Flood events have become more frequent as a result of seasonal changes, global warming, and changes in sea level. In terms of basin management, it is necessary to know the hydrodynamics of the basin in order to produce faster solutions in emergency action plans. The Filyos River is one of the two most important floodplains in the western Black Sea basin and has so far only been analyzed to a limited extent using modern hydrological and hydraulic models. In order to analyze the flood dynamics and determine the flood risks in the Filyos River. In this context, flood hydrographs, rainfall depths, peak flows, and excess water volumes were calculated for different return periods (2, 5, 10, 20, 50, and 100 years) using HEC-RAS, HEC-HMS, and Hyfran Plus software. The analyses showed that the rainfall depth and peak flow rate increased significantly as the return period increased. It was also observed that although the volume of precipitation increased, the amount of water converted into surface runoff remained limited due to infiltration and other losses. The results of the study contribute to the identification of high flood-risk areas in the Filyos River basin, the improvement of flood prevention infrastructure, and the development of sustainable water management policies. Analyses using modeling tools such as HEC-RAS and HEC-HMS provide a scientific basis to help local governments and decision makers strengthen flood prevention strategies, update risk maps, and make emergency response plans more effective while making flood scenarios more reliable.

This paper presents a multi-stage machine learning framework supported by explainable artificial intelligence (XAI) techniques to predict temperature trends using meteorological data from Zonguldak, Turkey. This hybrid approach aims to provide high prediction accuracy and model transparency by integrating XAI methods such as Principal Component Analysis (PCA) based dimensionality reduction, stable regression modeling, and permutation significance with SHapley Additive exPlanations (SHAP). As a result of the comparison, Linear Regression and Ridge Regression showed the most consistent performance, achieving the lowest error rates and highest coefficient of determination values in both cross-validation and test datasets. Also, methods such as Gradient Boosting and Support Vector Regression provided competitive results. The application of PCA has improved model efficiency by minimizing information loss while reducing the dimensional complexity of the dataset. SHAP and permutation analyses showed that one of the principal components has a significant impact on the model’s prediction results. Through these analyses, it’s clear which features the model makes decisions based on. The analyses for explainability emphasize the importance of interpretability in modern AI applications. The findings show that the proposed framework successfully balances high accuracy and explainability and is a powerful and scalable tool for environmental decision-making processes by analyzing climate trends. Future work could extend this approach by applying the framework to different climate regions, adding new meteorological variables and integrating more advanced modelling methods.

Seasonal changes, sea level rise, and global warming make flood events more frequent, which necessitates watershed management and efficient use of water resources. In this context, understanding the hydrodynamic behavior of basins is critical for the development of flood prevention strategies. The contributions of hydrological and hydraulic modeling techniques in this process are among the key determinants of sustainable water resources management. The Filyos Sub-Basin, located in the Western Black Sea Basin, stands out as one of the regions where flood risk assessment is a priority, as it has two important floodplains. This study aims to analyze the flood risk in the Filyos River Sub-Basin with hydraulic modeling methods, and to determine the Manning roughness coefficient. In the study, the parameters affecting the roughness of the river bed were analyzed using the Cowan method, and the effects of vegetation on river bed resistance were evaluated in the laboratory environment. Flood simulations were carried out for four different flow rates (Q₁₀₀₀, Q₅₀₀, Q₁₀₀ and Q₅₀) using the HEC-RAS model, and the performance of flood protection structures were analyzed. The findings show that a significant portion of the existing protection structures are unable to meet the potential flood flows, which can cause serious damage to residential and agricultural areas. In basins with limited historical discharge data, such as the Filyos River, these findings provide important contributions to sustainable water resources management and regional planning processes. The results of the study serve as a reference for flood risk assessment, not only for the Filyos River Basin, but also for other basins with similar hydrodynamic characteristics. It is envisaged that future research, supported by larger data sets, can improve the accuracy of flood simulations. Furthermore, the Cowan method and HEC-RAS model used in this study are expected to contribute to strategic planning and engineering solutions to minimize flood risk in other watershed management projects. In future studies, we plan to further develop methodological approaches for determining the roughness coefficient, and to address applications to increase the effectiveness of flood protection structures.

This study examines the historical and anticipated effects of climate change on essential hydroclimatic variables (temperature, precipitation, evapotranspiration, and soil moisture) in the Southeastern Anatolia Project (GAP) region of Türkiye, a semi-arid and agriculturally significant basin experiencing heightened water stress. The analysis employs a collection of CMIP6 Global Climate Models (GCM) and integrates three Shared Socioeconomic Pathways (SSP1-2.6, SSP2-4.5, and SSP5-8.5), utilizing statistical bias correction methods such as Delta Change, Quantile Mapping (QM), and Empirical Quantile Mapping (EQM) to improve the regional accuracy of the projections. The ACCESS-CM2 model, validated with data from Türkiye’s Meteorological General Directorate (MGM), was chosen for comprehensive spatial mapping, utilizing Inverse Distance Weighting (IDW) interpolation across seven temporal intervals encompassing past, present, and future periods. The findings indicate a steady increase in temperature and evapotranspiration, especially under high-emission scenarios, with temperature rises above +4 °C and considerable water losses anticipated by century’s end. Soil moisture exhibits a declining tendency, particularly in the southern and eastern regions, signifying increasing drought susceptibility. Precipitation patterns demonstrate significant spatial variability and rising uncertainty, with relative error (RE%) values increasing under SSP5-8.5. Historical data from 1963 to 2022 corroborate these conclusions, indicating a progressive shift towards a warmer and drier regional climate. These observations highlight the importance of climate adaptation strategies and water management in the GAP region. The research provides decision-makers a high-resolution, bias-corrected hydroclimatic dataset.

