다중모형 및 격자기반 CMIP5 기후변화 시나리오 상세화 자료를 이용한 극한기후지수 변동성 분석

조재필; 김재욱; 최순군; 황세운; 정휘철

doi:10.15531/KSCCR.2020.11.2.123

주제분류

...

저널정보

한국기후변화학회
한국기후변화학회지 학술저널
Journal of Climate Change Research Vol.11 No.2
2020.4 123 - 132 (10page)
DOI : 10.15531/KSCCR.2020.11.2.123

저자정보

조재필 (유역통합관리연구원)
김재욱 (한국환경정책·평가연구원)
최순군 (국립농업과학원)
황세운 (경상대학교)
정휘철 (한국환경정책·평가연구원)

이용수
내서재: 1

내서재에 추가
되었습니다. 내서재에서
삭제되었습니다.

초록·키워드

오류제보하기

Extreme climatic events have increased in frequency and intensity over tha last decades due to global warming. Our study evaluates spatial distribution of changes in climate extreme indices on the Korean Peninsula by downscaled high-resolution data based on an empirical quantile mapping technique and 29 global climate models. Grid‐based observation data with 3.0 km resolution derived from weather stations with data covering 30 years from 1976 to 2005 using a modified PRISM approach were used as reference data for bias-correction. Future projections until 2100 based on two Representative Concentration Pathway (RCP) scenarios of CMIP5 were considered for the variables of daily precipitation, minimum temperature, and maximum temperature. We finally estimated spatial changes in climate extreme indices at 3.0 km resolution. The reproducibility assessment of simple precipitation intensity index (SDII) and annual total precipitation in wet days (PRCPTOT) showed applicability of techniques for downscaling, with the biggest difference of 2.66% and 1.91%, respectively, compared to the observation. The annual maximum 5-day precipitation (Rx5day) and annual maximum value of maximum temperature (TXx) showed highest increase rate under the far future period and RCP8.5 scenario. The derived spatial distribution of climate extreme indices based on a multi‐model ensemble can contribute to vulnerability assessment at a national scale by reducing uncertainties caused by selecting a specific climate model.

#Quantile Mapping (QM) #ETCCDI #Representative Concentration Pathway (RCP) #Multi-Model Ensemble (MME) #Parameter-elevation Relationship on Independent Slopes Model (PRISM)