A low-cost hydroinformatics tool for monitoring inter-seasonal water availability in small earth dams: evidence from a data-scarce small sub-basin in Southern Africa

ABSTRACT

Small earth dams play a critical role in sustaining rural livelihoods and enhancing water and food security in semi-arid regions such as southern Zambia. However, their effective management is constrained by limited hydrological monitoring infrastructure and the absence of cost effective, user-friendly tools for tracking water availability. This study developed and evaluated a low-cost hydroinformatics framework that integrates bathymetric surveying, empirical modelling, and an Android-based mobile application to monitor inter-seasonal water availability in small earth dams. Using Makoye Earth Dam in Southern Zambia as a case study, bathymetric data were collected across rainy, cool-dry, and hot-dry seasons to establish depth–volume relationships. An empirical prediction model was trained and validated using standard performance metrics including R2, Nash–Sutcliffe efficiency (NSE), percent bias, root mean square error (RMSE), and RMSE–standard deviation ratio. The model was subsequently embedded into a mobile application (Madwapp 0.1) to enable real-time estimation of reservoir storage from field-measured water depths. Results show that the empirical model performed exceptionally well during the rainy season (R2 = 0.99; NSE = 0.99), demonstrating high predictive accuracy and minimal error. However, performance declined in the cool and dry seasons due to systematic underestimation linked to bathymetric complexity and seasonal water variability. Validation across multiple dams indicated that the volume–volume empirical model significantly outperforms conventional depth–volume based estimation methods, achieving near-perfect agreement with observed volumes (R2 = 0.99). The study highlights the strong seasonality and rapid depletion dynamics of small dams and provides a practical, scalable, and low-cost solution for improving water resource monitoring and decision-making in small to medium data-scarce rural catchments. The empirical regression model performance was strong during recharge conditions but deteriorated substantially during cool and hot-dry seasons, limiting direct transfer without seasonal recalibration. The study recommends multi-basin–based Android app development for enhanced water monitoring efficiency.