Canlı varlıklar yaşamlarını sürdürebilmek için suya ihtiyaç duyarlar. Su yeryüzünde sürekli olarak bir dolanım halinde bulunmaktadır. Bu dolanım sırasında çeşitli sebeplerden dolayı suya karışan maddeler suyun özelliklerinin değişmesine sebep olur. Gerek suyun kullanımı gerekse etrafındaki alanların kullanımı suyun kalitesinin değişmesine neden olmaktadır. Bu nedenle akarsular ve su rezervlerinde su kalitesinin araştırılması önem arz etmektedir. Bu çalışmada, Batı Karadeniz Havzasının en büyük alt havzası olan ve Filyos Çayı Alt Havzasında bulunan Filyos Çayı’nda askıda katı madde konsantrasyonu yapay sinir ağları yöntemi ile tahmin edilmeye çalışılmıştır. Ana kol üzerinde seçilen üç noktadan su numuneleri alınarak laboratuvarda su kalite parametrelerinden askıda katı madde, bulanıklık, demir ve krom analizleri gerçekleştirilmiştir. Daha sonra Askıda Katı Madde parametresinin tahmini, bulanıklık, debi, demir ve krom parametrelerine dayanarak, yapay sinir ağları yöntemiyle yapılmıştır.

Canlı yaşamının devamlılığı ve ekosistem dengesinin sürdürülebilirliği için hayati bir unsur olan su, çevresel ve insan kaynaklı faktörlerin etkisiyle fiziksel ve kimyasal özelliklerinde değişimlere uğrayabilmektedir. Su kalitesinin temel göstergelerinden biri olan askıda katı madde (AKM) konsantrasyonunun doğru bir şekilde tahmin edilmesi, sürdürülebilir su yönetimi açısından kritik öneme sahiptir. Bu çalışmada, Batı Karadeniz Havzası’ndaki Filyos Çayı Alt Havzası’na ait su kalitesi verileri kullanılarak, AKM konsantrasyonunun tahmini için yığma modeli temelli bir yaklaşım önerilmiştir. Modelde, temel model olarak rasgele orman (RO) ve gradyan artırma (GA) algoritmaları, meta-model olarak ise K-en yakın komşu (KNN), çok katmanlı algılayıcı (ÇKA), kategorik artırma (CatBoost), uyarlamalı artırma (AdaBoost) ve Torbalama (Bagging) algoritmaları kullanılmıştır. Model performansı, ortalama karesel hata (MSE), kök ortalama kare hatası (RMSE), ortalama mutlak hata (MAE), ortalama mutlak yüzde hata (MAPE) ve korelasyon katsayısı (R²) gibi metriklerle değerlendirilmiştir. Sonuçlara göre, en yüksek R² değeri (%87) RO-GA+CatBoost modeli tarafından elde edilmiştir. Bu model, diğer kombinasyonlara kıyasla daha düşük hata değerleri (MSE: 143.85, RMSE: 11.99, MAE: 9.03, MAPE: 0.21) ile en iyi tahmin performansını sergilemiştir. Hiperparametre optimizasyonu sürecinde Izgara Arama (Grid Search) ve Rasgele Arama (Randomized Search) yöntemleri kullanılarak modellerin performansı daha da iyileştirilmiştir. Özellikle RO-GA+CatBoost modeli, optimizasyon sonrası R² değerini %89’a çıkararak en yüksek performansı göstermiştir. Bu çalışma, doğru algoritma seçimi ve hiperparametre optimizasyonunun model performansını artırmada kritik bir rol oynadığını ortaya koymuştur. CatBoost ve KNN algoritmaları, uygun hiperparametre optimizasyon teknikleri (örneğin Grid Search veya Randomized Search) ile entegre edildiğinde, modelin tahmin doğruluğu üzerinde anlamlı bir iyileşme sağlamaktadır. Gerçekleştirilen optimizasyon süreçleri, modelin karar sınırlarını veri setinin yapısına daha iyi uyarlamasına olanak tanıyarak genelleme performansını artırmakta ve aşırı öğrenme riskini azaltmaktadır. Elde edilen bulgular, benzer regresyon problemlerinde algoritma seçimi ve optimizasyon adımlarının titizlikle planlanması gerektiğini ortaya koymaktadır.