Uncertainty in projections of future runoff depending on hydrological model calibration data

This study analyzed the impact of the length of calibration data and hydrological conditions in hydrological models on the uncertainty of future runoff projections. For this purpose, the upstream area of the Andong Dam was selected as the study area. Calibration data lengths ranging from 1 to 10 years and hydrological conditions (dry, normal, and wet) were set using observed inflow data from 1990 to 2023. Future climate data for runoff projections were constructed using 20 Coupled Model Intercomparison Project 6(CMIP) General circulation model (GCM)s and Shared Socioeconomic Pathway (SSP) scenarios (SSP2-4.5, SSP3-7.0, and SSP5-8.5), and bias correction was performed using the quantile mapping method. Subsequently, the uncertainty of the projected future runoff was analyzed using ANalysis Of VAriance (ANOVA) to evaluate the contributions of SSP, GCM, calibration data length, and hydrological conditions of the calibration data. As a result, it was found that the hydrological conditions of the calibration data had a significant impact on the validation performance of the hydrological model, while the influence of the calibration data length was relatively insignificant. The uncertainty in future runoff projections was mainly due to the General circulation model (GCM)s, contributing more than 70%, and the influence of the hydrological model calibration data varied depending on the season and period. The uncertainty due to the hydrological model calibration data contributed significantly (over 8%) during spring, when runoff is relatively low, and generally had a large impact during months with low runoff. This study provides foundational data that can contribute to the planning and development of future water resource management strategies using hydrological models.