In general, water-related hazards are caused by natural phenomena but sometimes anthropogenic factors from insufficient water resource management can become causes of water-related hazards. The presence and the amount of rainfall, among the natural causes, are often the direct causes of flood. Therefore, accurate rainfall estimation and rainfall occurrence time prediction are one of the ways to prevent and mitigate the damage caused by water-related hazards. However, as various uncertainties exist in rainfall forecasting, it is necessary to understand and reduce these uncertainties. Recently, with the rapid advancement in computer technology, the quality of rainfall prediction data has improved, as well as the rainfall-runoff prediction. In consideration of these points, this study attempts to investigate the uncertainty of the Korean rainfall ensemble prediction data and runoff analysis model to enhance reliability and to examine the possibility of improving the prediction by applying the bias correction technique. Therefore, the objectives of this study are: 1) to evaluate the spatial characteristics and applicability of LENS (Limited Area Ensemble Prediction System) data by calculating and comparing the point value of the rainfall observation station and the average of the areal value with respect to the actual amount of rainfall; 2) to evaluate whether the rainfall forecasting ability is improved after applying the bias correction technique to LENS data; 3) to understand the uncertainty using parameter correction and GLUE for the rainfall using the GRM (Grid-based Rainfall-Runoff Model); and 4) to apply and evaluate models before and after LENS-GRM correction. To achieve these goals, this study was conducted using the data from the Geumho River Basin, the Miryang River Basin, and the Wicheon Basin. In addition, informal likelihood, R2, NSE, PBIAS, and formal likelihood, lognormal, were used to evaluate applicability and uncertainty analysis for each goal. The results of this study confirmed the influence of the temporal uncertainty of rainfall forecasting of LENS data and the applicability of bias correction on rainfall and runoff prediction. Furthermore, this study was able to confirm the uncertainty of behavior model selection using the threshold of likelihood when applying the runoff model of rainfall forecasting data. Accordingly, by reducing the uncertainties of rainfall ensemble data and the runoff model when selecting the behavioral model of the user''s uncertainty analysis, our method is expected to deliver a more reliable flood prediction. In addition, the findings of this study are expected to be used as a basis for flood prediction studies that apply rainfall and geographical characteristics and expand analysis targets.