Understanding the hydrological responses of semi-arid watersheds to climatic variability is crucial for sustainable water resource management. This study demonstrates that integrating remote-sensing products with traditional hydrological indices offers a robust approach for comprehensive drought monitoring . An approach using multi-index was adopted by combining different remote-sensing products including the Automated Water Extraction Index (AWEI), the Normalized Difference Water Index (NDWI), and the Crop Water Stress Index (CWSI) computed over 10 years (2013–2022) using Google Earth Engine (GEE), with hydrological metrics such as the Standardized Runoff Index (SRI), the Streamflow Drought Index (SDI), and the Standardized Precipitation–Evapotranspiration Index (SPEI) analyzed over 44 years (1979–2022). Our findings reveal that remote-sensing products, particularly AWEI and NDWI, align most strongly with hydrological drought indices over 6 and 12 months, with AWEI–SRI/SDI correlations reaching r = 0.51/0.60 at 6 months and NDWI–SRI/SDI reaching r = 0.52/0.50 at 12 months. Although SPEI exhibits modest correlations with AWEI and NDWI (not exceeding r = 0.50), its integration with remote-sensing hydrological indices remains secondary. Trend analyses indicate significant declines in surface water extent during the study period: AWEI decreased by 1.03 km2/year and NDWI by 0.73 km2/year, underscoring intensifying drought conditions. These results highlight the indispensable role of remote sensing in delivering continuous, detailed spatio-temporal insights and supporting adaptive water governance strategies, particularly in regions facing increasing water scarcity.

Poor land management coupled with increasing climate extremes is affecting the livelihoods of Ethiopian communities. So, evaluating the possible impact of climate change on water resource is essential for planning a sustainable water use system. As a result, the aim of this study was to investigate impact of climatic variable changes on streamflow in Borkena River Catchment, Awash Basin, Ethiopia using projected climate data of canESM2 (Canadian Earth System Model of second generation) global climate model. Future scenario, analysis was performed for the 2020s, 2050s, and 2080s for the Representative Concentration Pathway of RCP4.5 and RCP8.5. Impact assessment of climatic variable changes on streamflow was done by HBV Light hydrological models. The model was calibrated using asemi–automated method and the performance was measured by the coefficient of determination (R²), Nash-Sutcliffe efficiency (NSE) and mean difference. The performance of HBV Light hydrological model during calibration and validation period showed a satisfactory agreement between observed and simulated flow with R², NSE and mean difference. The value of R², NSE and mean differences was 0.84, 0.82 and 0 for calibration and 0.83, 0.78 and 8.23 for validation periods, respectively. The downscaled precipitation and temperature result reveals decrease and increase value in all future three time horizon for both RCP8.5 and RCP4.5 scenarios in annual basis respectively. This condition was expected to result decreased in mean annual streamflow of Borkena River in the future three time horizon with the value of –16.81%,–17.25% and –18.28% for RCP8.5 scenario and –6.88%, –9.89%, and –10.66% for RCP4.5 scenario in the 2020s, 2050s, and 2080s, respectively. Inoculation future climate variation has significant impact on surface water potential of the catchment. So, for sustainable development of the River catchment in water use the streamflow of the catchment should be harvested.

Coastal and wetland regions play vital roles in ecosystem stability, economic development, and climate regulation, particularly along the rapidly changing coastline of eastern Asia. Timely and accurate mapping of these complex environments is essential for effective environmental monitoring and sustainable management. This study proposes a novel classification framework that integrates multi-sensor remote sensing data with deep feature extraction and machine learning algorithms to improve coastal and wetland mapping. High-resolution optical satellite images filtered by MLRSNet were used to classify six key coastal classes: beach, harbor port, wetland, river, lake, and island. Deep features were extracted using a pre-trained MobileNetV2 convolutional neural network and used to train Random Forest and Support Vector Classifier (SVC) models. Experimental results demonstrate that combining deep feature representations with multi-sensor data significantly enhances classification accuracy. The SVC model achieved a high overall classification accuracy of 97.3%, with nearly perfect performance in harbor port classification (F1-score = 0.999), along with high F1-scores for beach (0.986), island (0.981), river (0.960), lake (0.959), and wetland (0.952) classes. Despite the complex nature of wetland areas, the classifier maintained balanced performance across all classes, with macro-averaged accuracy, recall, and F1-score all reaching 0.973. These findings highlight the effectiveness of deep feature embeddings combined with SVC's non-linear decision boundaries for distinguishing natural and man-made coastal components. The framework shows strong potential for large-scale coastal ecosystem mapping and monitoring, even under subtle spectral differences, supporting future environmental management applications. Furthermore, the integration of Google Earth Engine (GEE) enables scalable visualization and operational deployment, demonstrating the framework's suitability for real-world coastal monitoring scenarios.