify the most important inputs. Moreover, the methodologies of decomposition-based preprocessing, Empirical Mode Decomposition (EMD) and Variational Mode Decomposition (VMD) were implemented on the lagged inputs to build the hybrid models augmented with intrinsic constituents. These augmented hybrid models were created to improve the data representation as well as predictive capability. Model performance was evaluated using key statistical metrics, including RMSE, MAE, NSE, KGE, R2, and bias-related indicators. CatBoost produced the lowest RMSE (1.57 mm/day) and MAE (1.13 mm/day), and the highest NSE (0.67), KGE (0.78), and R2 (0.67), along with the lowest AIC score (627.15), indicating strong and reliable predictive performance. Furthermore, the VMD based CatBoost model yielded even greater improvements, highlighting the added value of decomposition enhanced learning. On the contrary, DNN showed the weakest results, with the highest RMSE (2.39 mm/day) and lowest NSE (0.23). These findings suggest that CatBoost and its hybrid variants are highly effective and reliable tools for daily ET₀ prediction, particularly in data limited environments and mountainous regions.
RDL Kalıcı Kaydıİncele - ArticleYalnız künye2026Investigating the rubberized concrete-filled steel tube composite columns and developing artificial intelligence-based analytical models for ultimate axial strength prediction
Simsek, Oguz; Katipoğlu, Okan Mert; İpek, Süleyman; Güneyisi, Erhan; Güneyisi, Esra Mete
The application of rubberized concrete (RuC) in structural systems has led to the development of a novel composite system: the rubberized concrete-filled steel tube (RuCFST). The characteristics of infill materials, particularly RuC, significantly influence the structural behavior and performance of these composite systems. This study aims to evaluate the ultimate axial strength of RuCFST columns, considering rubber aggregate content up to 75% and infill concrete compressive strength up to 63 MPa, and propose artificial intelligence (AI)-based design models using deep learning (CNN, DNN, Autoencoder), machine learning (CatBoost), hybrid AI/neuro-fuzzy (ANFIS), and unsupervised neural network (SOM) to predict the ultimate axial strength. A robust dataset was compiled from the literature, comprising the results of 131 RuCFST column specimens. The experimental results underwent a comprehensive statistical assessment using various analytical methods, including Pearson correlation coefficient analysis, histogram analysis, distribution fitting curve analysis, and box-and-whisker plot analysis. Additional statistical metrics such as Kolmogorov–Smirnov, skewness, and kurtosis values were also used. Furthermore, by considering various variables covering the geometrical and material properties of the RuCFST columns, the design models were developed using AI techniques. The statistical evaluation results indicated that the compiled dataset is statistically significant and encompasses various input and output parameters. The design models developed in this study displayed varying prediction performance. However, those created using the Autoencoder and ANFIS methods showed statistically satisfactory performance in determining the load-carrying capacity of the columns.
RDL Kalıcı Kaydıİncele - ArticleAçık Erişim2026Physically constrained air–sea heat flux driven modeling of sea surface temperature using atmospheric and oceanographic variables with explainable AI
Katipoğlu, Okan Mert; Bazrafshan, Ommolbanin; Mohammadi, Mojtaba; Alsumaiei, Abdullah A.
