강우량 자료와 SPEI를 이용한 우리나라 습윤 및 건조 조건의 장기변동 경향성 분석 = Long-term Trend Analysis of Wet and Dry Conditions Using Rainfall Data and SPEI in South Korea

다국어 초록 (Multilingual Abstract)

Long-term Trend Analysis of Wet and Dry Conditions Using Rainfall Data and SPEI in South Korea

Kwon, Gi Ryang

Department of Civil & Environmental Engineering
Graduate School
Gyeongkuk National University

Abstract

Global climate change has led to increasingly frequent extreme precipitation events and prolonged droughts worldwide. In particular, changes in the hydrological cycle have intensified the spatial and temporal variability of precipitation, simultaneously aggravating the contrasting phenomena of floods and droughts. In this era of climate change, comprehensive analyses from long-term and large-scale perspectives are required. Therefore, this study aims to identify the overall variation in the hydrological cycle by integrating analyses of wet and dry conditions, thereby enhancing the understanding of the complex characteristics of extreme weather events.
The normality of precipitation data and the Standardized Precipitation Evapotranspiration Index (SPEI) was examined to verify whether the datasets satisfy normal distribution according to their periods and characteristics. The D'Agostino–Pearson and Shapiro–Wilk tests were applied for normality assessment. Once normality was determined, various statistical methods were used to detect long-term trends in the data. These methods were categorized into parametric and non-parametric tests. A linear regression model was used as the parametric approach, which assumes independence and normality of observations. However, because hydrological time series data often violate normality assumptions, non-parametric tests are frequently applied. Among them, the Mann–Kendall and Spearman’s rho (ρ) tests are the most widely used for detecting trends in hydrological time series.
In this study, long-term trends of precipitation and SPEI were analyzed. Both annual and rainy seasonal precipitation showed slightly increasing trends, though not statistically significant at the 10% level. Similarly, most monthly precipitation data failed to reach statistical significance; however, a general decreasing trend was observed in June, while July, August, and September exhibited increasing tendencies. The long-term trend analysis of SPEI also revealed distinct regional characteristics. Except for a slight worsening trend in short-term (3–6 months) minimum SPEI, no significant changes were found in mean, maximum, or long-term (9–12 months) SPEI. Nonetheless, some localized areas exhibited statistically significant variations, and their spatial patterns were identified.
Although extreme rainfall and heatwave events associated with climate change are increasingly frequent worldwide, only a few regions in Korea demonstrated statistically significant long-term trends. This suggests that Korea has not yet experienced changes strong enough to affect long-term climatic trends. However, as climate change continues to progress, ongoing monitoring and research will be essential to reflect these evolving patterns.

Keywords: precipitation, rainfall, SPEI, trend, normality, Mann-Kendall test.

목차 (Table of Contents)

• 제 1 장 서 론 1
• 1.1 연구 배경 및 목적 1
• 1.2 연구동향 3
• 1.3 연구내용 6
• 제 2 장 대상지역 및 자료 8
• 2.1 대상지역 8
• 2.1.1 우리나라의 기후특성 8
• 2.1.2 우리나라의 강우특성 9
• 2.2 사용자료· 11
• 제 3 장 연구방법· 15
• 3.1 정규성 검정 15
• 3.1.1 Shapiro-Wilk 검정 15
• 3.1.2 D'Agostino-Pearson 검정 16
• 3.2 경향성 분석 18
• 3.2.1 모수적 방법 18
• 3.2.2 비모수적 방법 20
• 3.3 표준화 강수-증발산 지수· 24
• 3.3.1 SPEI 24
• 3.3.2 SPEI의 산정 방법 25
• 제 4 장 적용 및 분석 29
• 4.1 강우량 분석 29
• 4.1.1 기초 통계 분석· 29
• 4.1.2 강우자료의 정규성 분석 35
• 4.1.3 강우자료의 경향성 분석 42
• 4.2 SPEI 분석 56
• 4.2.1 SPEI의 산정· 56
• 4.2.2 SPEI 자료의 정규성 분석· 58
• 4.2.3 SPEI 자료의 경향성 분석· 63
• 제 5 장 토론 및 고찰 78
• 5.1 강우량 해석 78
• 5.1.1 강우량의 정규성 해석· 78
• 5.1.2 강우량의 경향성 해석· 82
• 5.2 SPEI 해석 84
• 5.2.1 SPEI의 정규성 해석 84
• 5.2.2 SPEI의 경향성 해석 86
• 5.3 극한 건조화 위험도 분석· 88
• 5.3.1 위험도 계산 88
• 5.3.2 위험도의 지역적 특성· 90
• 5.4 실용적·정책적 시사점 94
• 제 6 장 결 론· 98
• 참고문헌 101
• 부 록 107
• Abstract 154