To understand the influence of rainfall patterns on rainfall–runoff processes, we propose two indices: skewness_PEAK (skew_p), representing the relative timing of peak rainfall, and the normalized root-mean-square error peak (NRMSE_p), which quantifies the concentration of rainfall near the peak. By analyzing approximately 25,000 rainfall scenarios, we examined the relationship between these indices and peak flood discharge in the rainfall–runoff process. The analysis revealed that peak flood discharge positively correlates with the NRMSE_p, indicating that concentrated rainfall near the peak substantially increases discharge. Conversely, a negative correlation with skew_p suggests that earlier peak rainfall reduces discharge. These insights were synthesized into a three-dimensional solution space providing a comprehensive framework for predicting how variations in rainfall distribution affect flood discharge. The findings underscore the importance of incorporating these indices into real-time flood forecasting models and urban flood risk management strategies.

We revisit empirical methods to prevent the overestimation of peak discharge in a small watershed, in particular investigating the time-area method, which has not been considered in the overestimation problem of peak discharge. To avoid misapplying the same inlet time between the unit hydrograph and rational formula, distinct parameter adjustments for each method are proposed. We adopt the secondary basin response time for the unit hydrograph, rainfall duration for the rational formula, and time of concentration for the time-area method, as suitable parameters to adjust the estimation of peak discharge. In conclusion, adding 10 minutes to secondary basin response time, 20 minutes to rainfall duration, and 30 minutes to time of concentration, respectively, yields estimates within a reasonable range of specific discharge in a small watershed.

In practice, methods arbitrarily increasing probable design rainfall or return periods have been adopted to derive design flood increases due to climate change. However, the method of arbitrarily increasing probable design rainfall is inadequate because probable rainfall estimated by the common frequency analysis method usually cannot be accepted in trend tests, meaning there is no increasing trend due to climate change. In addition, the method of increasing return periods to increase design flood is problematic. Rising flood damages—an opportunity to recognize climate change—are caused by concentrated flooding due to urbanization rather than increased probable rainfall. Thus, we proposed new methods for considering climate change in design flood estimation, by changing the temporal distribution of design rainfall or adjusting parameters in the rainfall-runoff process. Furthermore, the method of raising the freeboard criteria is proposed as an alternative for considering climate change.

A representative example of how extreme rainfall events caused by climate change can be directly recognized is the increase of property damage due to urban flood. This surge in urban flood damage is not merely corresponding to an augmentation in probable rainfall depth but rather a result of urbanization with densely populated and concentrated flooding. Despite this circumstance, we have mostly revolved around increasing the return period, primarily focusing on the augmentation of probable rainfall depth to mitigate the flood damage by climate change. In terms of the temporal distribution of design rainfall, conventional methods such as the Huff method and alternating blocking method are still in use; however, they cannot accurately capture the real rainfall distribution. In this context, we investigate the viability of enhancing design standards by adjusting the temporal distribution of design rainfall without artificially inflating a return period for design. To achieve that, it is necessary to understand the impact of temporal rainfall distribution on the design flood.We generate 449 dimensionless time-rainfall distributions for short-term(less than 6 hours) and 5,789 for long-term(6 hours or more) durations considering rainfall data from both meteorological observations by the Korea Meteorological Administration and d4PDF(Data for Policy Decision Making for Future) scenarios. Based on these distributions, total 25,860 hyetographs are synthesized for five rainfall durations. We repeatedly estimated the design flood using a rainfall-runoff model, revealing that the peak discharges varied up to 10 times depending on the time-rainfall distribution. Considering that increasing the return period from 50 years to 100 years generally results in only a 10% rise in probable rainfall depths, adjusting the temporal distribution of design rainfall provides a more adaptable approach to increasing design floods. These outcomes have the potential to broaden the perspective for applications of rainfall scenarios in data-based model or establishing the design flood standards. Acknowledgement: This research was supported by a grant(2022-MOIS61-002(RS-2022-ND 634021)) of ‘Development Risk Prediction Technology of Storm and Flood for Climate Change based on Artificial Intelligence’ funded by Ministry of Interior and Safety(MOIS, Korea).