This study aims to develop a robust multivariate regression framework for accurately predicting monthly Sea Surface Temperature (SST) using a core set of ten primary atmospheric and oceanographic variables, including relative humidity, specific humidity, temperature, mean square slope of waves, mean wave direction, 2-meter air temperature, top net solar radiation, total precipitation, mean evaporation rate, and potential evaporation. The originality of this research lies in the simultaneous integration of meteorological and oceanographic parameters into multiple machine learning architectures, coupled with a detailed interpretability analysis using SHAP values, which provide a comprehensive understanding of the influence of variables on SST prediction. The dataset was split into 60 % for training, 20 % for testing, and 20 % for validation. Four machine learning models: Random Forest, CatBoost, one-dimensional Convolutional Neural Network, and Support Vector Machine were optimized using RandomizedSearchCV, with MinMaxScaler applied to ensure consistent feature scaling across models. Model performance was evaluated using Mean Squared Error, Root Mean Squared Error, Mean Absolute Error, Coefficient of Determination, and Explained Variance metrics. Meanwhile, SHAP analysis identified the most influential features as 2-meter air temperature, mean evaporation rate, mean square slope of waves, air temperature, and specific humidity. Results revealed that the Support Vector Machine achieved the highest accuracy overall, followed by CatBoost, confirming that combining meteorological and oceanographic predictors significantly enhances SST forecasting performance. The physical consistency analyses, supported by the bulk air–sea heat flux formulations for sensible and latent heat fluxes, confirm that the machine-learning model’s most influential predictors: near-surface air temperature, specific humidity, and evaporation-related variables are fully consistent with the fundamental thermodynamic controls governing SST. This research not only advances SST prediction accuracy but also provides a transparent interpretability framework, making it a valuable reference for climate modeling, marine resource management, and environmental monitoring applications.
RDL Kalıcı Kaydıİncele - ArticleYalnız künye2026Beyond Conventional Methods: Predicting Novel Modified Groundwater Drought Indices with Advanced Computing Paradigms
Katipoğlu, Okan Mert; Kartal, Veysi; Aktürk, Gaye; Kilinc, Huseyin Cagan; Elshaboury, Nehal; Yamany, Mohamed S.; Pande, Chaitanya Baliram
To achieve accurate and reliable predictions of groundwater drought, it is essential to incorporate relevant meteorological variables and apply advanced predictive methodologies capable of modeling their complex interrelationships. In this study, groundwater drought prediction models were developed based on the seven modified drought indices, which are obtained from groundwater levels and multiple meteorological parameters. The analysis focuses on four districts in Türkiye (Tekirdağ-Çorlu, Tekirdağ-Çerkezköy, Amasya-Merzifon, and Amasya-Centre), which are hereafter referred to collectively as the study areas. Three machine learning models were applied such as a standalone Artificial Neural Network (ANN) and two hybrid models, ANN integrated with Shuffled Frog Leaping Algorithm (SFLA-ANN) and Firefly Algorithm (FFA-ANN). The hybrid metaheuristic algorithms were chosen for their ability to model nonlinear and complex relationships between meteorological variables and groundwater levels, which traditional statistical methods cannot fully capture. Modified drought indices were employed to represent the groundwater drought better. The findings indicate that the hybrid FFA-ANN model generally provides the highest predictive accuracy for most indices, with coefficient of determination values frequently exceeding 0.90. In contrast, the ANN model occasionally outperforms the hybrid FFA-ANN model in terms of root mean square error (RMSE) for specific stations within the study areas. These results highlight the robustness of hybrid ANN approaches, especially FFA-ANN, in predicting groundwater drought under changing climate conditions. The novelty of this study is the modification of meteorological drought indices widely used to represent groundwater drought and the application of hybrid metaheuristic ANN models to estimate these indices in Türkiye. The results demonstrated that such models provide a reliable framework for predicting groundwater drought and provide decision-makers with advanced tools for water resources management, irrigation planning, and climate adaptation strategies.
