지상관측자료 기반 강수 재분석자료의 우기(5~9월) 강우 특성 산출 수준 평가

Abstract

This study evaluated the reliability and applicability of high-resolution (5km) precipitation reanalysis data (Re_Da) developed by the Korea Meteorological Administration (KMA) by comparing it with surface observation data (SOD) in South Korea over a six-year period (2018-2023). Focusing on the rainy season (May-September), this study examined rainfall amounts (RA), number of rainy days (RD), rainfall intensity (RI), and the occurrence and intensity of extreme rainfall events at the daily, monthly, and yearly level. Our results show that Re_Da accurately reproduces the spatial distribution and temporal variations of rainfall during the rainy season, in close agreement with observations. However, Re_Da tends to overestimate RA and RD, while underestimating RI. Since 2021, the overestimation of RA and RD has increased, whereas RI has shown no change. Despite these biases, there is a strong correlation between the Re_Da and the SOD, with correlation coefficients of 0.98, 0.94, and 0.97 for RA, RD, and RI, respectively. Diurnal variation is well-represented, with correlation coefficients exceeding 0.99 across all variables. When evaluating the frequency and intensity of extreme rainfall events based on the KMA’s heavy rain advisories and warnings, the Re_Da reproduced extreme rainfall events similarly to observations. These results support the utility of this dataset in hydrology, agriculture, and climatology, but the current resolution is limited to regional extreme weather analysis. To improve the characterization of regional extreme rainfall and disaster prediction, improved spatial and temporal resolution along with improved accuracy of precipitation reanalysis data are needed.