In this study, rainfall characteristics with stationary and non-stationary perspectives were analyzed using generalized extreme value (GEV) distribution and Gumbel distribution models with rainfall data collected in major cities of Korea to reevaluate the return period of sewer flooding in those cities. As a result, the probable rainfall for GEV and Gumbel distribution in non-stationary state both increased with time(t), compared to the stationary probable rainfall. Considering the reliability of ξ1, a variable reflecting the increase of storm events due to climate change, the reliability of the rainfall duration for Seoul, Daegu, and Gwangju in the GEV distribution was over 90%, indicating that the probability of rainfall increase was high. As for the Gumbel distribution, Wonju, Daegu, and Gwangju showed the higher reliability while Daejeon showed the lower reliability than the other cities. In addition, application of the maximum annual rainfall change rate (ξ1·t) to the location parameter made possible the prediction of return period by time, therefore leading to the evaluation of design recurrence interval.

• 박규홍(중앙대학교 사회기반시스템공학부) | Kyoohong Park (Dept. of Civil Engineering at Chung-Ang University) Corresponding author
• 유순유(중앙대학교 사회기반시스템공학부) | Soonyu Yu (Dept. of Civil Engineering at Chung-Ang University)
• 뱜바도지엘베자르갈(울란바타르 상하수도국) | Elbegjargal Byambadorj (Water supply and sewerage authority in Ulanbataar City)