RDL Kalıcı Kaydıİncele - ArticleAçık Erişim2026Optimization of various machine learning concepts to evaluate landslide susceptibility: XGBoost, k-NN and MLP using PSO algorithm
Abdo, Hazem Ghassan; Richi, Sahar Mohammed; Zerouali, Bilel; Alqadhi, Saeed; Katipoğlu, Okan Mert; Prasad, Pankaj; Arman, Hasan; Albanai, Jasem A; Mallick, Javed
Landslides significantly threaten natural and built environments, necessitating accurate prediction models for effective hazard mitigation. There is an urgent need to further improve the performance of machine learning algorithms in predicting landslide susceptibility by monitoring the impact of optimization algorithms on the performance of these models. This study evaluates the performance of various machine learning classifiers, including k-Nearest Neighbors (kNN), Multi-Layer Perceptron (MLP), and Extreme Gradient Boosting (XGBoost), for landslide susceptibility mapping. Additionally, Particle Swarm Optimization (PSO) is employed to enhance model performance by optimizing hyperparameters. Mountainous areas in the eastern Mediterranean (the northern Kabir River basin in western Syria) were identified as a result of the high frequency of landslide events over the past two decades. Nineteen factors causing landslides were identified, with no factor excluded, as a result of a multicollinearity test. The results indicate that XGBoost achieves the highest performance among traditional models. When integrated with PSO, the PSO-XGBoost model further improves classification performance, demonstrating its robustness in handling complex spatial patterns. Feature importance analysis using SHAP confirms slope as the dominant factor, followed by TRI, rainfall, Aspect, TWI, and curvature, highlighting the role of topography and hydrology in landslide occurrence. Moderate lithology, NDVI, and LULC contributions and lower importance of Flow Accumulation and Soil Depth suggest complex environmental interactions. Model predictions show varying susceptibility distributions. PSO-MLP assigns the highest very high susceptibility (44.09%), while PSO-XGBoost provides a balanced classification (31.13%). The PSO-XGBoost model demonstrates superior predictive capability, offering reliable landslide susceptibility maps for disaster risk management and land-use planning.
RDL Kalıcı Kaydıİncele - ArticleYalnız künye2026Prediction of soil moisture via feature selection, model optimization, and climate data integration
Katipoğlu, Okan Mert; Kartal, Veysi; Akiner, Muhammed Ernur; Terzioğlu, Zeynep Özge; Kılıç, Zeyneb
Soil moisture is vital for water resource management and agricultural productivity. With the growing availability of reanalysis data and advanced machine learning methods, it is essential to explore reliable, scalable models for long-term estimation. Comparing different algorithms enables a comprehensive assessment of their strengths and limitations in modeling complex soil moisture dynamics under varying climatic and agricultural conditions. In this study, soil moisture data for the period 1950–2022 from six stations in the Konya Closed Basin of Türkiye were estimated using Echo State Networks (ESN), Adaptive Boosting (AdaBoost), Deep Neural Networks (DNN), and Light Gradient Boosting Machine (LGBM). Precipitation (PR), solar radiation (RSDS), wind speed (WS), mean temperature (Tave), maximum temperature (Tmax), minimum temperature (Tmin), and potential evapotranspiration (ET0) values calculated by the Thornthwaite equation were used as input variables of the models, and soil moisture (SM) was determined as the output variable. Model performance was evaluated using statistical metrics and visualizations derived from training, validation, and test datasets. In addition, a feature selection approach was used to identify the most effective parameters for soil moisture estimation. The analysis showed that the DNN model achieved the highest accuracy, with R2 values ranging from 0.87 to 0.97. The second most successful model was AdaBoost or ESN, while the LGBM model showed higher deviations than the others. These findings reveal that artificial intelligence-based models can effectively estimate soil moisture.
RDL Kalıcı Kaydıİncele - ArticleYalnız künye2026Assessment of future hydrometeorological variability in the Konya Basin, Türkiye using bias-corrected CMIP6 projections
Katipoğlu, Okan Mert; Kartal, Veysi; Akiner, Muhammed Ernur; Bazrafshan, Ommolbanin; Spor, Pinar; Çelik, Mehmet Ali
This study presents bias-corrected projections of soil moisture, precipitation, Temperature and evapotranspiration in the Konya Basin, derived using CMIP6 data under SSP1-2.6, SSP2-4.5, and SSP5-8.5 scenarios. Bias correction methods, including Delta, Quantile Mapping, and Empirical Quantile Mapping, were applied to the raw model outputs. Historical data from CMIP6 (1950–2014) and future projections (2015–2100) were corrected using ERA5 reanalysis data (1940–2024) to improve accuracy. Nine global climate models ACCESS-CM2, CESM2, CMCC-ESM2, EC-Earth3-CC, MIROC6, GFDL-ESM4, MPI-ESM1-2-LR, MRI-ESM2-0, and NorESM2-LM were evaluated, and ACCESS-CM2 provided the most reliable results for all three SSP scenarios. The data were divided into four time periods: the reference period (2015–2024), near future (2025–2049), mid-future (2050–2074), and far future (2075–2100). Relative errors were calculated as (Observed - Predicted)/Observed × 100, with temperature trends incorporated for comparative assessments. For spatial analysis, Universal Co-Kriging was employed to generate predictive maps for each future period, using elevation as a secondary variable. Therefore, the Universal Co-Kriging parameters determined through a trial-and-error approach namely, the use of a Gaussian kernel, standard neighborhood structure with 5 maximum and 2 minimum neighbors, and a 45-degree sectoral division in four directions proved to be the most effective combination for accurately mapping the spatial distribution of soil moisture, precipitation, temperature, and evapotranspiration in the Konya Basin. The study provides valuable insights into the future variability of hydrometeorological variables in the Konya Basin, offering critical information for water resource management and climate adaptation strategies.
