Investigating the causal impacts relationship between economic flood damage and extreme precipitation indices based on ARDL-ECM framework: A case study of Chungcheong region in South Korea

Abstract

The attribution of climate change to various natural disasters, including flood events, has continued to receive a considerable attention as it impacts the socioeconomic sustainability of society. However, the existing knowledge on the influential role of extreme precipitation on flood events and subsequent economic damage is limited. Thus, this study investigates the causal relationship between flood damage cost, flooded area, and extreme precipitation indices in the Chungcheong region of South Korea using the autoregressive distributed lag-error correction model (ARDL-ECM) framework and pairwise Granger causality analysis. Four extreme precipitation indices: consecutive wet days (CWD), number of very heavy precipitation days (R30mm), maximum 1-day precipitation amount (Rx1day), and simple daily precipitation intensity (SDII) that measure precipitation frequency, intensity, and duration were selected with time series data on the flooded areas and estimated flood damage cost from 1985 to 2020. The ARDL-ECM bound test indicates an existence of a long-run relationship among all the variables. The empirical results of pairwise Granger causality analysis further reveal that flooded area, R30mm, and Rx1day have significant positive causal impacts on the flood damage cost in both short and long-runs, in the current period. This implies that any increase in any of these variables will cause an increase in the flood damage cost. Unidirectional causality exists from the flooded area, R30mm, Rx1days, and SDII to flood damage cost, and from R30mm and SDII to flooded area. These extreme precipitation variables could serve as indicators of flood events and economic flood damage. This study advances the knowledge of the causal link between extreme precipitation indices and economic damage from flood events.