Groundwater is an invaluable natural resource that sustains human life and supports the economic development of nations. However, its unsustainable utilization has emerged as a critical issue, particularly in developing countries. This study investigates the groundwater potential of the Chemoga watershed to address these challenges. Conventional groundwater assessments have typically relied on labor-intensive and time-consuming field surveys, which are resource-demanding and often fail to provide accurate estimates due to the inherent complexity of groundwater systems. In response, this research utilizes geospatial and analytic hierarchy process (AHP) techniques to assess groundwater potential in the Chemoga Watershed, aiming to overcome these challenges. Eight critical biophysical and environmental factors: geology, slope, rainfall, land use/land cover (LULC), soil type, elevation, lineament density, and drainage density were selected for analysis using Saaty’s AHP methodology. Data was gathered from satellite imagery, existing thematic maps, local water offices, and national meteorological agencies. The integration of these thematic maps was performed through a weighted overlay analysis using ArcGIS 10.8, which resulted in the delineation of groundwater potential zones (GWPZ). The model was validated by cross-referencing the generated GWPZ with existing data from dug wells and boreholes. The results reveal five groundwater potential zones: very high (0.73%), high (24.39%), moderate (43.38%), poor (31.25%), and very poor (0.25%). The most suitable zones are in the south, southeast, and southwest of the watershed, particularly near Debre Markos Town. These high-potential zones were validated with a significant 81.5% match to ground truth data from shallow wells. The findings of this study provide crucial insights for decision-makers, enabling the formulation of more effective groundwater management strategies. By identifying cost-effective and suitable well sites, this research contributes to ensuring a sustainable water supply for Debre Markos Town.

Excessive runoff in floodplain areas, intensified by climate change and urbanization, poses a significant challenge, particularly during the rainy season. Accurate estimation of rainfall-induced runoff is crucial for effective watershed management and sustainable water resource utilization. This study applies the Soil Conservation Service-Curve Number (SCS-CN) method, integrated with Geographic Information Systems (GIS), to estimate runoff in the Wuseta Watershed, Upper Blue Nile Basin, Ethiopia. The SCS-CN method’s simplicity and reliability make it an effective tool for assessing runoff dynamics. Various datasets including rainfall, land use/land cover, soil characteristics, and Hydrologic Soil Group (HSG) data were utilized, with Antecedent Moisture Conditions (AMC) incorporated to capture runoff variability. Weighted Curve Numbers (C.N.) were computed for dry, normal, and wet conditions over a 20-year period, revealing significant runoff variability. The highest annual rainfall of 1,458.00 mm occurred in 2009, while the lowest (1,256.21 mm) was recorded in 2005, with an average of 1,358.84 mm. Correspondingly, peak runoff reached 458.6 mm in 2009, while the lowest was 261.43 mm in 2015. Over the study period, total rainfall amounted to 28,487.79 mm, generating 7,364.59 mm of runoff, accounting for 25.90% of total precipitation. The average runoff depth was 358.76 mm, exhibiting a strong correlation with rainfall (R² = 0.81). Flood frequency analysis using the Gumbel method indicated an increasing trend in peak discharge with return periods, ranging from 11.35 m³/s for a 5-year return period to 19.90 m³/s for a 100-year return period. These findings are essential for flood risk management, watershed planning, and the development of effective flood mitigation strategies. This study underscores the importance of accurate runoff estimation and rainfall-runoff analysis for sustainable water resource management and flood risk reduction.

Erosion-induced land degradation remains a persistent barrier to agricultural productivity and sustainability for smallholder farmers in Africa, where rapid population growth has intensified land pressure, especially in Ethiopia. Despite the implementation of several sustainable land management (SLM) initiatives, sustaining farmers' motivation and voluntary adoption remain problematic. This study examined the determinants of farmers' intention (INT) to adopt three SLM technologies-compost making, forage planting, and soil bund making-using an extended theory of planned behavior framework. A measurement model incorporating psychological constructs and contextual factors was developed and tested via partial least squares-structural equation modeling based on data from 202 respondents in the Ethiopian highlands. Attitude (ATT) significantly influenced adoption INT across all technologies (β = 0.38-0.64), while Perceived Behavioral Control (PBC) was crucial for forage (β = 0.43), a recently introduced practice. Subjective Norms affected INT indirectly through influencing PBC and ATT for compost and forage (β = 0.25-0.28). Interpersonal information-primarily from extension officers, relatives, and other farmers-emerged as the strongest contextual influence (β = 0.21-0.26), whereas access to media-based information was limited. Mobile phone ownership, experience with the SLM practices, prior-year profits, and land size had technology-specific effects. The findings provide practical insights to inform SLM promotion strategies through situational assessment and targeted interventions that enhance farmers' awareness and motivation for voluntary adoption.

Flood occurrence in tropical regions is intensifying due to climate variability and land-use change, increasing the need for reliable flood response time estimation. Accurate prediction of flood lag time (TL)—the interval between the centroid of excess rainfall and peak runoff—is critical for flood early warning and water resource planning. However, TL estimation remains challenging in data-scarce regions because of complex interactions among watershed morphology, rainfall characteristics, and runoff generation processes. This study evaluates four machine learning (ML) algorithms—Random Forest (RF), Artificial Neural Network (ANN), Extreme Gradient Boosting (XGBoost), and Support Vector Machine (SVM)—for predicting TL across twenty gauged watersheds in the Blue Nile Basin of Ethiopia. Fourteen physiographic and hydro-climatic watershed characteristics were used as predictors. Among the tested models, XGBoost achieved the highest training performance (R² = 0.98, NSE = 0.96), while RF showed better generalization in the test dataset (R² = 0.77, NSE = 0.70, KGE = 0.71). SVM produced the lowest prediction errors (MAE = 0.95; RMSE = 2.25) but had lower explanatory power (R² = 0.49). To enhance interpretability and practical applicability, ML-based feature importance was used to develop a parsimonious empirical model: TL = 0.8 + 0.011A − 0.023RI, where A is watershed area and RI is rainfall intensity. This model explained 51% of TL variability and retained much of the predictive skill of more complex ML models. The proposed hybrid ML–empirical framework provides a transparent and operational approach for flood response time estimation in tropical highland watersheds. Its broader applicability remains subject to additional watershed-level validation and regional calibration.