RDL Kalıcı Kaydıİncele - ArticleAçık Erişim2026Machine Learning-Based Hydrological Drought Prediction Integrating Teleconnections and Hydrological Memory in a Semi-Arid Basin, Algeria
Katipoğlu, Okan Mert; Achite, Mohammed; Kartal, Veysi; Çelik, Mehmet Ali; Pandey, Kusum
Hydrological drought forecasting in semi-arid basins is challenging due to the combined influence of meteorological forcing, large-scale atmospheric teleconnections, and basin memory processes, which are rarely jointly analysed within a leakage-free predictive framework. This study addresses this gap by evaluating gradient-boosted trees and neural forecasting models for one-month-ahead prediction of the Standardized Runoff Index (SRI) in two sub-basins of the Wadi Sahaouat Basin, Algeria. The models include gradient-boosted regression trees (GBRT), A-N-BEATS, A-N-HiTS, and TiDE, representing distinct forecasting architectures. Predictors consist of the Standardised Precipitation Index (SPI), seven teleconnection indices (NAO, AO, EAWR, SCAND, MEI, SOI, WeMO), and their one- to three-month lags. Two scenarios are tested: Scenario 1 uses SPI and teleconnection lags only, while Scenario 2 additionally includes lagged SRI values (SRI_lag1–3) to represent hydrological memory. A train-only Variance Inflation Factor (VIF > 10) procedure is applied to remove multicollinearity without data leakage. In Basin 1, SRI lags were excluded due to strong collinearity with SPI lags (r = 0.984), resulting in identical inputs for both scenarios. In Basin 2, SRI lags were retained to assess their predictive contribution. GBRT achieved the best overall performance across both basins and scenarios, with mean RMSE, NSE, and KGE values of 0.0682, 0.9907, and 0.8945, respectively. TiDE ranked second overall, with a mean RMSE of 0.1166, followed by A-N-HiTS in third place with a mean RMSE of 0.1203 and A-N-BEATS with the weakest overall performance, with a mean RMSE of 0.2159. These results indicate that gradient-boosted trees remain highly competitive with neural models for small monthly hydrological datasets and that the value of hydrological memory is basin-dependent and varies according to its independence from concurrent meteorological forcing.
