GR modellerinde uydu görüntüleri kullanarak hidrolojik modelleme

Water scarcity is a critical issue that needs attention to achieve sustainability. Remote sensing technology has advanced in recent decades, providing important data on hydrological variables for effective water resource management. This study analyzed two regions with different hydroclimatic conditions: Türkiye, with a temperate climate where snowmelt significantly contributes to streamflow, and Guinea, with a humid tropical climate characterized by monsoon rains. Rainfall-runoff modeling was performed at daily, monthly, and annual time steps using the split-sample test method and single objective calibration. Three algorithms—Calibration Michel, DEoptim, and hydroPSO— were used at the daily time step to effectively explore the parameter space. Satisfactory results were achieved for both regions at all time steps, although the standard CemaNeige for snow accounting and GR1A in the Guinean watershed were less satisfactory. A multi objective calibration was then conducted to further refine the daily models by considering hysteresis between snow accumulation and melting phases in snowy basins and soil moisture in others. This approach aimed to enhance model accuracy and robustness by incorporating multiple hydrological variables, providing a comprehensive understanding of the water balance in these regions. By considering hysteresis in the CemaNeige snow module, snow cover dynamics were accurately simulated without compromising discharge performance. Additionally, when both soil moisture and discharge were considered in the GR4J model, simulations of soil moisture closely matched SMAP data during the rainy season, especially with increased weighting on soil moisture.