Impact of Self-Calibrating on the Effective Drought Index: A Case Study of the south Korean Peninsula

Effective Drought Index (EDI) was proposed to monitor daily propagations of an emerging drought. The EDI, like other drought indices, uses the last 30 years of the daily precipitation record for the reference period for effective precipitation (EP) climatology, resulting that the drought characteristics can be solely estimated by a recent climatology. To overcome this weakness, this study proposes a self-calibrating EDI, a modified EDI with time-varying EP climatology via the 30-year moving time windows. In this study, the scEDI is calculated from the 240-year daily precipitation records (1777–2020) in Seoul, the south Korean Peninsula, and is compared with and the EDIs with different reference periods. The scEDI successfully adapts multi-decadal variability of precipitation, leading to robust (temporally consistent) estimates of drought severity while the EDIs, particularly with the 1885–1915 (dry) and 1990–2020 (wet) reference periods, over- and under-estimate drought severity, respectively. Furthermore, the droughts estimated by the scEDI are compared with the drought damage records in the Annals of the Joseon Dynasty (1778–1907) and the recent search frequency about droughts in Google and NAVER portals (2016–18) to investigate scEDI threshold values that is linked to actual socioeconomic impacts or favorable drought stages for a high level of drought awareness. Results confirmed that -1.0  and -2.0 of the scEDI can be good threshold values to detect severe droughts that cause socioeconomic impacts for agrarian and industrial societies, respectively. It is also found that the persistent and recovery stage of recent droughts surged the Internet search activities while public interest in drought was low during the onset stage. The findings of this study suggest that the importance of self-calibrating on improving the EDI-based drought assessment.