RDL Kalıcı Kaydıİncele - ArticleAçık Erişim2026Decoding the dynamics of daily total global solar radiation: Computational learning strategies powered by explainable AI
Akiner, Muhammed Ernur; Tınğ, Naim Süleyman; Alsumaiei, Abdullah A.; Katipoğlu, Okan Mert
This study investigates the prediction of daily total global solar radiation using advanced machine learning and hybrid models. The input variables include various meteorological and environmental parameters such as total precipitation, average snow height, total solar intensity, existing snow height, soil temperature, actual pressure, average temperature, cloudiness, minimum cloud height, and relative humidity. Based on SHAP value analysis, 5 cm Soil Temperature (+67.08) and Total Solar Intensity (+34.76) emerged as the most influential features for the models, while parameters like Current Snow Depth (+0.33) and Average Snow Depth (+0.36) contributed minimally. The analysis spans a period from 2006 to 2022 using daily data. The dataset is divided into 70 % for training and 20 % for validation, with 10 % of the training data allocated for testing. Several models were employed, including AdaBoost, CatBoost, Deep Belief Network (DBN), Light Gradient Boosting Machine (LGBM), Adaptive Neuro-Fuzzy Inference System (ANFIS), Echo State Neural Networks (ESN), and Autoencoder-based regression. According to the analysis results, the Catboost model produces the most accurate predictions with RMSE: 97.25 KW.hr/m2, MAE: 73.64 KW.hr/m2, AIC: 17075, NSE: 0.65 and, KGE: 0.72 values, while the DBN model shows the second best model with RMSE: 98 KW.hr/m2, MAE: 74.48 KW.hr/m2, AIC: 17103, NSE: 0.64 and, KGE: 0.74. The analysis results have the potential to provide information to decision makers and planners, especially in the evaluation of the performance of solar panels, the optimization of agricultural processes, and the development of climate change adaptation strategies.
RDL Kalıcı Kaydıİncele - ArticleYalnız künye2026Change-point detection and trend magnitude estimation in long-term streamflow records of a transboundary river basin under climate change
Tanas Savaş, Gülçin; Ertugay, Neşe; Katipoğlu, Okan Mert
Global climate change is reshaping basin hydrology by altering precipitation timing, evaporative demand, runoff generation, and groundwater recharge, which makes long-term streamflow trend detection essential for sustainable water resources management and climate adaptation. In this study, monthly, seasonal, and annual streamflow trends from three gauging stations in the upper Euphrates basin (D21A001, E21A019, and E21A051) were evaluated using an integrated non-parametric framework that combines Mann–Kendall (MK), Spearman’s rank correlation (SRC), Sequential Mann–Kendall (SMK), innovative trend analysis (ITA), and ITA slope (ITAS). This multi-method design allowed consistent identification of trend direction, magnitude, and temporal transition behavior, while also improving interpretive robustness across stations with different streamflow characteristics. The results show a clear common signal: streamflow weakening is most pronounced in the spring to early summer period, although the strength and timing vary by station. E21A019 shows a relatively coherent declining trend across methods, particularly during April–June and at seasonal and annual scales, indicating a stronger tendency toward streamflow reduction compared to the other stations. E21A051 shows a seasonally contrasting regime, with increasing tendencies in winter and late summer/early autumn (notably January–February and August–September), but decreasing trends during spring and early summer, implying intra-annual redistribution rather than uniform change. D21A001 presents weaker trend magnitudes overall, yet decreasing tendencies are still evident in late spring, early summer, and autumn. Across methods, MK and SRC consistently support the trend directions, while ITA/ITAS better highlight magnitude and seasonal contrast. Overall, the findings provide robust evidence of long-term streamflow change and hydrological pressure in the upper Euphrates basin. They can support adaptive and transboundary water management planning, while attribution to specific climatic or anthropogenic drivers requires direct analysis of the relevant forcing variables.
RDL Kalıcı Kaydıİncele - ArticleYalnız künye2026A New Paradigm in Flood Management: Hydrological, Hydraulic, and Deep Learning Methods for Flood Routing
Sarıgöl, Metin; Katipoğlu, Okan Mert; Aktürk, Gaye
Flood routing is essential for water resources management and flood mitigation, yet traditional methods often struggle to represent complex hydraulic processes and rapidly changing flow conditions. In this study, hydrological models (Muskingum and SCS), hydraulic methods (Kinematic Wave, Muskingum–Cunge, and Dynamic Wave), and deep learning (DL) algorithms (Autoencoder, Deep Neural Network (DNN), Gated Recurrent Unit (GRU), Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), and Recurrent Neural Network (RNN)) were applied to predict flood routing. Model performance was evaluated using several statistical indicators, including RMSE, AIC, MAE, KGE, NSE, R², Pbias, and MBE. The results showed that the Autoencoder model provided the best predictive performance (RMSE: 0.30, MAE: 0.17, NSE: 0.97, R²: 0.99), while the RNN model produced the weakest results. Other DL models, particularly DNN, CNN, and LSTM, also demonstrated strong predictive capability. Among the hydraulic approaches, the Kinematic Wave method yielded the most accurate results, whereas the SCS model showed the best performance among hydrological routing methods. Overall, the findings indicate that DL models generally outperform classical routing approaches in flood prediction, with the Autoencoder architecture emerging as the most effective model under the conditions of the present study.
RDL Kalıcı Kaydıİncele - ArticleYalnız künye2026Advancing Forest Fire Susceptibility Mapping in the Eastern Mediterranean Using Multi-Algorithm Machine Learning
Abdo, Hazem Ghassan; Richi, Sahar Mohammed; Alqadhi, Saeed; Prasad, Pankaj; Katipoğlu, Okan Mert; Nguyen-Huy, Thong; Das, Jayanta; Almohamad, Hussein; Mallick, Javed
Forest fires are among the most severe natural hazards threatening ecological sustainability in the Eastern Mediterranean, with Syria being a prominent example where ongoing conflict has undermined forest management and monitoring systems. Despite numerous global studies on forest fire susceptibility, conflict-affected areas in the Eastern Mediterranean remain underrepresented in the scientific literature due to limited data and complex environmental and social conditions. In particular, few studies have systematically compared multiple machine learning algorithms for predicting forest fire susceptibility in these regions. To address this gap, this study evaluates the performance of six machine learning algorithms—Extreme Gradient Boost (XGBoost), Random Forest (RF), Support Vector Machine (SVM), Multilayer Perceptron (MLP), K-Nearest Neighbor (KNN), and Linear Regression (LR)—using 3,589 fire events and 13 conditioning factors in the Latakia governorate, western Syria. Results indicate excellent predictive performance for all models, with XGBoost achieving the highest AUC of 0.993, followed by RF (0.992), KNN (0.977), MLP (0.975), SVM (0.968), and LR (0.963). These findings provide a robust basis for supporting fire management and mitigation strategies in Syria and offer insights applicable to other conflict-affected regions of the Eastern Mediterranean.
RDL Kalıcı Kaydıİncele - ArticleAçık Erişim2026Robust Monthly Lake Water Level Forecasting Using an Ensemble of Singular Spectrum Analysis Based Deep Learning and Categorical Boosting Models
Katipoğlu, Okan Mert; Kartal, Veysi; Tuğrul, Türker; Yavuz, Veysel Süleyman; Ariman, Sema
This study developed an innovative hybrid modelling approach to predict one-month-ahead water levels (LWL (T + 1)) for two major lakes in Türkiye (Eber and Eğirdir), aiming to model complex dynamics of hydrological systems and optimize sustainable water resources management. In the proposed methodological framework, Singular Spectrum Analysis (SSA) was employed to decompose hydrometeorological time series into sub-components, while Mutual Information (MI) and Partial Autocorrelation Function (PACF) techniques were utilized to determine optimal input variables. The hybrid model architecture, incorporating Deep Neural Networks (DNN), Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), and CatBoost algorithms, was optimized using 70% of the data for training and 30% for testing, and its performance was evaluated through deterministic and probabilistic metrics. The analyses revealed that SSA-based data preprocessing significantly improved the performance of prediction models, and depending on lake characteristics, CNN-based models demonstrated superior performance for Lake Eber, while DNN architecture excelled for Lake Eğirdir. The developed hybrid approach presents an original methodology that can be integrated into decision support systems for water resources management, contributing to the development of sustainable water management strategies and adaptability to similar hydrological systems.
RDL Kalıcı Kaydıİncele - ArticleYalnız künye2026Data-driven forecasting of photovoltaic performance with explainable artificial intelligence: Experimental validation using environmental and thermal predictors
Karakoyun, Yakup; Katipoğlu, Okan Mert; Dogan, Ahmet; Canbaz, Ali
PV systems meet energy demand, but environmental and operational factors affect their performance. This study used a three-scenario ablation framework to predict PV power from baseline meteorological and operational variables to under-panel temperature and finally to the full thermal configuration, including water temperature (a newly studied variable). Scenario 1 included time, air temperature, panel angle, relative humidity, and average wind speed; Scenario 2 added under-panel temperature; and Scenario 3 added water temperature. Direct thermal predictors were evaluated separately from meteorological inputs using this design. Four sophisticated ML algorithms, Random Forest (RF), CatBoost, Support Vector Machine (SVM), and one-dimensional Convolutional Neural Network (1D CNN), were refined by Randomized Search (RS) and evaluated using metrics. According to the first group of power prediction results under the full Scenario 3 configuration, the CatBoost-RS model demonstrated superior performance with RMSE = 10.61, MAE = 8.41, AIC = 222.22, KGE = 0.957, and R2 = 0.99. For the second group, the SVM-RS model outperformed others, achieving RMSE = 8.02, MAE = 5.50, AIC = 200.36, KGE = 0.994, and R2 = 0.992. Comparative analysis showed that CatBoost-RS was the most robust predictor, while RF exhibited proficiency in managing nonlinearity in this experimental setting. One-way ANOVA, paired t -test, Wilcoxon signed-rank test, bootstrap confidence intervals, and effect size (Cohen’s d) analyses across both groups show that optimised ensemble and kernel-based models outperform the convolutional approach in predictive performance. Experimental validation and high predictive accuracy improve PV performance prediction in the proposed method.
RDL Kalıcı Kaydıİncele - ArticleAçık Erişim2026Estimation of the pan evaporation coefficient in semi-arid climate conditions via machines learning models
Achite, Mohammed; Katipoğlu, Okan Mert; Vishwakarma, Dinesh Kumar; Pandey, Kusum; Salem, Ali; Elbeltagi, Ahmed
In this study, class A pan coefficient (KPan) values were simulated via five machine learning models, namely the ANN, the AAN-REPTree, the ANN-SMO SVM, the ANN-Linear regression, and the ANN-Bagging model, by using daily meteorological data of the meteorological station of Ouled Ben Abdelkader region, which has semi-arid microclimate in the northwest region of Algeria. To determine the optimal combination of inputs, a variety of input-target pairs were tested by a variety of machine learning models, resulting in seven possible input scenarios: Tmean, RHmin, RHmax and Wind Speed were found to be the best input combinations. The results of models were analyzed (i.e., correlation coefficient (R), mean absolute error (MAE), root mean squared error (RMSE), relative absolute error (RAE), root relative squared error (RRSE)) to find their accuracy. As the best optimal condition, four input variables were introduced for the models ANN, ANN-REPTree, ANN-SMO SVM, ANN-Linear regression, and ANN-Bagging, with R = 0.9941 and 0.984, MAE = 0.0018 and 0.0037, RMSE = 0.0068 and 0.0016, RAE = 3.8863 and 7.6858, and RRSE = 10.9301 and 18.0686 in the training and testing phases, respectively. This hybrid model (ANN-Bagging) has demonstrated its utility in a scenario where there is a strong connection among the variables which includes KPan, and its feasibility to display the model in a feasible state.
RDL Kalıcı Kaydıİncele - CorrectionAçık Erişim2026Correction: Estimation of the pan evaporation coefficient in semi-arid climate conditions via machines learning models (Applied Water Science, (2026), 16, 2, (47), 10.1007/s13201-025-02709-6)
Achite, Mohammed; Katipoğlu, Okan Mert; Vishwakarma, Dinesh Kumar; Pandey, Kusum; Salem, Ali; Elbeltagi, Ahmed
In this article, the figures 3 and 5 appeared incorrectly and have now been corrected in the original publication. For completeness and transparency, the old incorrect and correct versions are displayed below: Simple Artificial Neural Networks (ANN) structure Typical architecture of Bagging algorithm The original article has been corrected.
RDL Kalıcı Kaydıİncele - ArticleYalnız künye2025Intercomparison of sediment transport curve and novel deep learning techniques in simulating sediment transport in the Wadi Mina Basin, Algeria
Achite, Mohammed; Katipoğlu, Okan Mert; Elshaboury, Nehal; Tuğrul, Türker; Pandey, Kusum
The accurate estimation of sediment discharge is crucial for the design and operation of engineering structures such as dams, water treatment facilities, and erosion control systems. This study evaluates the performance of various machine learning (ML) and deep learning (DL) models in predicting sediment transport in the Mina Basin, Algeria, at two stations: Oued Abtal and Sidi Abdelkader Djillali. The models include the sediment rating curve, category boosting, convolutional neural network, deep neural network (DNN), gated recurrent unit, and multilayer perceptron. Among these, the DNN model consistently demonstrated superior performance. For Oued Abtal station, the DNN achieved RMSE = 243.72 kg/s, MAE = 102.17 kg/s, NSE = 0.99, and PBIAS = 6.81%. At Sidi Abdelkader Djillali station, it recorded RMSE = 91.27 kg/s, MAE = 46.51 kg/s, NSE = 0.99, and PBIAS = 38.06%. Error analysis revealed that the DNN model offers the most reliable predictions, outperforming both traditional and other ML/DL methods. This study underscores the potential of deep learning models in advancing sediment transport prediction, particularly in semi-arid regions, and highlights their implications for sustainable water resource management.
RDL Kalıcı Kaydıİncele - ArticleYalnız künye2025Optimizing river flow rate predictions: integrating cognitive approaches and meteorological insights
Kartal, Veysi; Karakoyun, Erkan; Akiner, Muhammed Ernur; Katipoğlu, Okan Mert; Kuriqi, Alban
The models used in this study make it possible to make more accurate predictions about river discharge. These results can influence flood protection strategies, water resource management, and hydropower generation. Due to their ability to capture the underlying temporal relationships in the data, time series forecasts have become increasingly popular in recent years. This study examines the critical processes in river forecasting for the Kizilirmak River basin. We begin with a look at data collection and preparation, followed by an overview of time series forecasting models. Finally, we look at the process of model testing and selection. Seven techniques were used to predict streamflow from meteorological data: Artificial Neural Network (ANN), Firefly-based ANN (FFA-ANN), Random Forest (RF), K-Nearest Neighbors (KNN), Generalized Linear Regression (GLR), Support Vector Machines (SVM), Least Squares Boosted Trees (LSBT). The performance of the models was evaluated using the statistical indicators. The LSBT, RF, and ANN models provided the best results for Kayseri, Kırşehir, and Gemerek stations, respectively. The RF, ANN and GLR models provided second best results for these stations, respectively.
RDL Kalıcı Kaydıİncele - ArticleAçık Erişim2025A coupled extreme gradient boosting-MPA approach for estimating daily reference evapotranspiration
Achite, Mohammed; Nasiri, Hamid; Katipoğlu, Okan Mert; Abdallah, Mohammed; Moazenzadeh, Roozbeh; Mohammadi, Babak
Reference evapotranspiration (ET0) modeling is pivotal for irrigation scheduling and water resources planning. This study presents a hybrid approach integrating Extreme Gradient Boosting (XGB) with Marine Predators Algorithm (MPA) for daily ET0 estimation in northern Algeria. The proposed XGB-MPA model was evaluated against traditional empirical models and assessed using statistical methods. Shapley Additive Explanations (SHAP) was employed to enhance model interpretability. Various combinations of meteorological variables were tested as inputs, including air temperature, relative humidity, sunshine hours, wind speed, and extraterrestrial solar radiation. The XGB-MPA hybrid model achieved superior prediction accuracy during testing (R2 = 0.9958, RMSE = 0.1713 mm/day) compared to traditional empirical models and the standard XGB model. The study demonstrated that ET0 prediction accuracy increased with the number of meteorological inputs used. Our findings highlight the XGB-MPA hybrid model's potential for accurate ET0 estimation in northern Algeria, which can be used for water resource management and irrigation planning.
RDL Kalıcı Kaydıİncele